228 86 54MB
English Pages 758 [4501] Year 2010
Handbook of Disease Burdens and Quality of Life Measures
Victor R. Preedy, Ronald R. Watson (Eds.)
Handbook of Disease Burdens and Quality of Life Measures
With 570 Figures and 1001 Tables
Editors: Prof. Victor R. Preedy Dept. Nutrition and Dietetics Nutritional Sciences Research Division School of Biomedical & Health Sciences King’s College London Franklin‐Wilkins Building 150 Stamford Street London SE1 9NH UK Prof. Ronald R. Watson Mel and Enid Zuckerman College of Public Health University of Arizona Health Science Center 1295 N. Martin P.O. Box 245155 Tucson, AZ 85724–5155
ISBN–13: 978–0–387–78664–3 This publication is available also as: Electronic publication under 978–0–387–78665–0 and Print and electronic bundle under ISBN 978–0–387–78666–7 Library of Congress Control Number: 2009927296 ß Springer Science+Business Media LLC 2010 (USA) All rights are reserved. This work may not be translated or copied in whole or in part without the written permission of the publisher (Springer Science+Business Media, LLC, 233 Spring Street, New York, NY 10013, USA), except for brief excerpts in connection with reviews or scholarly analysis. Use in connection with any form of information storage and retrieval, electronic adaption, computer software, or similar or dissimilar methodology now known or hereafter developed is forbidden. The use in this publication of trade names, trademarks, service marks, and similar terms, even if they are not identified as such, is not to be taken as an expression of opinion as to whether or not they are subject to proprietary rights. Product liability: The publishers cannot guarantee the accuracy of any information about the application of operative techniques and medications contained in this book. In every individual case the user must check such information by consulting the relevant literature. springer.com Printed on acid‐free paper
SPIN: 11980131
2109 –5 4 3 2 1
Preface
Disease and pathologies have devastating consequences on the individual, the family unit and society in general including effect on nations or entire geographical regions. Assessing these impacts at the individual or international level has been problematical due to the diverse nature of the diseases themselves. There is also a variable response between individuals or regions. However, there is now an increasing awareness that the imposition of disease or ill health can be measured and assessed in quantitative terms. These measures of disease encompass a variety of facets: from the financial costs of treatment to the effects of a particular disease condition on the quality of life of individuals. The development and application of tools to measure these aspects have been particularly evident over the past decade. For example, suitable questionnaires developed for the general assessment of Quality of Life have now been revised and refined so that they are directly applicable to specific disease entities or even different cultures. Hitherto, there has never been a coherent publication that allocated to a single volume the many questionnaires that have been developed nor the quantitative aspects of disease in terms of finance, mortality, morbidity and quality of life measures. These aspects are addressed in The Handbook of Disease Burdens and Quality of Life Measures. It is structured into 3 main sections as follows: Part [1] Instruments and methological aspects; Part [2] Disease Burdens and Economics Impacts; Part [3] Quality of life measures and indices. The various subsections of The Handbook of Disease Burdens and Quality of Life Measures reflect the diverse nature of diseases and their impact, Sections include those pertaining to geographical aspects of disease, pathologies and metabolic disorders, early life stages and aging, cancer, cardiovascular disease, immune disorders, viral, bacterial, microbiological, infectious and parasitic diseases, psychosocial, social, behavioural, psychiatric, neurological conditions and addictions to name just a few examples. In dividing the chapters the Editors recognise the problems that this entails. Some chapters may be equally at home in more than one subsection. To a certain extent this is covered by the comprehensive coverage and excellent indexing. Essentially, The Handbook of Disease Burdens and Quality of Life Measures represents a ‘‘one-stop-shopping’’ of information with suitable tables and figures. Each chapter is written by internationally or nationally recognised experts or Institutions. Each chapter is also ‘‘stand alone’’, self contained and well illustrated with appropriate tables and figures. The articles are written in such a way that material from one area can be readily transferable to other areas. In other words the material truly bridges the trans-disciplinary divide. The Handbook’s broad coverage and meticulous, up-to-date detail make it essential for public health researchers, medical and health practitioners, and for those involved in allocating resources and setting priorities, such as epidemiologists, sociologists, health economists, and policymakers. It is also suitable for those who specially want to broaden their knowledge-base in a rapidly expanding area of medical sciences. Professors Victor R Preedy and Ronald Ross Watson
#
Springer Science+Business Media LLC 2010 (USA)
Table of Contents
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxix
Volume 1 Part 1 Instruments and Methodological Aspects . . . . . . . . . . . 1 Part 1.1 Instruments Used in the Assessment of Disease Impact . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1
The International Classification of Functioning, Disability and Health: A Tool to Classify and Measure Functioning . . . . . . . . . . . . . . . . . . . . . . . . 3 G. Stucki . N. Kostanjsek . A. Cieza
2
The Keele Assessment of Participation . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 R. Wilkie
3
The Global Person Generated Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 F. Martin . L. Camfield . D. Ruta
4
The Total Illness Burden Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 S. Greenfield . J. Billimek . S. H. Kaplan
5
The EQ-5D Health-Related Quality of Life Questionnaire . . . . . . . . . . . . . 87 N. Gusi . P. R. Olivares . R. Rajendram
6
The University of Washington Quality of Life Scale . . . . . . . . . . . . . . . . . 101 S. N. Rogers . D. Lowe
7
Comparison of Three Quality of Life Questionnaires in Urinary Incontinence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129 Ja Hyeon Ku . Seung-June Oh
8
Overview of Instruments Used to Assess Quality of Life in Dentistry . . . . 145 C. McGrath . S. N. Rogers
9
SQOR-V: A Patient Reported Outcome Specifically Dedicated to Chronic Venous Disorders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161 J.-J. Guex . S. E. Zimmet . S. Boussetta . C. Taieb
viii
Table of Contents
10
The Uniscale Assessment of Quality of Life: Applications to Oncology . . . 179 E. Ballatori . F. Roila . B. Ruggeri . A. A. Bruno . S. Tiberti . F. di Orio
11
The Bone Metastases Quality of Life Questionnaire . . . . . . . . . . . . . . . . . 195 X. Badia . A. Vieta . M. Gilabert
12
The Impact of Weight on Quality of Life Questionnaire . . . . . . . . . . . . . . 209 J. Manwaring . D. Wilfley
13
The Quality in Later Life Questionnaire . . . . . . . . . . . . . . . . . . . . . . . . . . 227 S. Evans
14
The MacDQOL Individualized Measure of the Impact of Macular Disease on Quality of Life . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247 Jan Mitchell . Alison Woodcock . Clare Bradley
15
Development and Assessment of Chinese General Quality of Life Instrument . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265 Y. Wu . G. Xie
16
The Japanese Version of the EORTC Quality of Life Questionnaire . . . . . . 285 G. To´th . M. Tsukuda
Part 1.2 Contemporary Issues in Assessment . . . . . . . . . . . . 311 17
Calculating QALYs and DALYs: Methods and Applications to Fatal and Non-Fatal Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313 F. Sassi
18
Accuracy of Death Certifications and the Implications for Studying Disease Burdens . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 329 J. R. Pierce . A. V. Denison
19
Completeness and Accuracy of Death Dates and the Implications for Studying Disease Burdens: Focus on Alternative Data Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 345 Min-Woong Sohn . Elissa Oh
20
Utility Scores for Comorbid Conditions: Methodological Issues and Advances . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 359 C. N. McIntosh
21
Subjective Quality of Life Measures – General Principles and Concepts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 381 C. L. K. Lam
22
Standard Expected Years of Life Lost as a Measure of Disease Burden: An Investigation of Its Presentation, Meaning and Interpretation . . . . . . 401 R. J. Marshall
Table of Contents
Part 2 Disease Burdens and Economic Impacts . . . . . . . . . . . 415 Part 2.1 General Aspects, Geographical Aspects, Pathologies and Metabolic Disorders . . . . . . . . . . . 415 23
Health-Adjusted Life Expectancy: Concepts and Estimates . . . . . . . . . . . . 417 J. A. Labbe
24
Individual Disability-Adjusted Life Year: A Summary Health Outcome Indicator Used for Prospective Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . 425 P. Zhang . M. Woodward . J. Shen . Y. Wu
25
Integration of Quality of Life and Survival for Comparative Risk/Outcome Assessment in Healthcare Industry . . . . . . . . . . . . . . . . . . 437 J.-D. Wang . J.-S. Hwang
26
Disability-Adjusted Life Years in Occupational Injuries and Accidents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 447 M. Concha-Barrientos . J. Labbe´ Cid
27
The Income-Associated Burden of Disease in the United States . . . . . . . . 459 P. Muennig . M. Gold . E. Lubetkin . H. Jia
28
Financial Burdens and Disability-Adjusted Life Years in Los Angeles County . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 473 G. F. Kominski . P. A. Simon . A. Y. Ho . J. E. Fielding
29
Burden of Disease Between Two Time Frames: Mexico Perspectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 483 G. Rodriguez-Abrego . J. Escobedo-de la Pen˜a . R. B. Zurita Garza . T. Ramı´rez-Sa´nchez
30
East/West Differences in Health in Europe: Rates, Expectancies and DALYs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 505 J. Powles . H. Gouda
31
The Burden of Neglected Diseases in Developing Countries . . . . . . . . . . . 517 A. Boutayeb
32
The Burden of Communicable and Non-Communicable Diseases in Developing Countries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 531 A. Boutayeb
33
Economic Evaluation of Health Interventions: Tanzania Perspectives . . . . 547 B. Robberstad . Yusuf Hemed
34
The Burden of Disease and Injury in Serbia . . . . . . . . . . . . . . . . . . . . . . . 587 S. Jankovic´ . H. Vlajinac . V. Bjegovic´ . J. Marinkovic´ . S. Sˇipetic´-Grujicic´ . Markovic´-Denic´ . N. Kocev . M. Sˇantric´-Milic´evic´ . Z. Terzic´-Sˇupic´ . N. Maksimovic´ . U. Laaser
ix
x
Table of Contents
35
Disease Burdens and Disability-Adjusted Life Years in Aboriginal and Non-Aboriginal Populations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 603 Y. Zhao
36
The Relationship Between the National Health Insurance Expenditures and the Burden of Disease Measures in the Iran . . . . . . . . . . . . . . . . . . . . 629 M. Russel . H. R. Jamshidi
37
The Burden of Maternal Mortality and Morbidity in the United States and Worldwide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 647 C. T. Lang . J. C. King
38
The Financial Implications of Pancreas Transplant Complications . . . . . . 661 J. A. Cohn . M. J. Englesbe
39
The Use of Pharmacoepidemiological Databases to Assess Disease Burdens: Application to Diabetes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 671 H. Støvring
40
Years of Life Lost Due to Air Pollution in Switzerland: A Dynamic Exposure-Response Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 685 M. Ro¨o¨sli
41
Quantification of Deaths and DALYs Due to Chronic Exposure to Arsenic in Groundwaters Utilized for Drinking, Cooking and Irrigation of Food-Crops . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 701 D. A. Polya . D. Mondal . A. K. Giri
Part 2.2 Early Life Stages and Aging . . . . . . . . . . . . . . . . . . . . 729 42
Disability-Adjusted Life Years in Children and Adolescents in Europe . . . 731 F. Valent . S. Di Bartolomeo
43
Disease Burden of Diarrheal and Respiratory Disorders in Children: Hong Kong Perspectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 751 E. A. S. Nelson
44
Burden of Disease in Elderly Mexican Population . . . . . . . . . . . . . . . . . . . 763 B. Rico-Verdı´n . G. Rodriguez-Abrego . I. Villasen˜or-Ruiz . J. L. Torres-Cosme
Volume 2 Part 2.3 Cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 783 45
Years of Life Lost from Cancer and Applications to Research Funding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 785 N. G. Burnet . S. J. Jefferies . F. P. Treasure
Table of Contents
46
Worldwide Burden of Gynecological Cancer . . . . . . . . . . . . . . . . . . . . . . . 803 R. Sankaranarayanan . J. Ferlay
47
Years of Life Lost from Lung, Stomach, Liver and Cervical Cancers: An Evaluation of the Top Cancer Killers . . . . . . . . . . . . . . . . . . . . . . . . . . 825 B. Y. Goldstein . F. I. Bray . D. M. Parkin . J. W. Sellors . Z. F. Zhang
48
Burden of Cancer in Serbia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 843 H. Vlajinac . S. Sipetic-Grujicic . S. Jankovic . L. Markovic-Denic . J. Marinkovic
49
The Disease Burden of Mastectomy: Turkish Perspective and Impact on the Patient and Family . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 865 ¨ zkan . M. O ¨ zkan . V. O ¨ zmen . Z. Armay S. O
50
The Burden of Chemotherapy Induced Nausea and Vomiting on Patients’ Daily Lives: Italian Perspectives . . . . . . . . . . . . . . . . . . . . . . . . . 885 E. Ballatori . F. Roila . B. Ruggeri . A. A. Bruno . S. Tiberti . F. di Orio
Part 2.4 Cardiac, Vascular, Pulmonary and Dietary . . . . . . . 899 51
Atherosclerotic Burden and Mortality . . . . . . . . . . . . . . . . . . . . . . . . . . . . 901 J. Roquer . Angel Ois
52
Burden of Cardiovascular Diseases Among Aboriginal and Torres Strait Islander Peoples: Mortality, Hospitalization and Risk Factors . . . . . 919 A. G. Thrift . S. L. Gall . A. D. Brown
53
Burden of Ischemic Heart Diseases in Serbia . . . . . . . . . . . . . . . . . . . . . . 933 S. Sipetic-Grujicic . H. Vlajinac . J. Marinkovic . V. Bjegovic . I. Ratkov . J. Maksimovic
54
Burden of Cerebrovascular Diseases (Stroke) in Serbia . . . . . . . . . . . . . . . 949 T. Pekmezovic . H. Vlajinac . S. Sipetic-Grujicic . N. Kocev . D. K. Tepavcevic . L. B. Bumbasirevic
55
DALYs and Public Health Programs for Stroke: Australian Perspectives . . . 965 D. A. Cadilhac . M. L. Moodie . E. E. Lalor
56
Burden of Stroke: Indian Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . 991 P. M. Dalal . M. Bhattacharjee
57
Disability-Adjusted Life Years, Years of Life Lived with Disability, and Years of Life Lost in Stroke: Italian Perspectives . . . . . . . . . . . . . . . 1007 S. Mariotti
58
Burden of Ischemic Stroke and Benefits of Stroke Unit Care and Thrombolytic Therapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1029 A. Meretoja . M. Kaste . T. Tatlisumak
xi
xii
Table of Contents
59
Economic Burden of Complications During Percutaneous Coronary Intervention . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1061 M. Dewan . K. M. Jacobson . C. S. Rihal
60
Features of Mediterranean Diet and Burden of Cardiovascular Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1073 D. B. Panagiotakos . C. Pitsavos . D. P. Mikhailidis
61
Obesity’s Final Toll: Influence on Mortality Rate, Attributable Deaths, Years of Life Lost and Population Life Expectancy . . . . . . . . . . . 1085 K. R. Fontaine . S. W. Keith . J. A. Greenberg . S. J. Olshansky . D. B. Allison
62
Financial Impact of Obesity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1107 L. Barrios . D. B. Jones
63
Burden of Disease Attributable to Obesity and Overweight: Korean Focus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1119 Seok-Jun Yoon . Jae-Hyun Park
64
Economic Burden of the Components of the Metabolic Syndrome . . . . . 1135 P. J. Marangos . L. J. Okamoto . J. J. Caro
65
Zinc Deficiency and DALYs in India: Impact Assessment and Economic Analyses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1151 A. J. Stein . M. Qaim . P. Nestel
Part 2.5 Immune Disorders, Viral, Bacterial, Microbiological, Infectious and Parasitic Diseases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1171 66
The Impact of Infectious Diseases on the Development of Africa . . . . . . 1173 A. Boutayeb
67
Measuring the Global Burden of Tuberculosis . . . . . . . . . . . . . . . . . . . . 1189 I. Onozaki . N. Ishikawa . D. A. Enarson
68
Burden of Tuberculosis: Serbian Perspectives . . . . . . . . . . . . . . . . . . . . . 1211 Z. Gledovic . H. Vlajinac . T. Pekmezovic . S. Grujicic-Sipetic . A. Grgurevic . D. Pesut
69
DALYs and Diarrhea . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1221 R. Oria . R. Pinkerton . AAM Lima . R. L. Guerrant
70
The Burden of Rotavirus Acute Gastroenteritis in Europe . . . . . . . . . . . . 1233 J. Bilcke . P. Van Damme . P. Beutels
Table of Contents
71
The Economic Burden of Rotavirus Diarrhea: Taiwan Perspectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1243 Kow-Tong Chen
72
Disease Burden of Dengue Fever and Dengue Hemorrhagic Fever . . . . . 1263 J. A. Suaya . D. S. Shepard . Mark E. Beatty . J. Farrar
73
Cost-Effectiveness of a Dengue Vaccine in Southeast Asia and Panama: Preliminary Estimates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1281 D. S. Shepard . J. A. Suaya
74
Burden of Sexually Transmitted Chlamydia trachomatis Infections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1297 L. M. Niccolai . D. Berube
75
Disease Burden from Group A Neisseria meningitidis Meningitis in Hyperendemic Countries of the African Meningitis Belt . . . . . . . . . . . . . 1313 C. Suraratdecha . C. Levin . F. M. LaForce
76
DALYs in Chronic Hepatitis C : A Paneuropean Perspective . . . . . . . . . . 1323 U. Siebert . A. Conrads-Frank . R. Schwarzer . B. Lettmeier . G. Sroczynski . S. Zeuzem . N. Mu¨hlberger
77
Economics and Vaccines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1335 J. Bos . M. Postma
78
Economic Costs and Disability-Adjusted Life Years in Polio Eradication: A Long-Run Global Perspective . . . . . . . . . . . . . . . . . . . . . . 1353 M. M. Khan
79
Financial Burdens and Disability-Adjusted Life Years in Echinococcosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1373 P. R. Torgerson
80
Measuring Japanese Encephalitis (JE) Disease Burden in Asia . . . . . . . . 1391 W. Liu . D. Ding . J. D. Clemens . N. T. Yen . V. Porpit . Z.‐Y. Xu
81
Pandemic Influenza: Potential Contribution to Disease Burden . . . . . . . 1401 M. Nun˜o
82
Prophylaxis of Healthcare Workers in an Influenza Pandemic . . . . . . . . 1419 S. M. Moghadas
83
The Burden of Human African Trypanosomiasis . . . . . . . . . . . . . . . . . . . 1433 A. Shaw . J. Robays . E. M. Fe`vre . P. Lutumba . M. Boelaert
xiii
xiv
Table of Contents
84
The Economic Burden of Malaria in Nigeria and Willingness to Pay . . . . 1443 A. Jimoh
85
Measurement of Adverse Health Burden Related to Sexual Behavior . . . 1459 S. H. Ebrahim . M. McKenna
Volume 3 Part 2.6 Psychosocial, Social, Behavioural, Psychiatric, Neurological and Addictions . . . . . . . . . . . . . . . . . . 1471 86
Global Burden of Mental Health . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1473 M. Kastrup
87
Disease Burden and Disability-Adjusted Life Years Due to Schizophrenia and Psychotic Disorders . . . . . . . . . . . . . . . . . . . . . . . . . 1493 A. Theodoridou . W. Ro¨ssler
88
The Societal Costs of Anxiety and Mood Disorders: An Epidemiological Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1509 R. C. Kessler . P. S. Wang . H.-U. Wittchen
89
Estimating the Disease Burden of Combat-Related Posttraumatic Stress Disorder in United States Veterans and Military Service Members . . . . . 1527 M. C. Freed . R. K. Goldberg . K. L. Gore . C. C. Engel
90
Estimating the Disease Burden of Seasonal Affective Disorder . . . . . . . . 1549 M. C. Freed . R. L. Osborn . K. J. Rohan
91
The Disease Burden of Suicide: Conceptual Challenges and Measurement Standards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1569 C. A. Claassen . R. M. Bossarte . S. M. Stewart . E. Guzman . P. S. F. Yip
92
The Disease Burden Due to Epilepsy in Rural China . . . . . . . . . . . . . . . . 1591 D. Ding . W. Z. Wang . Z. Hong
93
Alcohol Consumption and Burden of Disease: Germany and Switzerland . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1603 M. Roerecke . J. Rehm . J. Patra
94
Alcoholic Beverage Preference, Morbidity and Mortality . . . . . . . . . . . . 1619 T. E. Strandberg
95
Burden of Disease Due to Alcohol and Alcohol Related Research . . . . . . 1633 R. Rajendram . G. Lewison . V. R. Preedy
96
Years Life Lost Due to Smoking: A Korean Focus . . . . . . . . . . . . . . . . . . 1649 S. Yoon
Table of Contents
97
The Health and Economic Consequences of Smoking and Smoking Cessation Interventions: The Dutch Perspective . . . . . . . . . . . . . . . . . . . 1661 M. P. M. H. Rutten-van Mo¨lken . T. Feenstra
98
Life Years Saved, Quality-Adjusted Life Years Saved and Cost-Effectiveness of a School-Based Tobacco Prevention Program . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1681 L. Y. Wang . C. Linda . R. Lowry . G. Tao
Part 2.7 Sensory and Musculoskeletal . . . . . . . . . . . . . . . . . 1699 99
Burden of Disease and Years Lived with Disability Associated with Musculoskeletal Pain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1701 M. Suka . K. Yoshida
100
Financial Burdens and Disability-Adjusted Life Years in Loss of Vision Due to Trachoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1717 K. D. Frick
101
The Economic Burden of Rheumatoid Arthritis: Asia/Thailand Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1733 M. Osiri . A. Maetzel
Part 3 Quality of Life Measures and Indices . . . . . . . . . . . . 1751 Part 3.1 General Aspects, Pathologies and Metabolic Disorders . . . . . . . . . . . . . . . . . . . . . . . . . . 1751 102
Quality of Life-Related Concepts: Theoretical and Practical Issues . . . . . 1753 A. A. J. Wismeijer . A. J. J. M. Vingerhoets . J. De Vries
103
Alternative Therapies and Quality of Life . . . . . . . . . . . . . . . . . . . . . . . . 1767 J. X. Zhang
104
Leisure-Time Physical Activity and Quality of Life . . . . . . . . . . . . . . . . . 1781 A. Vuillemin
105
Spa Therapy and Quality of Life . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1799 G. Blasche
106
Health-Related Quality of Life and Prioritization Strategies in Waiting Lists: Spanish Aspects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1811 M. Nu´n˜ez . E. Nu´n˜ez . J. M. Segur
xv
xvi
Table of Contents
107
Hirsutism and Quality of Life . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1825 S. Davies
108
Oral Health-Related Quality of Life . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1839 U. Schu¨tte . M. Walter
109
Quality of Life Issues in Chronic Fatigue Syndrome . . . . . . . . . . . . . . . . 1855 P. G. McKay . C. R. Martin
110
Hemoglobin Fluctuations and Correlation with Quality of Life and Fatigue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1867 G. Caocci . R. Baccoli . G. La Nasa
111
Anemia and Quality of Life: Association with Diagnosis and Treatment of Anemias . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1881 D. R. Thomas
112
Quality of Life in Hemophilia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1895 S. V. Mackensen . A Gringeri
113
Quality of Life in Amenorrhea and Oligomenorrhea . . . . . . . . . . . . . . . . 1921 William W. K. To
114
Quality of Life Among Japanese Oral Contraceptive Users . . . . . . . . . . . 1937 Y. Matsumoto . S. Yamabe . K. Ideta
115
Premenstrual Syndrome and Premenstrual Dysphoric Disorder: Issues of Quality of Life, Stress and Exercise . . . . . . . . . . . . . . . . . . . . . 1951 M. Kathleen B. Lustyk . W. G. Gerrish
116
Quality of Life and Infertility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1977 A. Montazeri
117
Speech Determines Quality of Life Following Total Laryngectomy: The Emperors New Voice? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1989 P. Farrand . R. Endacott
118
Health-Related Quality of Life of Living Kidney Donors . . . . . . . . . . . . . 2003 Ja Hyeon Ku . Hyeon Hoe Kim
119
Quality of Life and Tryptophan Degradation . . . . . . . . . . . . . . . . . . . . . 2027 D. Fuchs . K. Schroecksnadel . G. Neurauter . R. Bellmann-Weiler . M. Ledochowski . G. Weiss
120
L-Carnitine Supplementation on Quality of Life and Other Health Measures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2047
G. Mantovani . A. Maccio` . C. Madeddu . G. Gramignano
Table of Contents
121
Quality of Life Among Diabetic Subjects: Indian Perspectives . . . . . . . . 2071 K. Vijayakumar . R. T. Varghese
122
Healthy Lifestyle Habits and Health-Related Quality of Life in Diabetes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2095 C. Li . E. S. Ford
123
Quality of Life in Patients with Diabetic Foot Ulcers . . . . . . . . . . . . . . . . 2115 L. Ribu
124
Obstructive Sleep Apnea Hypopnea Syndrome and Quality of Life . . . . 2135 M. Hirshkowitz . A. Sharafkhaneh . H. Sharafkhaneh
125
Quality of Life and Pruritus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2151 J. C. Szepietowski . A. Reich
126
Quality of Life and Costs in Atopic Dermatitis . . . . . . . . . . . . . . . . . . . . 2163 R. J. G. Arnold . R. K. Kuan
127
Quality of Life in Crohn’s Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2183 S. D. Wexner . J. C. Frattini
128
Impact of Self-Perceived Bothersomeness, Quality of Life and Overactive Bladder . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2195 Ja Hyeon Ku . Soo Woong Kim
129
Quality of Life in Men with Chronic Prostatitis/Chronic Pelvic Pain Syndrome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2211 D. A. Tripp . J. C. Nickel
130
Quality of Life in Kidney Transplantation . . . . . . . . . . . . . . . . . . . . . . . . 2227 M. Veroux . D. Corona . V. B. Patel . P. Veroux
131
Quality of Life in Liver Cirrhosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2239 E. Kalaitzakis
Volume 4 Part 3.2 Surgical . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2255 132
Quality of Life and Functional Outcome in Pediatric Patients Requiring Surgery: Italian Perspectives . . . . . . . . . . . . . . . . . . . . . . . . . 2257 M. Castagnetti
133
Breast Reduction Surgery and Quality of Life and Clinical Outcomes . . . 2271 A. Thoma . L. McKnight
xvii
xviii
Table of Contents
134
Quality of Life and Postoperative Anesthesia in Gastrointestinal Surgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2287 R. Kennelly . A. M. Hogan . J. F. Boylan . D. C. Winter
135
Health-Related Quality of Life After Surgery for Crohn’s Disease . . . . . . 2305 M. Scarpa . I. Angriman
136
Postoperative Quality of Life Assessment in the Over 80’s After Cardiac Surgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2319 Christoph H. Huber
137
Quality of Life and Financial Measures in Surgical and Non-Surgical Treatments in Emphysema . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2335 J. D. Miller . F. Altaf
138
Quality of Life After Revascularization and Major Amputation for Lower Extremity Arterial Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2353 M. Deneuville
139
Quality of Life After Laser Surgery for Eye Disorders . . . . . . . . . . . . . . . 2379 K. Pesudovs . D. B. Elliott
Part 3.3 Early Life Stages and Aging . . . . . . . . . . . . . . . . . . . 2395 140
Intrauterine Growth Restriction and Later Quality of Life . . . . . . . . . . . . 2397 D. Spence
141
Assessment of Quality of Life During Pregnancy and in the Postnatal Period . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2411 C. R. Martin . J. Jomeen
142
Generic Quality of Life Measures for Children and Adolescents . . . . . . . 2423 K. J. Zullig . M. R. Matthews . R. Gilman . R. F. Valois . E. S. Huebner
143
Quality of Life in Children with Cerebral Palsy . . . . . . . . . . . . . . . . . . . . 2453 A. Aran
144
Quality of Life Measures in Children with Cancer . . . . . . . . . . . . . . . . . . 2469 C. H. Yeh . Y.-P. Kung . Y.-C. Chiang
145
Quality of Life in Healthy and Chronically Ill Icelandic Children: Agreement Between Child’s Self-Report and Parents’ Proxy-Report . . . . 2483 E. K. Svavarsdottir
146
Health-Related Quality of Life in Obese Children and Adolescents . . . . . 2503 M. de Beer . R. J. B. J. Gemke
Table of Contents
147
Implementing Interventions to Enhance Quality of Life in Overweight Children and Adolescents . . . . . . . . . . . . . . . . . . . . . . . . . . 2517 J. Lamanna . N. Kelly . M. Stern . S. E. Mazzeo
148
Adolescent Quality of Life in Australia . . . . . . . . . . . . . . . . . . . . . . . . . . 2537 A. H. Lee . L. B. Meuleners . M. L. Fraser
149
Health-Related Quality of Life Among University Students . . . . . . . . . . . 2555 M. Vaez . M. Voss . L. Laflamme
150
The Quality of Life and the Impact of Interventions on the Health Outcomes of Looked After and Accommodated Young People . . . . . . . . 2579 D. Carroll . C. R. Martin
151
Quality of Life Measures During the Menopause . . . . . . . . . . . . . . . . . . 2593 G. D. Mishra . D. Kuh
152
Low Testosterone Level in Men and Quality of Life . . . . . . . . . . . . . . . . 2615 S. Horie
153
Measuring Quality of Life in Macular Degeneration . . . . . . . . . . . . . . . . 2633 J. Mitchell . C. Bradley
154
Quality of Life Measures in the Elderly and Later Life . . . . . . . . . . . . . . . 2649 S. Evans
155
Cochlear Implant Outcomes and Quality of Life in the Elderly . . . . . . . . 2667 S. R. Saeed . D. J. Mawman
156
Back Pain and Quality of Life in Elderly Women . . . . . . . . . . . . . . . . . . . 2675 K. Zhu . R. L. Prince
157
Measuring Quality of Life at the End of Life . . . . . . . . . . . . . . . . . . . . . . 2687 L. A. Roscoe . D. D. Schocken
158
Quality of Life Measures in the Elderly and the Role of Social Support in Elderly Chinese . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2705 L. Zhang . R. Hunter . C. Shao
159
Quality of Life in Elderly Dyspnea Patients . . . . . . . . . . . . . . . . . . . . . . . 2725 A. Hooshiaran . F. van der Horst . G. Wesseling . J. J. M. H. Strik . J. A. Knottnerus . A. Gorgels . A. Fastenau . M. van den Akker . J. W. M. Muris
160
Quality of Life and Age Urinary Incontinence Severity: Turkish Perspectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2745 T. M. Filiz . P. Topsever
xix
xx
Table of Contents
161
Quality of Life Measures in Elderly Patients with Chronic Obstructive Pulmonary Disease: Japanese Perspectives . . . . . . . . . . . . . . . . . . . . . . 2759 K. Kida . T. Motegi . T. Ishii . K. Yamada
Part 3.4 Cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2779 162
Chemotherapy for Brain Metastasis and Quality of Life . . . . . . . . . . . . . 2781 R. Addeo . G. Cimmino . S. D. Prete
163
Quality of Life Measures in Patients with Esophageal Cancer . . . . . . . . . 2795 R. Parameswaran . J. C. Clifton . J. M. Blazeby
164
Quality of Life Measures in Head and Neck Cancer . . . . . . . . . . . . . . . . . 2809 C. D. Llewellyn
165
Quality of Life in Breast Cancer Patients: An Overview of the Literature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2829 A. Montazeri
166
Quality of Life with Localized Prostate Cancer: Japanese Perspectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2857 S. Namiki . L. Kwan . Y. Arai
167
Quality of Life in Men Undergoing Radical Prostatectomy for Prostate Cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2875 M. Pearson . E. M. Wallen . R. S. Pruthi
168
Myeloproliferative Disorders and the Chronic Leukemias: Symptom Burden and Impact in Quality of Life . . . . . . . . . . . . . . . . . . . . . . . . . . . 2887 R. A. Mesa . D. P. Steensma . T. Shanafelt
169
Quality of Life in Advanced Renal Cell Carcinoma: Effect of Treatment with Cytokine Therapy and Targeted Agents . . . . . . . . . . . . 2905 S. Shah . K. Gondek
170
Quality of Life for Patients Receiving Cancer Chemotherapy: The Japanese Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2923 H. Uramoto . J. Tsukada
171
Quality of Life Measures in Caregivers of Patients with Cancer . . . . . . . 2935 E. K. Grov . A. A. Dahl
172
Cancer: Influence of Nutrition on Quality of Life . . . . . . . . . . . . . . . . . . . 2947 M. M. Marı´n Caro . C. Pichard
Table of Contents
Volume 5 Part 3.5 Cardiovascular and Pulmonary . . . . . . . . . . . . . . . . 2965 173
Quality of Life, Drugs and Diet in Hypertensive Patients . . . . . . . . . . . . 2967 H. G. Kirpizidis
174
Measurement Issues in the Assessment of Quality of Life in Patients with Coronary Heart Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2987 D. R. Thompson . C. R. Martin
175
Influence of Age, Sex and Episode Recurrence on Quality of Life in Atrial Fibrillation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2999 M. R. Reynolds . A. S. Fein
176
Quality of Life Measures in Acute Coronary Syndromes: The Evaluation of Predictors in this Field of Research . . . . . . . . . . . . . . 3015 R. Coelho . J. Prata
177
Home Mechanical Ventilation and Quality of Life Measures . . . . . . . . . . 3035 J. L. Lo´pez-Campos . W. Windisch . I. Failde
178
Quality of Life in Children with Asthma . . . . . . . . . . . . . . . . . . . . . . . . . 3055 M. L. Marsac
179
Efficacy of Environmental Interventions on Asthma-Related Quality of Life . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3073 J. E. Clougherty
180
Chronic Obstructive Pulmonary Disease, Lung Function and Quality of Life in Adult Chinese . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3085 G. Xie . A. Paice . Y. Wu
Part 3.6 Dietary and Nutritional . . . . . . . . . . . . . . . . . . . . . . . 3097 181
Health-Related Quality of Life in Eating Disorders . . . . . . . . . . . . . . . . . 3099 ´ . Padierna . P. Mun˜oz C. Las Hayas . J. A
182
Weight-Related Stigmatization: Effects on the Quality of Life of Obese Adolescents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3137 N. R. Kelly . R. W. Gow . M. Stern . S. E. Mazzeo
183
Malnutrition in Chronic Kidney Disease and Relationship to Quality of Life . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3159 B. Kalender
xxi
xxii
Table of Contents
184
Nutrition and Quality of Life in Hemodialysis – the Impact of Nutritional Status and Quality of Life on Morbidity and Mortality in Hemodialysis Patients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3171 B. Feldt-Rasmussen . T. A. Ikizler . K. Kalantar-Zadeh . J. D. Kopple
185
Nutritional Wasting in Cancer and Quality of Life: The Value of Early Individualized Nutritional Counseling . . . . . . . . . . . . . . . . . . . . . . . . . . . 3189 P. Ravasco . I. M. Grillo . M. Camilo
Part 3.7 Immune Disorders, Viral, Bacterial, Microbiological, Infectious and Parasitic Diseases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3205 186
Quality of Life Measures in HIV Low Income Women: How to Use Quality of Life to Design, Implement and Evaluate Programs . . . . . . . . . 3207 K. A. McDonnell
187
Quality of Life and Financial Measures in HIV/AIDS in Southern Africa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3223 M. O. Bachmann . G. Louwagie . L. R. Fairall
188
Quality of Life in Immune Thrombocytopenic Purpura: China Perspectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3245 R. Yang . Z. Zhou
189
Health-Related Quality of Life in Adults with Systemic Lupus Erythematosus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3261 L-S. Teh . K. McElhone . J. Abbott
190
Translation, Cultural Adaptation and Validation of Health-Related Quality of Life Assessment Tools: A Brazilian Perspective on Patients with Systemic Lupus Erythematosus . . . . . . . . . . . . . . . . . . . . . 3281 E. A. M. Freire . R. M. Ciconelli
191
Cognitive Function, Mood and Health-Related Quality of Life in Hepatitis C Virus-Infected Individuals . . . . . . . . . . . . . . . . . . . . . . . . . . . 3299 Hla-Hla Thein . G. J. Dore
192
Quality of Life in Urticaria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3327 ¨ zkan . S. O ¨ zkan . N. Kocaman Yildirim M. O
193
Quality of Life in Group-Based Intervention Program in Inflammatory Bowel Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3339 L. Oxelmark
Table of Contents
Part 3.8 Psychosocial, Social, Behavioural, Psychiatric, Neurological and Addictions . . . . . . . . . . . . . . . . . . 3361 194
Housing and Quality of Life: An Ecological Perspective . . . . . . . . . . . . . 3363 G. Nelson . S. Saegert
195
Quality of Life of Urban Slum Residents . . . . . . . . . . . . . . . . . . . . . . . . . 3383 T. Izutsu . A. Tsutsumi
196
Quality of Life and Chronic Illness among Refugee Populations . . . . . . . 3397 R. T. Mikolajczyk . A. E. Maxwell . A. Eljedi
197
Health-Related Quality of Life in Prisoners . . . . . . . . . . . . . . . . . . . . . . . 3413 G. J. Dore
198
Quality of Life in War Veterans . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3425 N. Shamspour . S. Assari
199
Post-Traumatic Stress Disorder and Quality of Life in Women . . . . . . . . 3439 C. S. Rodgers . C. B. Allard . P. Wansley
200
Relation Between Sexuality and Health-Related Quality of Life . . . . . . . 3457 N. Shamspour . S. Assari . M. Moghana Lankarani
201
The Correlations Between the Presence of Comorbidities, Psychological Distress and Health-Related Quality of Life . . . . . . . . . . . 3475 M. Ekici . A. Ekici
202
Quality of Life and Stigma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3489 A. Tsutsumi . T. Izutsu
203
Depressive Symptoms and Health-Related Quality of Life . . . . . . . . . . . 3501 A. A. Dan . Z. M. Younossi
204
Quality of Life and Depression in Patient-Giver Scenarios: Reference to Amyotrophic Lateral Sclerosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3511 A. Chio`
205
Quality of Life and Depression in Police Officers: Perspectives from Chinese in Taiwan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3541 F. H.-C Chou . M.-H. Kuo . K.-Y. Tsai
206
Health-Related Quality of Life in Obsessive-Compulsive Disorder Subjects and their Relatives [1]: Overview . . . . . . . . . . . . . . . . . . . . . . . 3557 U. Albert . G. Maina . F. Bogetto
xxiii
xxiv
Table of Contents
207
Health-Related Quality of Life in Obsessive-Compulsive Disorder Subjects and Their Relatives [2]: A Systematic Review and Original Data of Assessment of Quality of Life Measured by WHOQOL-BREF . . . . 3579 K. Stengler
208
Quality of Life in Bipolar Disorder . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3591 B. M. Cardoso . V. V. Dias . B. N. Frey . F. K. Gazalle . F. Kapczinski . M. Kauer-Sant’Anna . A. R. Rosa . J. C. Walz
209
Issues in Quality of Life Assessment in Schizophrenia . . . . . . . . . . . . . . 3607 C. R. Martin . M. Fleming
210
Health-Related Quality of Life in Parents of Children with Asperger Syndrome and High-Functioning Autism . . . . . . . . . . . . . . . . . . . . . . . . 3625 H. Allik . J.-O. Larsson . H. Smedje
211
Quality of Life and Neuropsychological Symptoms in Primary Hyperparathyroidism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3643 T. Weber . M. Keller
212
Children with Cerebral Palsy, Psychometric Analysis and Quality of Life . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3657 E. Davis . E. Waters
213
Quality of Life in Dementia Patients and Their Proxies: A Narrative Review of the Concept and Measurement Scales . . . . . . . . . . . . . . . . . . 3671 C. J. M. Scho¨lzel-Dorenbos . P. F. M. Krabbe . M. G. M. Olde Rikkert
Volume 6 214
Quality of Life and Drug Abuse . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3691 S. Assari . M. Jafari
215
Quality of Life in HIV Positive Injecting Drug Users . . . . . . . . . . . . . . . . 3705 M. Pre´au . A. D. Bouhnik . M. P. Carrieri . F. M. B. Spire
216
Quality of Life Measurement and Alcoholism: A Nursing Perspective . . . 3727 J. H. Foster
217
Quality of Life and Psychiatric Symptomatology in Alcohol Detoxification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3747 M. Ginieri-Coccossis . I. A. Liappas
218
Quality of Life in Patients Affected by Multiple Sclerosis: A Systematic Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3769 F. Patti . A. Pappalardo
Table of Contents
219
Quality of Life in People with Lower-Limb Amputation . . . . . . . . . . . . . 3785 D. Desmond . P. Gallagher
220
Quality of Life in Spinal Cord Injury . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3797 B. Celik
221
Quality of Life in Sporadic Adult-Onset Ataxia . . . . . . . . . . . . . . . . . . . . 3809 M. Abele
222
Quality of Life in Systemic Sclerosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3823 L. Mouthon . F. Rannou . A. Berezne´ . S. Poiraudeau
223
Quality of Life in Toenail Onychomycosis . . . . . . . . . . . . . . . . . . . . . . . . 3837 A. Reich . J. C. Szepietowski
Part 3.9 Sensory, Musculoskeletal and Exercise . . . . . . . . . 3851 224
Quality of Life Measures in the Deaf . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3853 J. Fellinger . D. Holzinger . J. Gerich . D. Goldberg
225
Hearing Aids and Quality of Life . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3871 C. E. Johnson . J. L. Danhauer
226
Cochlear Implantation and Quality of Life in Deafness . . . . . . . . . . . . . . 3887 G. W. J. A. Damen . E. A. M. Mylanus . A. F. M. Snik
227
Quality of Life Measures for the Visually Impaired: Sub-Sahara Africa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3905 A. Leplege . J.-F. Schemann
228
Health-Related Quality of Life in Pain Medicine: A Review of Theory and Practice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3917 T. R. Vetter
229
Relationship Between Pain and Quality of Life . . . . . . . . . . . . . . . . . . . . 3933 M. Azizabadi Farahani . S. Assari
230
Quality of Life in Ankylosing Spondylitis and Undifferentiated Spondyloarthropathy: Chinese Perspectives . . . . . . . . . . . . . . . . . . . . . . 3955 J. R. Gu . Z. T. Liao . Z. M. Lin . R. Srirajaskanthan
231
Quality of Life Measures in Fibromyalgia . . . . . . . . . . . . . . . . . . . . . . . . 3965 N. Gusi . Pedro R. Olivares . J. Carmelo Adsuar . A. Paice . P. Tomas-Carus
232
Quality of Life and Low Back Pain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3979 A. Montazeri . S. J. Mousavi
xxv
xxvi
Table of Contents
233
Disease Burden, Quality of Life and Other Measures in Polymyalgia Rheumatica . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3995 B. Dasgupta . H. M.-Kremers . E. L. Mattesson
234
Health-Related Quality of Life in Movement Disorders . . . . . . . . . . . . . . 4013 R. Dodel . A. Schrag
235
Quality of Life Measures in Lower Limb Ischaemia . . . . . . . . . . . . . . . . . 4035 F. A. K. Mazari . T. A. Mehta . I. C. Chetter
236
Quality of Life among Primary Caregivers of Rheumatic Patients – a South American Experience . . . . . . . . . . . . . . . . . . . . . . . . . 4053 F. Jennings . A. Jones . J. Natour
237
Quality of Life in Conservatively Treated Lumbar Disc Disease . . . . . . . 4071 C. Schneider . M. Hefti . H. Landolt
238
Quality of Life and Stress in Wheelchair-Users . . . . . . . . . . . . . . . . . . . . 4087 M. Furlong . J. Connor
239
Exercise and Quality of Life in Menopause . . . . . . . . . . . . . . . . . . . . . . . 4103 A. J. Daley . H. Stokes-Lampard . C. MacArthur
240
Exercise and Quality of Life in COPD . . . . . . . . . . . . . . . . . . . . . . . . . . . 4119 J. A. Alison . Z. J. McKeough
Glossary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4133 Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4357
Editors-in-Chief Professor Victor R. Preedy Dept Nutrition and Dietetics Nutritional Sciences Research Division School of Biomedical & Health Sciences King’s College London Franklin-Wilkins Building 150 Stamford Street London SE1 9NH UK
Professor Ronald R. Watson Mel and Enid Zuckerman College of Public Health University of Arizona Health Science Center 1295 N. Martin P.O. Box 245155 Tucson, AZ 85724–5155
Editorial Advisors and Board Dr. Ross Hunter Dept Nutrition and Dietetics Nutritional Sciences Research Division School of Biomedical & Health Sciences King’s College London Franklin-Wilkins Building 150 Stamford Street London SE1 9NH UK Professor Colin R Martin School of Health, Nursing and Midwifery University of the West of Scotland Ayr Campus Beech Grove, AYR Ayrshire KA8 0SR Scotland UK Dr. Alistair Paice Department of Nutrition and Dietetics Nutritional Sciences Division School of Biomedical and Health Sciences King’s College of London, London UK Dr. Vinood Patel Senior Lecturer Department of Biomedical Sciences
School of Life Sciences University of Westminster 115 New Cavendish Street London W1W 6UW UK
Dr. Rajkumar Rajendram Specialist Registrar Departments of General Medicine and Intensive Care John Radcliffe Hospital Oxford OX3 9DU Dept Nutrition and Dietetics Nutritional Sciences Research Division School of Biomedical & Health Sciences King’s College London Franklin-Wilkins Building 150 Stamford Street London SE1 9NH UK
Dr. Rajaventhan Srirajaskanthan Department of Gastroenterology St. Thomas’ Hospital London SE1 9RT UK
Contributors Janice Abbott University of Central Lancashire, Preston UK Michael Abele Department of Neurology University of Bonn, Bonn Germany Raffaele Addeo Oncology Unit, “S. Giovanni di Dio” HospitaL, Frattaminore Naples Italy J. Carmelo Adsuar University of Extremadura, Caceres Spain Marjan van den Akker Department of General Practice, School for Public Health and Primary Care: Caphri Maastricht University, Maastricht The Netherlands Umberto Albert Mood and Anxiety Disorders Unit Department of Neurosciences University of Turin, Torino Italy Jennifer A. Alison Discipline of Physiotherapy, Faculty of Health Sciences, The University of Sydney NSW Australia Carolyn B. Allard VA San Diego Healthcare System, Department of Psychiatry, University of California, San Diego, CA USA Hiie Allik Department of Woman and Child Health Karolinska Institute, Child and Adolescent #
Springer Science+Business Media LLC 2010 (USA)
Psychiatric Unit, Stockholm Sweden David B. Allison Department of Biostatistics, University of Alabama at Birmingham, UAB Station USA Fawaz Altaf McMaster University, Ancaster, ON Canada Imerio Angriman Department of Surgical and Gastroenterological Science University of Padova Italy Yoichi Arai David Geffen School of Medicine (SN, YA) Jonsson Comprehensive Cancer Center University of California Los Angeles (LK) USA Adi Aran Shaare Zedek Medical Center Neuropediatric Unit, Jerusalem Israel Zeyneo Armay Department of Consultation Liaison Psychiatry, University of Istanbul, Istanbul Turkey Rene´e J. G. Arnold Department of Community and Preventive Medicine, Mount Sinai School of Medicine, Arnold Consultancy & Technology LLC, New York, NY USA
xxx
Contributors
Shervin Assari Psychology and Psychiatry Research Department, Medicine and Health Promotion Institute, Baqiyatallah University of Medical Sciences, Tehran Iran Roberto Baccoli Engineering and Territory Development Physical Technical Institute University of Cagliari Cagliari Italy Max O. Bachmann School of Medicine, Health Policy and Practice, University of East Anglia Norwich UK Xavier Badia Health Economics and Outcomes Research IMS Health Spain Enzo Ballatori Biostatistician Freelance, AP Italy Limaris Barrios Beth Israel Deaconess Medical Center Boston, MA USA
Rosa Bellmann-Weiler Department of Internal Medicine Innsbruck Medical University, Innsbruck Austria Alice Berezne´ Department of Internal Medicine, Paris Descartes University, Reference Center for Necrotizing Vasculitides and Systemic Sclerosis, Cochin Hospital, Assistance Publique-Hoˆpitaux de Paris (AP-HP), Paris France Philippe Beutels Center for Health Economics Research and Modeling of Infectious Diseases (CHERMID) Center for the Evaluation of Vaccination (CEV), (WHO Collaborating Center) Vaccine & Infectious Disease Institute (VAXINFECTIO) Antwerp University, Wilrijk (Antwerp) Belgium Madhumita Bhattacharjee Lilavati Hospital, L. K. M. M. Trust Research Centre, Bandra Reclamation, Mumbai India
Stefano Di Bartolomeo Cattedra di Igiene ed Epidemiologia – Universita´ degli Studi di Udine Udine Italy
Joke Bilcke Center for Health Economics Research and Modeling of Infectious Diseases (CHERMID) Center for the Evaluation of Vaccination (CEV), Vaccine & Infectious Disease Institute (VAXINFECTIO) Antwerp University, Wilrijk (Antwerp) Belgium
Mark E. Beatty Pediatric Dengue Vaccine Initiative International Vaccine Institute, Seoul Korea
John Billimek Center for Health Policy Research University of California, Irvine, CA USA
Marieke de Beer Department of Pediatrics, VU University Medical Center, Amsterdam The Netherlands
Vesna Bjegovic´ Institute of Social Medicine, School of Medicine, University of Belgrade, Belgrade Serbia
Contributors
Gerhard Blasche Department of Physiology, Centre of Physiology and Pathophysiology Medical University of Vienna, Vienna Austria
John F. Boylan Institute for Clinical Outcomes Research and Education (iCORE), St. Vincent’s University Hospital, Dublin Ireland
Jane M. Blazeby Department of Social Medicine, University of Bristol, Bristol UK
Clare Bradley Department of Psychology, Royal Holloway, University of London Surrey UK
Marleen Boelaert Epidemiology and Disease Control Unit Institute of Tropical Medicine Antwerp Belgium Filippo Bogetto Mood and Anxiety Disorders Unit Department of Neurosciences, University of Turin Italy Jasper Bos Associate Director, Merch Serono Ventures Geneva Switzerland Robert M. Bossarte Department of Psychiatry, University Rochester, Rochester, NY USA Anne-De´borah Bouhnik Southeastern Health Regional Observatory (ORS-PACA), Research Unit UMR912, Economic & Social Sciences Health Systems & Societies, INSERM Marseille France and IRD, Aix Marseille Universite´, Marseille France Sami Boussetta Departement Pharmaco-Economique Pierre Fabre SA, Boulogne-sur-Seine France Abdesslam Boutayeb University Mohamed Ier Oujda Morocco
Freddie I. Bray Department of Clinical- and Registry-Based Research, Cancer Registry of Norway Montebello, Oslo Norway Alex D. Brown Centre for Indigenous Vascular and Diabetes Research, Baker IDI Heart & Diabetes Institute, NT Australia Anna A. Bruno Dip. di Riabilitazione, Ospedale ‘S Salvatore’ L’Aquila Italy Ljiljana B. Bumbasirevic Institute of Neurology, Clinical Centre of Serbia, Belgrade Serbia Neil Burnet Department of Oncology, University of Cambridge, Cambridge UK Dominique A. Cadilhac National Stroke Research Institute Heidelberg Heights, Vic Australia Laura Camfield Department of International Development Young Lives Research Group Oxford UK
xxxi
xxxii
Contributors
Maria Camilo Unidade de Nutric¸a˜o e Metabolismo Instituto de Medicina Molecular Faculdade de Medicina da Universidade de Lisboa Portugal Giovanni Caocci Bone Marrow Transplant Centre R. Binaghi Hospital University of Cagliari Cagliari Italy Betina M. Cardoso Alameda Victor Adalberto Kessler Porto Alegre-RS Brasil J. Jaime Caro Health Care Analytics United BioSource Corporation Lexington, MA USA Marie P. Carrieri Research Unit ‘‘Economic & Social Sciences Health Systems & Societies’’, Marseille France and IRD, Aix Marseille Universite´ Marseille France Denise Carroll Kibble Care and Education Centre Paisley Scotland Marco Castagnetti Section of Paediatric Urology, Urology Unit Department of Oncological and Surgical Sciences, University Hospital of Padova Padua Italy Berna Celik Department of Physical Medicine and Rehabilitation, Istanbul Physical Medicine Rehabilitation Teaching and Research Hospital, Bahcelievler – Istanbul Turkey
Kow-Tong Chen Department of Public Health, College of Medicine, National Cheng Kung University Tainan Taiwan Ian C. Chetter Academic Vascular Surgical Unit, Hull UK Yi-Chien Chiang Department of Nursing, Chang Gung Institute of Technology, Tao Yuen Taiwan Adriano Chio Department of Neuroscience, ALS Center University of Torino, Torino Italy Frank Chou Department of Community Psychiatry Kai-Suan Psychiatric Hospital, Kaohsiung Taiwan Rozana M. Ciconelli Department of Medicine, Universidade Federal de Sa˜o Paulo Sa˜o Paulo Brazil Alarcos A. Cieza ICF Research Branch of WHO FIC CC (DIMDI), IHRS, Ludwig Maximilian University, Munich, Mu¨nchen Germany Gaetano Cimmino Surgery Unit, ‘‘S. Giovanni di Dio’’ Hospital Frattaminore, Naples Italy Cynthia A. Claassen Department of Psychiatry, The University of Texas Southwestern Medical Center, Dallas TX USA John D. Clemens International Vaccine Institute, Kwanak-ku Seoul Republic of Korea
Contributors
Joanne C. Clifton Department of Surgery, Division of Thoracic Surgery, University of British Columbia, British Columbia Canada Jane E. Clougherty Department of Environmental Health Harvard School of Public Health, Boston, MA USA Rui Coelho Department of Psychiatry, Hospital de Sa˜o Joa˜o, Porto Portugal Joshua A. Cohn University of Michigan Health System Ann Arbor, MI USA Marisol Concha-Barrientos Health Department, Asociacio´n Chilena de Seguridad, Santiago Chile Jason Connor Discipline of Psychiatry, The University of Queensland, Mental Health Centre, Royal Brisbane and Women’s Hospital, Herston QLD Australia Annette Conrads-Frank Department of Public Health, Information Systems and Health Technology Assessment, UMIT - University for Health Sciences, Medical Informatics and Technology and ONCOTYROL Center for Personalized Cancer Medicine Hall i.T./Innsbruck Austria Daniela Corona Organ Transplant Unit, Department of Surgery, Transplantation and Advanced Technologies, University Hospital of Catania Italy
Linda Crossett Division of Adolescent and School Health National Center for Chronic Disease Prevention and Health Promotion Centers for Disease Control and Prevention Atlanta, GA USA Alv A. Dahl The Norwegian Radium Hospital Oslo University Hospital Rikshospitalet, Oslo Norway Praful M. Dalal Lilavati Hospital, L. K. M. M. Trust Research Centre, Bandra Reclamation, Mumbai India Amanda J. Daley Primary Care Clinical Sciences, School of Health and Population Sciences University of Birmingham UK Godelieve W. J. A. Damen Department of Otorhinolaryngology University Medical Centre Sint Radboud Nijmegen The Netherlands Pierre Van Damme Center for the Evaluation of Vaccination (CEV), (WHO Collaborating Center) Vaccine & Infectious Disease Institute (VAXINFECTIO) Antwerp University, Wilrijk (Antwerp) Belgium Amy A. Dan Michigan State University, Environmental Science and Policy Program East Lansing, MI USA Jeffrey L. Danhauer Department of Speech and Hearing Sciences, University of California Santa Barbara, Santa Barbara, CA USA
xxxiii
xxxiv
Contributors
Bhaskar Dasgupta Department of Rheumatology, Southend University Hospital, Essex UK
Bipolar Disorders Research Program, Hospital Santa Maria, Faculty of Medicine, University of Lisbon, (FMUL) Portugal
Shan Davies Institute of Health Research, School of Health Science, Swansea University Singleton Park, Swansea, Wales UK
Ding Ding Department of Biostatistics and Epidemiology, Institute of Neurology Fudan University, WHO Collaborating Center for Research and Training in Neurosciences Shanghai China
Elise Davis McCaughey Centre, VicHealth Centre for the Promotion of Mental Health and Community Wellbeing, School of Population Health, University of Melbourne, VIC Australia Michel Deneuville Service de Chirurgie Vasculaire et Thoracique, University Hospital (CHU de Pointe-a`-Pitre-Abymes), Guadeloupe French West Indies Anne V. Denison Texas Tech University Health Sciences Center, Amarillo, TX USA Deirdre Desmond Department of Psychology, National University of Ireland Maynooth, Maynooth Ireland Misha Dewan Drexel University College of Medicine, Philadelphia, PA USA Vasco V. Dias Bipolar Disorders Program and Molecular Psychiatry Unit, Hospital de Clı´nicas Federal University, UFRGS, Porto Alegre Brazil and
Richard Dodel Department of Neurology Philipps-University Marburg, Marburg Germany Gregory J. Dore National Centre in HIV Epidemiology and Clinical Research, The University of New South Wales, Darlinghurst NSW Australia Shahul H. Ebrahim Centers for Disease Control and Prevention, US Department of Health and Human Services, Atlanta, GA USA Aydanur Ekici Department of Chest Diseases, Kirikkale University, Kirikkale Turkey Mehmet Ekici Department of Chest Diseases, Kirikkale University, Kirikkale Turkey Ashraf Eljedi School of Nursing, The Islamic University of GazaGaza, Gaza Strip Palestinian Territories
Contributors
David B. Elliott Bradford School of Optometry, University of Bradford, Bradford, West Yorkshire UK Donald A. Enarson International Union against Tuberculosis and Lung Disease, Paris France Ruth Endacott School of Nursing and Community Studies University of Plymouth, Drake Circus Plymouth UK Charles C. Engel Department of Psychiatry Chief Deployment Health Clinical Center, Walter Reed Army Medical Center, Uniformed Services University of the Health Sciences Bethesda, MD USA Michael J. Englesbe University of Michigan Health System, Ann Arbor, MI USA Sherrill Evans Centre for Social Work and Social Care Research, School of Human Sciences Swansea University, Singleton Park Swansea UK Immaculade Failde Area de Medicina Preventiva y Salud Pu´blica. E.U. Ciencias de la Salud Universidad de Cadiz, Cadiz Spain Lara R. Fairall Knowledge Translation Unit, University of Cape Town Lung Institute and Department of Medicine, University of Cape Town Cape Town South Africa
Mahdi A. Farahani Clinical Research Department, Medicine and Health Promotion Institute Baqiyatallah University of Medical Sciences Tehran Iran Paul Farrand School of Psychology University of Exeter, Exeter UK Jeremy Farrar Hospital for Tropical Diseases, Oxford University’s Clinical Research Unit, Ho Chi Minh City Vietnam Annemieke Fastenau Department of General Practice, School for Public Health and Primary Care: Caphri Maastricht University, Maastricht The Netherlands Talitha Feenstra Netherlands Institute for Public Health and the Environment, Bilthoven The Netherlands Adam S. Fein Beth Israel Deaconees Medical Center Boston, MA USA Bo Feldt-Rasmussen Division of Nephrology, Rigshospitalet University of Copenhagen, Copenhagen Denmark Johannes Fellinger Health Centre for the Deaf Hospital St. John of God, Linz Austria Jacques Ferlay Descriptive Epidemiology, Data Analysis and Interpretation Group, International Agency for Research on Cancer, Lyon France
xxxv
xxxvi
Contributors
Eric M. Fe`vre Ashworth Laboratories, Centre for Infectious, Diseases and Centre for Infection, Immunity and Evolution University of Edinburgh, Edinburgh UK Jonathan E. Fielding Los Angeles County Department of Public Health, Los Angeles, CA USA Tuncay M. Filiz Department of Family Medicine, Kocaeli University, Kocaeli Turkey Mick Fleming Department of Health Sciences, University of York, York UK Kevin R. Fontaine Johns Hopkins University, Baltimore, MD USA
Michael C. Freed Deployment Health Clinical Center Walter Reed Army Medical Center Department of Psychiatry Uniformed Services University of the Health Sciences Capital Behavioral Health & Wellness, LLC Washington, DC USA Eutilia A. M. Freire Department of Internal Medicine Universidade Federal da Paraı´ba, Paraı´ba Brazil Benicio N. Frey Department of Psychiatry and Behavioural Neurosciences McMaster University, Hamilton, ON Canada Kevin D. Frick Johns Hopkins Bloomberg School of Public Health, Baltimore, MD USA
Earl S. Ford Behavioral Surveillance Branch, Division of Adult and Community Health National Center for Chronic Disease Prevention and Health Promotion Centers for Disease Control and Prevention Atlanta, GA USA
Dietmas Fuchs Division of Biological Chemistry, Biocenter Innsbruck Medical University, Innsbruck Austria
John H. Foster Department of Health and Social Sciences Archway Campus, Middlesex University London UK
Seana L. Gall Menzies Research Institute, Hobart TAS Australia
Michelle L. Fraser School of Public Health, Curtin University of Technology, Perth, WA Australia Jared C. Frattini Department of Colorectal Surgery Cleveland Clinic Florida Weston, FL USA
Michele Furlong Diabetes and Endocrinology Unit, Princess Alexandra Hospital, QLD Australia
Pamela Gallagher School of Nursing, Dublin City University Dublin Ireland Fernando K. Gazalle Bipolar Disorders Program and Laboratory of Molecular Psychiatry Hospital de Clı´nicas de Porto Alegre Brazil and INCT Translational Medicine
Contributors
Reinoud J. B. J. Gemke Department of Pediatrics, VU University Medical Center, Amsterdam The Netherlands Joachim Gerich Department of Sociology, unit for empirical social research, Johannes Kepler University Linz Austria Winslow G. Gerrish Department of Clinical Psychology School of Psychology, Family and Community, Seattle Pacific University Seattle, WA USA Montserrat Gilabert Novartis Oncology Spain Rich Gilman Psychology and Special Education Programs, Division of Developmental and Behavioral Pediatrics, Cincinnati Children’s Hospital Medical Center Department of General Pediatrics University of Cincinnati Medical School Ohio USA
Marthe Gold Department of Community Health and Social Medicine, Sophie Davis School of Biomedical Education/CUNY Medical School, New York, NY USA David Goldberg Institute of Psychiatry, King’s College London UK Robert K. Goldberg Deployment Health Clinical Center Walter Reed Army Medical Center Washington, DC USA Binh Y. Goldstein Sexually Transmitted Disease Program Los Angeles County Department of Health, Los Angeles, CA USA Kathleen Gondek Global Health Economics, Outcomes and Reimbursement, Bayer HealthCare Pharmaceuticals, Montville, NJ USA
Maria Ginieri-Coccossis Eginition Hospital, Department of Psychiatry, Medical School, University of Athens Greece
Kristie L. Gore Deployment Health Clinical Center, Department of Psychiatry, USUHS Walter Reed Army Medical Center NW Washington, DC USA
Ashok K. Giri Molecular and Human Genetics Division Indian Institute of Chemical Biology, West Bengal India
Anton Gorgels Department of Cardiology, Maastricht University Medical Centre: MUMC Maastricht University, Maastricht The Netherlands
Zorana Gledovic Institute of Epidemiology, School of Medicine, University of Belgrade, Belgrade Serbia
Hebe Gouda Department of Public Health and Primary Care, Institute of Public Health, Robinson Way, Cambridge UK
xxxvii
xxxviii
Contributors
Rachel W. Gow Department of Psychology, Virginia Commonwealth University Richmond, VA USA
Richard L. Guerrant Centers for Global Health, University of Virginia, Charlottesville, Virginia and Federal University of Ceara´, Fortaleza Brazil
Giulia Gramignano Department of Medical Oncology University of Cagliari, Cagliari Italy
Jean-Jerome Guex Vascular Medicine & Phlebology, Nice France
James A. Greenberg Department of Health and Nutrition Sciences, Brooklyn College of the City University of New York, Brooklyn, NY USA Sheldon Greenfield Center for Health Policy Research University of California, Irvine, CA USA Anita Grgurevic Institute of Epidemiology, School of Medicine, University of Belgrade, Belgrade Serbia Isabel M. Grillo Servic¸o de Radioterapia Hospital Universita´rio de Santa Maria Centro Hospitalar de Lisboa Norte Lisboa Portugal Alessandro Gringeri Angelo Bianchi Bonomi Haemophilia and Thrombosis Centre, IRCCS Maggiore Hospital and University of Milan, Milan Italy
Narcis Gusi University of Extremadura, Ca´ceres Spain Ellen Guzman Psychiatry Resident, Thomas Jefferson University (Jefferson Medical College) Philadelphia, PA USA Carlota Las Hayas CIBER in Epidemiology and Public Health Galdakao – Usansolo Hospital, Galdakao Vizcaya Spain Martin Hefti Department of Neurosurgery Kantonsspital Aarau, Aarau Switzerland Yusuf Hemed MEASURE Evaluation, Dar Es Salaam Tanzania Max Hirshkowitz VAMC Sleep Center (111-i), Houston, TX USA
Ellen K. Grov The Norwegian Radium Hospital Oslo University Hospital Rikshospitalet, Oslo Norway
Alex Y. Ho Office of Health Assessment & Epidemiology Los Angeles County Department of Public Health, Los Angeles, CA USA
Jieruo R. Gu Division of Rheumatology, The Third Affiliated Hospital of Sun Yat-sen University, Guangzhou People’s Republic of China
Aisling M. Hogan Institute for Clinical Outcomes Research and Education (iCORE), St. Vincent’s University Hospital, Dublin Ireland
Contributors
Daniel Holzinger Health Centre for the Deaf Hospital St. John of God, Bischofstrasse Linz Austria Zhen Hong Department of Biostatistics and Epidemiology, Institute of Neurology Fudan University, WHO Collaborating Center for Research and Training in Neurosciences, Shanghai China
T. Alp Ikizler Vanderbilt University Medical Center Vanderbilt University School of Medicine Nashville, TN USA Takeo Ishii Department of Pulmonary Medicine Infection, and Oncology, Respiratory Care Clinic, Nippon Medical School, Tokyo Japan
Afshin Hooshiaran Institute for Education, Medical Programme Maastricht University, Maastricht The Netherlands
Nobukatsu Ishikawa The Research Institute of Tuberculosis Japan Ant-Tuberculosis Association Matsuyama, Kiyose, Tokyo Japan
Shigeo Horie Department of Urology, Teikyo University School of Medicine, Tokyo Japan
Takashi Izutsu Department of Mental Health, Graduate School of Medicine, The University of Tokyo, Tokyo Japan
Frans van der Horst Department of General Practice, School for Public Health and Primary Care: Caphri Maastricht University, Maastricht The Netherlands Christoph H. Huber Cardiovascular Surgery Division, Centre Hospitalier Universitaire Vaudois (CHUV) Lausanne Switzerland E. Scott Huebner Department of Psychology, University of South Carolina, Columbia, SC USA Jing-Shiang Hwang Institute of Statistical Science, Academia Sinica, Taipei Taiwan Kazuhisa Ideta Department of International Cooperation Yodogawa Christian Hospital, Osaka Japan
Kurt M. Jacobson Department of Cardiovascular Medicine University of Wisconsin Hospitals & Clinics Madison, WI USA Mehrdad Jafari Department of studies in Addiction Medicine and Health Promotion Institute Tehran Iran Hamidreza Jamshidi University of Shahidbeheshti Medical Science University, Tehran Iran Slavenka Jankovic´ Institute of Epidemiology, School of Medicine, University of Belgrade Belgrade Serbia
xxxix
xl
Contributors
Sarah J. Jefferies Oncology Centre, Addenbrooke’s Hospital Cambridge UK Fabio Jennings Universidade Federal de Sa˜o Paulo Rheumatology Division, Sa˜o Paulo, SP Brazil Haomiao Jia School of Nursing, Mailman School of Public Health and School of Nursing Columbia University, New York, NY USA Ayodele Jimoh Department of Economics University of Ilorin, Ilorin, Kwara State Nigeria Carole E. Johnson Department of Communication Disorders Auburn University, AL USA Julie Jomeen University of Hull Hull UK Anamaria Jones Universidade Federal de Sa˜o Paulo Rheumatology Division, Sa˜o Paulo, SP Brazil Daniel B. Jones Harvard Medical School, Beth Israel Deaconess Medical Center, Boston, MA USA Evangelos Kalaitzakis Section of Gastroenterology and Hepatology, Department of Internal Medicine, Sahlgrenska University Hospital Gothenburg Sweden
Kamyar Kalantar-Zadeh Harold Simmons Center for Kidney Disease Research and Epidemiology Los Angeles BioMedical Research Institute at Harbor-UCLA Medical Center, the David Geffen School of Medicine at UCLA and the UCLA School of Public Health Torrance Betu¨l Kalender Departments of Nephrology, University of Kocaeli Turkey Fla´vio Kapczinski Bipolar Disorders Program and Laboratory of Molecular Psychiatry Hospital de Clinicas de Porto Alegre Brazil and INCT Translational Medicine Sherrie H. Kaplan Center for Health Policy Research University of California Irvine, CA USA Markku Kaste Department of Neurology Helsinki University Central Hospital University of Helsinki, Helsinki Finland Marianne Kastrup Centre for Transcultural Psychiatry Psychiatric Centre, University Hospital Copenhagen, Rigshospitalet Copenhagen Denmark Marcia Kauer-Sant’Anna Bipolar Disorders Program and Laboratory of Molecular Psychiatry Hospital de Clinicas de Porto Alegre Brazil and INCT Translational Medicine
Contributors
Scott W. Keith Department of Biostatistics University of Alabama at Birmingham Birmingham, AL USA
Soo Woong Kim Department of Urology, Seoul National University College of Medicine Seoul National University Hospital Seoul Korea
Monika Keller Department of Surgery, University Hospital Steinhoevelstr, Ulm Germany
Jeffrey C. King Maternal-Fetal Medicine Department of Obstetrics, Gynecology & Women’s Health University of Louisville College of Medicine Louisville, KY USA
Nichole R. Kelly Department of Psychology, Virginia Commonwealth University Richmond, VA USA Rory Kennelly Institute for Clinical Outcomes Research and Education (iCORE), St. Vincent’s University Hospital, Dublin Ireland Ronald C. Kessler Department of Health Care Policy Boston, MA USA Mahmud M. Khan Department of Health Systems Management, Tulane University School of Public Health and Tropical Medicine New Orleans, LA USA Kozui Kida Department of Pulmonary Medicine Infection, and Oncology, Respiratory Care Clinic, Nippon Medical School, Tokyo Japan Hyeon Hoe Kim Department of Urology, Seoul National University College of Medicine, Seoul National University Hospital Seoul Korea
Hristos G. Kirpizidis Department of Cardiology, Hospital Thessaloniki Greece J. Andre Knottnerus Department of General Practice, School for Public Health and Primary Care: Caphri Maastricht University, Maastricht The Netherlands Nazmiye Kocaman Department of Psychiatry, Department of Consultation Liaison Psychiatry, University of Istanbul, Istanbul Faculty of Medicine C¸apa, Istanbul Turkey Nikola Kocev Institute of Medical Statistics and Informatics, School of Medicine, University of Belgrade, Belgrade Serbia Gerald F. Kominski Department of Health Services, UCLA School of Public Health, Los Angeles, CA USA Joel D. Kopple Division of Nephrology and Hypertension Los Angeles Biomedical Research Institute at Harbor-UCLA Medical Center, the David Geffen School of Medicine at UCLA and the UCLA School of Public Health, Torrance, CA USA
xli
xlii
Contributors
Nenad Kostanjsek World Health Organization, Classification Assessment, Surveys and Terminology (CTS) Team, Geneva Switzerland Paul F. M. Krabbe Multidisciplinary Memory Clinic Slingeland Hospital/Alzheimer Centre Nijmegen University Medical Centre Nijmegen Kruisbergseweg The Netherlands Ja Hyeon Ku Department of Urology, Seoul National University College of Medicine, Seoul National University Hospital, Seoul Korea Renee K. Kuan Arnold Consultancy & Technology LLC New York, NY USA Diana Kuh MRC Unit for Lifelong Health and Ageing Royal Free and University College Medical School, Department of Epidemiology and Public Health London UK Yi Ping Kung School of Nursing, Chang Gung University Tao Yuen Taiwan Ming Hui Kuo Department of Occupational Rehabilitation, Kai-Suan Psychiatric Hospital, Kaohsiung Taiwan Lorna Kwan David Geffen School of Medicine(SN, YA) Jonsson Comprehensive Cancer Center University of California, LA USA
Ulrich Laaser Section of International Public Health University of Bielefeld, Bielefeld Germany Javier Labbe´ Health Department, Association Chilena de Seguridad, Santiago Chile Lucie Laflamme Department of Public Health Sciences Division of International Health, Karolinska Institutet, Stockholm Sweden F. Marc LaForce PATH, Ferney-Voltaire France Erin E. Lalor National Stroke Foundation, Melbourne, Vic Australia Cindy L. K. Lam Family Medicine Unit, the University of Hong Kong Hong Kong SAR Jennifer Lamanna Department of Psychology, Virginia Commonwealth University Richmond, VA USA Hans Landolt Department of Neurosurgery Kantonsspital Aarau Switzerland M. Moghana Lankarani Department of Psychology and Psychiatry Medicine and Health Promotion Institute Tehran Iran Christopher T. Lang Department of Obstetrics and Gynecology The Ohio State University College of Medicine Columbus, OH USA
Contributors
Jan Olov Larsson Department of Woman and Child Health Karolinska Institute, Child and Adolescent Psychiatric Unit, Stockholm Sweden Maximilian Ledochowski Division of Nutrition Medicine, Innsbruck Medical University, Innsbruck Austria Andy H. Lee School of Public Health, Curtin University of Technology, Perth, WA Australia Alain Leplege Department of Philosophy and History of Sciences and REHSEIS, University Paris Diderot, Paris France Beate Lettmeier Department of Public Health, Information Systems and Health Technology Assessment, UMIT - University for Health Sciences, Medical Informatics and Technology and ONCOTYROL Center for Personalized Cancer Medicine Hall i.T./Innsbruck Austria Carol Levin PATH, Seattle, WA USA G. Lewison School of Library, Archive and Information Studies, University College London London UK Chaoyang Li Behavioral Surveillance Branch, Division of Adult and Community Health, National Center for Chronic Disease Prevention and Health Promotion, Centers for Disease Control and Prevention, Atlanta, GA USA
Z. T. Liao Division of Rheumatology, The Third Affiliated Hospital of Sun Yat-sen University, Guangzhou People’s Republic of China Iannis A. Liappas Department of Psychiatry, Medical School University of Athens Greece AAM Lima Centers for Global Health, University of Virginia, Charlottesville, Virginia and Federal University of Ceara´ Fortaleza Brazil Z. M. Lin Division of Rheumatology, The Third Affiliated Hospital of Sun Yat-sen University, Guangzhou People’s Republic of China Wei Liu International Vaccine Institute Kwanak-ku, Seoul Republic of Korea Carrie D. Llewellyn Department of Primary Care & Public Health, Brighton & Sussex Medical School Brighton UK Jose L. Lo´pez-Campos Unidad Me´dico-Quiru´rgica de Enfermedades Respiratorias, Hospital Universitario Virgen del Rocı´o., Seville Spain Goedele Louwagie School of Health Systems and Public Health, University of Pretoria, Pretoria South Africa
xliii
xliv
Contributors
Derek Lowe Evidence base Practice Research Centre (EPRC), Edge Hill University Liverpool, UK
Clelia Madeddu Department of Medical Oncology University of Cagliari Italy
Richard Lowry Division of Adolescent and School Health National Center for Chronic Disease Prevention and Health Promotion Centers for Disease Control and Prevention Atlanta, GA USA
Andreas Maetzel Amgen (Europe) GmbH, Zug, Switzerland Department of Health Policy, Management and Evaluation, University of Toronto Toronto, ON Canada
Erica Lubetkin Department of Community Health and Social Medicine, Sophie Davis School of Biomedical Education/CUNY Medical School New York, NY USA M. Kathleen B. Lustyk Department of Psychology School of Psychology, Family and Community, Seattle Pacific University Seattle, WA USA Pascal Lutumba Unive´rsite´ de Kinshasa Democratic Republic of Congo Christine MacArthur Unit of Public Health, Epidemiology and Biostatics School of Health and Population Sciences, University of Birmingham UK Antonio Maccio` Department of Medical Oncology University of Cagliari Italy Sylvia V. Mackensen Institute and Policlinics of Medical Psychology, University Medical Centre Hamburg-Eppendorf Martinistr Hamburg Germany
Giuseppe Maina Mood and Anxiety Disorders Unit Department of Neurosciences University of Turin Turin Italy Jadranka Maksimovic Institute of Epidemiology, School of Medicine, Belgrade University Belgrade Serbia Natasa Maksimovic´ Institute of Epidemiology, School of Medicine, University of Belgrade, Belgrade Serbia Giovanni Mantovani Department of Medical Oncology University of Cagliari, Cagliari Italy Jamie Manwaring Department of Psychology Washington University St. Louis, MO USA Hilal Maradit-Kremers Department of Health Sciences Research Mayo Clinic, Rochester, MN USA Peter J. Marangos Health Care Analytics United BioSource Corporation Bethesda, MD USA
Contributors
Monica M. Marı´n Caro Geneva University Hospital, Geneva Switzerland Jelena Marinkovic Institute for Medical Statistics and Informatics, School of Medicine, Belgrade University, Silos Belgrade Serbia Sergio Mariotti National Center for Epidemiology Surveillance and Health Promotion, Istituto Superiore di Sanita Italy Ljiljana Markovic-Denic Institute of Epidemiology, School of Medicine, Belgrade University, Belgrade Serbia Meghan L. Marsac Department of Psychology, University of Toledo & Children’s Hospital of Philadelphia, Philadelphia, PA USA
Eric L. Mattesson Division of Rheumatology Mayo Clinic College of Medicine Rochester, Minnesota USA Molly R. Matthews Department of Community Medicine School of Medicine, West Virginia University, Morgantown, WV USA Deborah J. Mawman University Department of Otolaryngology Head-Neck Surgery, Manchester Royal Infirmary, Manchester UK Annette E. Maxwell Division of Cancer Prevention and Control Research, School of Public Health, University of California, Los Angeles, CA USA Fayyaz A. K. Mazari Academic Vascular Surgical Unit, Alderson House, Hull Royal Infirmary, Hull UK
Roger J. Marshall Section of Epidemiology and Biostatistics School of Population Health, University of Auckland New Zealand
Suzanne E. Mazzeo Department of Psychology, Virginia Commonwealth University, Richmond, VA USA
Colin R. Martin School of Health, Nursing and Midwifery University of the West of Scotland, Scotland UK
Karen A. McDonnell Department of Prevention and Community Health, George Washington University School of Public Health and Health Services Washington, DC USA
Faith Martin Department of Primary Health Care University of Oxford UK
Kathleen McElhone Department of Rheumatology Royal Blackburn Hospital, Blackburn UK
Yasuyo Matsumoto Department of International Cooperation Yodogawa Christian Hospital, Osaka Japan
Colman McGrath Faculty of Dentistry, University of Hong Kong, Hong Kong SAR People’s Republic of China
xlv
xlvi
Contributors
Cameron N. McIntosh ARC Epidemiology Unit, University of Manchester Rutherford House (Unit 4) Manchester Science Park, Manchester UK Pamela G. McKay Health Information and Research Division Statistics Canada Canada Matthew McKenna Centers for Disease Control and Prevention, US Department of Health and Human Services Atlanta, GA USA Zoe J. McKeough Faculty of Health Sciences, The University of Sydney, Lidcombe Australia Leslie McKnight Department of Surgery McMaster University, Hamilton Ontario Canada Tapan A. Mehta Academic Vascular Unit University of Hull, Hull UK Atte Meretoja Department of Neurology Helsinki University Central Hospital University of Helsinki, Helsinki Finland Ruben A. Mesa Division of Hematology, Mayo Clinic Rochester, MN USA Lynn B. Meuleners School of Public Health, Curtin University of Technology, Perth Australia
Dimitri P. Mikhailidis Department of Clinical Biochemistry (Vascular Disease Prevention Clinics) Royal Free Hospital Campus University College London Medical School University College London, London UK Rafael Mikolajczyk Department of Public Health Medicine School of Public Health, University of Bielefeld, Bielefeld Germany John D. Miller Department of Thoracic Surgery, Thoracic Surgery McMaster University St. Joseph’s Hospital, Hamilton, ON Canada Gita D. Mishra MRC Unit for Lifelong Health and Ageing Royal Free and University College Medical School, Department of Epidemiology and Public Health, London UK Jan Mitchell Department of Psychology, Royal Holloway University of London, Surrey UK Seyed M. Moghadas Institute for Biodiagnostics, National Research Council Canada Winnipeg, MB Canada Maureen P. M. H. Rutten-van Mo¨lken Institute for Medical Technology Assessment, Erasmus MC, Rotterdam The Netherlands Debapriya Mondal School of Earth, Atmospheric and Environmental Sciences, The University of Manchester UK
Contributors
Ali Montazeri Iranian Institute for Health Sciences Research (IHSR), Tehran Iran Marjory L. Moodie Deakin Health Economics Deakin University, Vic Australia Takashi Motegi Department of Pulmonary Medicine Infection, and Oncology, Respiratory Care Clinic, Nippon Medical School, Tokyo Japan Sayed J. Mousavi Department of Physical Therapy Tehran University of Medical Sciences Tehran Iran Luc Mouthon Department of Internal Medicine Paris Descartes University Reference Center for Necrotizing Vasculitides and Systemic Sclerosis Cochin Hospital, Assistance Publique-Hoˆpitaux de Paris (AP-HP) Paris France Peter Muennig Mailman School of Public Health, Columbia University, New York, NY USA Nikolai Mu¨hlberger Department of Public Health, Information Systems and Health Technology Assessment UMIT - University for Health Sciences, Medical Informatics and Technology and ONCOTYROL Center for Personalized Cancer Medicine Hall i.T./Innsbruck Austria
Pedro Mun˜oz Department of Psychiatry Mental Health Centre Ortuella Ortuella, Vizcaya Spain Jean W. M. Muris Research School Public Health and Primary Care (Caphri) Medicine and Life Sciences Department of General Practice Maastricht University, MD Maastricht The Netherlands Emmanuel A. M. Mylanus Department of Otorhinolaryngology University Medical Centre Sint Radboud Nijmegen The Netherlands Shunichi Namiki Department of Urology, Tohoku University Graduate School of Medicine Sendai Japan Giorgia La Nasa Bone Marrow Transplant Centre, R. Binaghi Hospital, University of Cagliari Cagliari Italy Jamil Natour Universidade Federal de Sa˜o Paulo, Rheumatology Division Sa˜o Paulo, SP Brazil E. A. S. Nelson Department of Paediatrics, The Chinese University of Hong Kong, Prince of Wales Hospital, Shatin, Hong Kong SAR People’s Republic of China Geoffrey Nelson Department of Psychology, Wilfrid Laurier University, Waterloo, ON Canada
xlvii
xlviii
Contributors
Penelope Nestel Institute of Human Nutrition, Southampton General Hospital, Southampton UK
Angel Ois Unitat d’Ictus. Servei de Neurologia Hospital del Mar, Barcelona Spain
Gabriele Neurauter Division of Biological Chemistry, Biocenter Innsbruck Medical University Innsbruck Austria
Lynn J. Okamoto Health Care Analytics United BioSource Corporation Bethesda, MD USA
Linda M. Niccolai Yale School of Medicine – Epidemiology and Public Health New Haven, CT USA
Pedro R. Olivares University of Extremadura, Ca´ceres Spain
J. Curtis Nickel Department of Urology, Queen’s University, Kingston, ON Canada Esther Nu´n˜ez Health Services. Institut Catala` de la Salut. Av. Drassanes, Barcelona Spain Montserrat Nu´n˜ez Educational and Functional Readaptation Unit, Rheumatology Department (ICEMEQ) Hospital Clı´nic, Barcelona Spain Miriam Nun˜o UCLA School of Public Health, Department of Biostatistics, Los Angeles, CA USA
S. Jay Olshansky School of Public Health, University of Illinois at Chicago, Chicago, IL USA Ikushi Onozaki The Research Institute of Tuberculosis, Japan Anti-Tuberculosis Association Kiyose Tokyo Japan Reinaldo Oria Centers for Global Health, University of Virginia, Charlottesville, Virginia and Federal University of Cear Fortaleza Brazil Ferdiando di Orio Department of Internal Medicine and Public Health, University of L’Aquila, P.le S. Tommasi Italy
Elissa Oh Institute for Healthcare Studies Northwestern University, Chicago, IL USA
Robyn L. Osborn Department of Medical and Clinical Psychology, Uniformed Services University of the Health Sciences Bethesda, MD USA
Seung-June Oh Department of Urology, Seoul National University College of Medicine, Seoul National University Hospital, Seoul Korea
Manathip Osiri Division of Rheumatology Department of Medicine Chulalongkorn University Thailand
Contributors
Lena Oxelmark Department of Neurobiology, Care Sciences, Division of Nursing Karolinska Institutet, Huddinge Sweden ¨ zkan Mine O Department of Psychiatry Department of Consultation Liaison Psychiatry University of Istanbul Istanbul Faculty of Medicine, C¸apa Istanbul Turkey ¨ zkan Sedat O Department of Psychiatry Department of Consultation Liaison Psychiatry, University of Istanbul Istanbul Faculty of Medicine, C¸apa Istanbul Turkey ¨ zmen Vahit O Department of Consultation Liaison Psychiatry, University of Istanbul, C¸apa Istanbul Turkey Jesus A´. Padierna Department of Psychiatry Hospital de Galdakao Usansolo Galdakao, Vizcaya Spain Demosthenes B. Panagiotakos Department of Nutrition Science – Dietetics, Harokopio University Athens Greece Angelo Pappalardo Department of Neurology, Multiple Sclerosis Centre – University of Catania Italy
Rajeev Parameswaran Royal Devon & Exeter NHS Foundation Trust, Exeter UK Jae-Hyun Park National Cancer Control Research Institute National Cancer Center Goyang-si Korea Donald M. Parkin Clinical Trials Service Unit and Epidemiological Studies Unit University of Oxford, Oxford UK Jayadeep Patra Public Health and Regulatory Policies Section, Centre for Addiction and Mental Health, Toronto, ON Canada Francesco Patti Department of Neurology, Multiple Sclerosis Centre- University of Catania Italy Mathew Pearson Division of Urologic Surgery, The University of North Carolina at Chapel Hill Chapel Hill NC USA Tatjana Pekmezovic Institute of Epidemiology, School of Medicine, University of Belgrade Belgrade Serbia Jorge Escobedo-de la Pen˜a Mexican Institute of Social Security, Regional General Hospital ‘‘Carlos Macgregor Sanchez-Navarro’’ Epidemiologic Research Unit, Mexico City Mexico
xlix
l
Contributors
Konrad Pesudovs NH&MRC Centre for Clinical Eye Research Department of Ophthalmology, Flinders Medical Centre and Flinders University Bedford Park, SA Australia Dragica Pesut Institute of Lung Diseases and Tuberculosis, Clinical Centre of Serbia2 Visegradska, Belgrade Serbia Claude Pichard Clinical Nutrition, Geneva University Hospital, Geneva Switzerland J. Rush Pierce Department of Internal Medicine, Texas Tech University Health Sciences Center, TX USA Relana Pinkerton Centers for Global Health, University of Virginia, Charlottesville Virginia and Federal University of Ceara´, Fortaleza Brazil Christos Pitsavos First Cardiology Clinic, Hippokration Hospital, School of Medicine University of Athens, Athens Greece Serge Poiraudeau Department of Rehabilitation, Paris Descartes University, Cochin Hospital INSERM Institut Fe´deratif de Recherche sur le Handicap, Paris France David A. Polya University of Manchester UK Varinsathein Porpit Department of Disease Control, Ministry of Public Health, Immunization Program
Bureau of General Communicable Diseases Bangkok Thailand Maarten Postma Groningen University Institute for Pharmacy, Department of Social Pharmacy and Pharmacoepidemiology, University of Groningen The Netherlands John Powles Department of Public Health and Primary Care, Institute of Public Health Cambridge UK Joana Prata Department of Psychiatry, CHVNGaia/ Espinho Portugal Marie Pre´au Research Unit UMR912, Economic & Social Sciences, Health Systems & Societies INSERM, Marseille, France and IRD, Aix Marseille Universite´, Marseille France Victor R. Preedy Department of Nutrition and Dietetics Nutritional Sciences Division School of Biomedical & Health Sciences King’s College London, London UK Salvatore D. Prete Oncology Unit, S. Giovanni di Dio Hospital Naples Italy Richard L. Prince School of Medicine and Pharmacology University of Western Australia Department of Endocrinology & Diabetes Sir Charles Gairdner Hospital, Perth, WA Australia
Contributors
Raj S. Pruthi Division of Urologic Surgery, The University of North Carolina at Chapel Hill Chapel Hill, NC USA
Mathew R. Reynolds Division of Cardiology, Beth Israel Deaconess Medical Center, Harvard Medical School, Harvard USA
Matin Qaim Department of Agricultural Economics and Rural Development, Georg-AugustUniversity of Goettingen, Goettingen Germany
Lis Ribu Oslo University College, Oslo, Norway
Teresita Ramı´rez-Sa´nchez Mexican Institute of Social Security Assessment Coordination, Mexico City Mexico Francois Rannou Department of Rehabilitation Paris Descartes University, Cochin Hospital, AP-HP, INSERM Institut Fe´deratif de Recherche sur le Handicap (IFR 25) Paris Isidora Ratkov Institute of Epidemiology, School of Medicine, Belgrade University, Belgrade Serbia Paula Ravasco Unidade de Nutric¸a˜o e Metabolismo Instituto de Medicina Molecular Faculdade de Medicina da Universidade de Lisboa Portugal Jurgen Rehm Dalla Lana School of Public Health University of Toronto and Public Health and Regulatory Policies Section Centre for Addiction and Mental Health Toronto, ON Canada Adam Reich Department of Dermatology, Venereology and Allergology, Wroclaw Medical University, Wroclaw Poland
Beatriz Rico-Verdı´n Institute of Services of Social Security to Civil Servants National Medical Centre ‘‘20 de Noviembre’’, Hospital Epidemiology Division, Me´xico City Me´xico Charanjit S. Rihal Cardiac Catheterization Laboratory, Mayo Clinic, Rochester, MN USA Marcel G. M. Olde Rikkert Multidisciplinary Memory Clinic Slingeland Hospital/Alzheimer Centre Nijmegen University Medical Centre Nijmegen Doetinchem The Netherlands Jo Robays Epidemiology and Disease Control Unit Institute of Tropical Medicine, Antwerp Belgium Bjarne Robberstad Department of Public Health and Primary Health Care and Centre for International Health, University of Bergen, Bergen Norway Carie S. Rodgers VA San Diego Healthcare System, Mission Valley Outpatient Services San Diego, CA USA
li
lii
Contributors
Gabriela Rodriguez-Abrego Mexican Institute of Social Security, Regional General Hospital ‘‘Carlos Macgregor Sanchez-Navarro’’ Epidemiologic Research Unit, Mexico City Mexico Michael Roerecke Public Health and Regulatory Policies Section, Centre for Addiction and Mental Health, Toronto, ON Canada Simon N. Rogers Evidence base Practice Research Centre (EPRC), Edge Hill University, Liverpool UK Kelly J. Rohan Department of Psychology John Dewy Hall The University of Vermont Burlington, VT USA Fausto Roila Medical Oncology Division Regional Hospital Ospedale ‘S. Maria della Misericordia’, S Andrea delle Fratte, Perugia Italy Martin Ro¨o¨sli Institute of Social and Preventive Medicine University of Bern Switzerland Jaume Roquer Unitat d’Ictus. Servei de Neurologia Hospital del Mar, Barcelona Spain Adriane R. Rosa Bipolar Disorders Program, Clinical Institute of Neuroscience, Hospital Clinic of Barcelona Barcelona Spain
Lori A. Roscoe Center for Hospice, Palliative Care & End-of-Life Studies at USF, University of South Florida, Tampa, FL USA W. Ro¨ssler Department of General and Social Psychiatry, University of Zu¨rich, Zurich Switzerland Benedetta Ruggeri ASUR Marche Ospedale ‘Mazzoni’ AP Italy Mehdi Russel University of Social Welfare and Rehabilitation, Tehran Iran Danny Ruta Institute of Health Society, University of Newcastle UK Shakeel R. Saeed University College London Ear Institue The Royal National Throat, Nose & Ear Hospital and Royal Free Hospital London UK Susan Saegert Department of Human and Organizational Development, Peabody College Vanderbilt University, Nashville, Tennesse USA Rengaswamy Sankaranarayanan Screening Group, International Agency for Research on Cancer, Lyon France Milena Sˇantric´-Milic´evic´ Institute of Epidemiology, School of Medicine, University of Belgrade, Belgrade Serbia
Contributors
Franco Sassi Department of Social Policy, London School of Economics and Political Science London UK Marco Scarpa Department of Surgical and Gastroenterological Science, Sezione di Clinica Chirurgica I, University of Padova Italy Jean Francois Schemann University Victor Se´galen – Bordeaux Bordeaux France Christian Schneider Department of Neurosurgery Kantonsspital Aarau, Aarau Switzerland Douglas D. Schocken Department of Internal Medicine Division of Cardiovascular Disease, USF College of Medicine and H. Lee Moffitt Cancer Center and Research Institute, Tampa, FL USA Carla J. M. Scho¨lzel-Dorenbos Multidisciplinary Memory Clinic Slingeland Hospital/Alzheimer Centre Nijmegen University Medical Centre Nijmegen Doetinchem The Netherlands Anette Schrag University Department of Clinical Neurosciences, Royal Free and University College Medical School London UK Katharina Schroecksnadel Department of Internal Medicine Innsbruck Medical University, Innsbruck Austria
Ursula Schu¨tte Department of Prosthetic Dentistry Technical University Dresden, Dresden Germany Ruth Schwarzer Department of Public Health, Information Systems and Health Technology Assessment, UMIT - University for Health Sciences, Medical Informatics and Technology and ONCOTYROL Center for Personalized Cancer Medicine Hall i.T./Innsbruck Austria Josep M. Segur Knee Section Orthopaedic Surgery Department (ICEMEQ), Hospital Clı´nic Barcelona Spain John W. Sellors Department of Family Medicine McMaster University, Hamilton, ON Canada Sonalee Shah Global Health Economics Outcomes and Reimbursement Bayer HealthCare Pharmaceuticals Montville, NJ USA Navvab Shamspour Department of Psychology and Psychiatry Medicine and Health Promotion Institute Tehran Iran Tait Shanafelt Division of Hematology, Mayo Clinic Rochester, MN USA Chen Shao Department of Urology, Xinjing Hospital Fourth Military Medical University, Shanxi Province People’s Republic of China
liii
liv
Contributors
Amir Sharafkhaneh VAMC Sleep Center (111-i), Houston TX USA Hossein Sharafkhaneh VAMC Sleep Center (111-i), Houston, TX USA Alexandra Shaw AP Consultants Andover, Hampshire UK Jianzhao Shen Pharmaceutical Product Development Wilmington, NC USA Donald S. Shepard Schneider Institutes for Health Policy Heller School, Brandeis University Waltham, MA USA Uwe Siebert Department of Public Health, Information Systems and Health Technology Assessment, UMIT - University for Health Sciences, Medical Informatics and Technology and ONCOTYROL Center for Personalized Cancer Medicine Hall i.T./Innsbruck Austria and Institute for Technology Assessment Department of Radiology, Massachusetts General Hospital, Harvard Medical School/ Department of Health Policy and Management Harvard School of Public Health, Boston, MA USA Paul A. Simon Office of Chronic Disease and Injury Prevention, Los Angeles County Department of Public Health Los Angeles, CA USA
Sandra B. Sipetic-Grujicic Institute of Epidemiology, School of Medicine, Belgrade University Belgrade Serbia Hans Smedje The Department of Neuroscience Uppsala University, Child and Adolescent Psychiatric Unit, Uppsala Sweden Ad F. M. Snik Department of Otorhinolaryngology University Medical Centre Sint Radboud Nijmegen The Netherlands Min-Woong Sohn Center for Management of Complex Chronic Care Hines VA Hospital, Hines, IL USA Dale Spence Nursing and Midwifery Research Unit Queen’s University Belfast, Belfast Northern Ireland UK Bruno Spire Southeastern Health Regional Observatory (ORS-PACA), Marseille France Research Unit UMR912 ‘‘Economic & Social Sciences Health Systems & Societies’’ INSERM, Aix Marseille Universite´ Marseille France Gaby Sroczynski Department of Public Health, Information Systems and Health Technology Assessment, UMIT - University for Health Sciences, Medical Informatics and Technology and ONCOTYROL Center for Personalized Cancer Medicine Hall i.T./Innsbruck Austria
Contributors
David P. Steensma Division of Hematology, Mayo Clinic Rochester, MN USA Alexander J. Stein Institute for Prospective Technological Studies, Sevilla Spain Katarina Stengler Department of Psychiatry, University of Leipzig, Semmelweisstr, Leipzig Germany Marilyn Stern Departments of Psychology & Pediatrics Virginia Commonwealth University Richmond, VA USA Sunita M. Stewart Department of Community Medicine Hong Kong University and Department of Psychiatry, The University of Texas Southwestern Medical Center Dallas, TX USA Helen Stokes-Lampard Primary Care Clinical Sciences, School of Health and Population Sciences University of Birmingham UK Henrik Støvring Research Unit for General Practice University of Southern Denmark, Odense Denmark Timo E. Strandberg Department of Health Sciences/Geriatrics University of Oulu, and Oulu University Hospital, Unit of General Practice, Oulu Finland Jacqueline J. M. H. Strik Department of Psychiatry and Psychology, Maastricht University Medical Centre:
MUMC, Maastricht University, Maastricht The Netherlands Gerold Stucki Department of Physical Medicine and Rehabilitation, Munich University Hospital Ludwig Maximilian University, Munich Germany Jose A. Suaya Schneider Institutes for Health Policy Heller School, Brandeis University Waltham, MA USA Machi Suka Department of Preventive Medicine St. Marianna University School of Medicine, Kawasaki, Kanagawa Japan Chutima Suraratdecha PATH, Seattle, WA USA Erla K. Svavarsdottir University of Iceland, Eirbergi Iceland Jacek C. Szepietowski Department of Dermatology Venereology and Allergology Wroclaw Medical University, UI. Wroclaw Poland Charles Taieb Department Pharmaco-Economique Pierre Fabre SA, Boulogne, sur Seine France Guoyu Tao Division of Sexually Transmitted Disease Prevention, National Center for HIV STD and TB Prevention Centers for Disease Control and Prevention, Atlanta, GA USA
lv
lvi
Contributors
Turgut Tatlisumak Department of Neurology Helsinki University Central Hospital University of Helsinki Helsinki Finland
David R. Thompson Department of Health Sciences and Department of Cardiovascular Sciences University of Leicester, Leicester UK
Lee-Suan Teh Department of Rheumatology, Royal Blackburn Hospital, Blackburn UK
Amanda G. Thrift Baker IDI Heart & Diabetes Institute, Melbourne VIC Australia
Darija K. Tepavcevic Institute of Epidemiology School of Medicine University of Belgrade Belgrade Serbia
Sergio Tiberti Department of Internal Medicine and Public Health, University of L’Aquila, P.le S. Tommasi Italy
Zorica Terzic´-Sˇupic´ Institute of Social Medicine School of Medicine University of Belgrade, Belgrade Serbia Hla-Hla Thein National Centre in HIV Epidemiology and Clinical Research, The University of New South Wales, Sydney Australia Anastasia Theodoridou Department of General and Social Psychiatry, University of Zurich, Zu¨rich Switzerland Achilleas Thoma Departments of Surgery and Clinical Epidemiology and Biostatistics McMaster University, Hamilton Ontario Canada David R. Thomas Division of Geriatric Medicine, Saint Louis University Health Sciences Center, Saint Louis, MO USA
William W. K. To Department of Obstetrics & Gynaecology United Christian Hospital, Kwun Tong Hong Kong Pablo Tomas-Carus Department of Sports and Health University of Evora Portugal Pinar Topsever Department of Family Medicine, Kocaeli University, Kocaeli Turkey Paul R. Torgerson Institute of Parasitology, University of Zurich, Winterthurestrasse, Zurich Switzerland Jose L. Torres-Cosme National Institute of Perinatology Public Health Division, Mexico City Mexico Gabor To´th Department of Biology and Function in the Head and Neck (A603), Yokohama City University Graduate School of Medicine Yokohama Japan
Contributors
F. P. Treasure Eastern Cancer Registration & Information Centre, Unit C - Magog Court Shelford Bottom, Cambridge UK
Robert F. Valois Department of Community Medicine School of Medicine, West Virginia University, Morgantown, WV USA
Dean A. Tripp Departments of Psychology Anesthesiology, Urology Queen’s University, Kingston, ON Canada
Ron T. Varghese Department of Community Medicine Government Medical College Thiruvananthapuram, Kerala India
Kuan-Yi Tsai Department of Community Psychiatry Kai-Suan Psychiatric Hospital, Kaohsiung Taiwan
Massimiliano Veroux Department of Surgery, Transplantation and Advanced Technologies Organ Transplant Unit, University Hospital of Catania Italy
Junichi Tsukada Cancer Chemotherapy Center, University of Occupational and Environmental Health Japan Mamoru Tsukuda Department of Biology and Function in the Head and Neck (A603), Yokohama City University Graduate School of Medicine Kanazawa-ku, Yokohama Japan Atsuro Tsutsumi Institute of Biomedical Research and Innovation (IBRI), Kobe Japan Hidetaka Uramoto Cancer Chemotherapy Center, University of Occupational and Environmental Health Japan Marjan Vaez Department of Clinical Neuroscience Section of Personal Injury Prevention Karolinska Institutet, Stockholm Sweden Francesca Valent Istituto di Igiene ed Epidemiologia Azienda Ospedaliero-Universitaria di Udine, Udine, Italy
Thomas R. Vetter Department of Anesthesiology, University of Alabama School of Medicine and Department of Health Policy and Organization, University of Alabama at Birmingham School of Public Health Birmingham, AL USA Ana Vieta Health Economics and Outcomes Research IMS Health Spain K. Vijayakumar Department of Community Medicine Government Medical College Thiruvananthapuram India Ignacio Villasen˜or-Ruiz Analytical Control and Coverage Extension Commission Direction, Ministry of Health Mexico City Mexico Ad J. J. M. Vingerhoets Department of Clinical, Developmental and Cultural Psychology Tilburg University, Tilburg The Netherlands
lvii
lviii
Contributors
Hristina Vlajinac Institute of Epidemiology School of Medicine University of Belgrade Belgrade Serbia Margaretha Voss Department of Clinical Neuroscience Section of Personal Injury Prevention Karolinska Institutet, Stockholm Sweden Jolanda De Vries CoRPS, Department of Medical Psychology Tilburg University, Tilburg The Netherlands Anne Vuillemin EA 4003, Faculty of Medicine, Nancy Nancy-University School of Public Health France Eric M. Wallen Division of Urologic Surgery, The University of North Carolina at Chapel Hill Chapel Hill NC USA Michael Walter Department of Prosthetic Dentistry Faculty of Medicine Carl Gustav Carus, Technical University Dresden, Dresden Germany Julio C. Walz Bipolar Disorders Program and Laboratory of Molecular Psychiatry Hospital de Clinicas de Porto Alegre Brazil and INCT Translational Medicine Jung-Der Wang Institute of Occupational Medicine and Industrial Hygiene, College of Public Health, National Taiwan University Taipei Taiwan
Li Yan Wang Division of Adolescent and School Health National Center for Chronic Disease Prevention and Health Promotion Centers for Disease Control and Prevention Atlanta, GA USA Philip S. Wang Division of Services and Intervention Research, National Institute of Mental Health, Bethesda, MD USA Wen-Zhi Wang Department of Neuroepidemiology Beijing Neurosurgical Institute, Beijing China Patti Wansley VA San Diego Healthcare System, Mission Valley Outpatient Services (116A4Z), CA USA Elizabeth Waters McCaughey Centre, VicHealth Centre for the Promotion of Mental Health and Community Wellbeing, School of Population Health, University of Melbourne, VIC Australia Theresia Weber Department of Surgery, University Hospital Steinhoevelstr, Ulm Germany Guenter Weiss Department of Internal Medicine, Innsbruck Medical University Innsbruck Austria Geertjan Wesseling Department of Respiratory Medicine, Maastricht University Medical Centre: MUMC, Maastricht University Maastricht The Netherlands
Contributors
Steven D. Wexner Department of Colorectal Surgery Cleveland Clinic Florida, FL USA Denise Wilfley Departments of Psychiatry, Medicine, Pediatrics and Psychology Washington University St. Louis, MO USA Ross Wilkie Primary Care Musculoskeletal Research Centre, Keele University, Keele. Newcastleunder-Lyme, Staffordshire UK Wolfram Windisch Department of Pneumology, University Hospital Freiburg, Killianstrasse 5 Freiburg Germany Desmond C. Winter Institute for Clinical Outcomes Research and Education (iCORE), St. Vincent’s University Hospital, Dublin Ireland Andreas A. J. Wismeijer Department of Clinical, Developmental and Cultural Psychology, Tilburg University Tilburg The Netherlands Hans-Ulrich Wittchen Institute of Clinical Psychology and Psychotherapy and Centre of Clinical Epidemiology and Longitudinal Studies (CELOS) Technische Universitaet Dresden Germany Alison Woodcock Department of Psychology, Royal Holloway, University of London, Egham Surrey UK
Mark Woodward Department of Medicine Mount Sinai School of Medicine New York USA Yangfeng Wu Department of Epidemiology and Biostatistics, Peking University School of Public Health, The George Institute, China Beijing China Gaoqiang Xie Division for CVD Prevention and Control Network, Cardiovascular Institute and Fuwai Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing China Zhi-Yi Xu Shi-Ji-Tong-Le, Shangai China Shingo Yamabe Department of International Cooperation Yodogawa Christian Hospital Osaka Japan Koich Yamada Department of Pulmonary Medicine Infection, and Oncology, Respiratory Care Clinic, Nippon Medical School, Tokyo Japan Renchi Yang Department of Hematology Institute of Hematology and Blood Diseases Hospital Chinese Academy of Medical Sciences Peking Union Medical College Tianjin China Chao-Hsing Yeh School of Nursing, University of Pittsburgh US
lix
lx
Contributors
Nguyen T. Yen Department of Epidemiology, National Institute of Hygiene and Epidemiology (NIHE), Hanoi Vietnam Paul S. F. Yip Centre for Suicide research and Prevention The University of Hong Kong Hong Kong China Seok-Jun Yoon Department of Preventive Medicine College of Medicine, Korea University Seongbuk-ku, Seoul Republic of Korea Katsumi Yoshida Department of Preventive Medicine St. Marianna University School of Medicine Kawasaki, Kanagawa Japan Zobair M. Younossi Department of Preventive Medicine St. Marianna University School of Medicine Sugao, Miyamae-ku, Kawasaki Kanagawa Japan Stefan Zeuzem Department of Internal Medicine Gastroenterology, Hepatology Pneumology and Endocrinology Johann Wolfgang Goethe-University Frankfurt am Main Germany James X. Zhang Department of Pharmacy, Virginia Commonwealth University, VA USA Lei Zhang Department of Epidemiology, Fourth Military Medical University of PLA, Shanxi
Province China Puhong Zhang Division of NCD Control and Community Health, Chinese Center for Disease Control and Prevention Beijing China Zuo-Feng Zhang Department of Epidemiology, UCLA School of Public Health, Los Angeles, CA USA Yuejen Zhao Department of Health and Community Services, Northern Territory Australia Zeping Zhou Department of Hematology, Institute of Hematology and Blood Diseases Hospital Chinese Academy of Medical Sciences Peking Union Medical College, Tianjin China Kun Zhu School of Medicine and Pharmacology University of Western Australia Department of Endocrinology and Diabetes Sir Charles Gairdner Hospital Perth, WA Australia Steven E. Zimmet Dermatology & Phlebology department Austin, TX USA Keith J. Zullig Department of Community Medicine School of Medicine, West Virginia University, Morgantown, WV USA R. Beatriz Zurita Garza Abt Associates Inc. Bethesda, MD USA
Part 1
1
Instruments and Methodological Aspects 1.1 Instruments Used in the Assessment of Disease Impact
1 The International Classification of Functioning, Disability and Health: A Tool to Classify and Measure Functioning G. Stucki . N. Kostanjsek . A. Cieza 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2 2.1 2.2 2.3 2.4 2.5 2.6 2.6.1 2.6.2 2.6.3
The ICF . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 The ICF in the Historical Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 The ICF in the WHO and the UN Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Development of the ICF . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Up-Date and Revision Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 The Structure of the ICF . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 Validity of the ICF . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 Exhaustiveness or Width . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 Precision or Depth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 ICF Framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
3
Implementation of the ICF . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
4 4.1 4.2 4.2.1 4.2.2 4.3 4.3.1 4.3.2 4.4 4.4.1 4.5
ICF-based Classification and Measurement of Functioning . . . . . . . . . . . . . . . . . . . . . . 18 ICF Categories: Building Blocks and Reference Units . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 ICF-based Practical Tools: ICF Checklist and ICF Core Sets . . . . . . . . . . . . . . . . . . . . . . . 19 ICF Checklist . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 ICF Core Sets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 Mapping the World of Measures to the ICF . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 Linkage Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 ICF-based Measurement of Functioning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 Measuring a Single ICF Category . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 Measuring Across ICF Categories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
4.5
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
#
Springer Science+Business Media LLC 2010 (USA)
4
1
The International Classification of Functioning, Disability and Health
Abstract: The endorsement of the > International Classification of Functioning, Disability and Health (ICF) by the 54th World Health Assembly in May 2001 mirrors an important shift in the priorities by the > World Health Organization (WHO). Although WHO has traditionally focused on infection control and mortality reduction WHO now also emphasizes the importance of reducing the burden associated with non-fatal health conditions. WHO has developed the ICF to provide a unified, international and standardized language for describing and classifying health and health-related domains and hence to provide a common framework for health outcome measurement. With the approval of the ICF the WHO member states are now called upon to implement the ICF in the health, education, labour and social sector. The ICF has found immediate interest in the health sciences with currently more than 600 ICF related publications reflecting the interest, relevance, and impact of its application in health research worldwide. The > ICF categories are the discrete, meaningful, universally shared and understood elements which allow users to comprehensively classify and measure > functioning, > disability and health of individuals and populations. They are thus the building blocks for the construction of ICF-based practical tools such as the > ICF Checklist and the > ICF Core Sets as well as clinical measurement instruments such as the ICF Core Set Indices and self-reported measurement instruments such as the World Health Organization Disability Assessment Schedule II (WHODAS II). While ICF-based practical tools such as the ICF Core Sets allow the classification of functioning states, clinical and self-reported measurement instruments allow the measurement and hence the estimation of functioning status or aspects of it in relation to specific purposes. Vice versa, the ICF categories serve as meaningful and universal reference units for reporting and communicating results of measurements of aspects of functioning made with any measurement instrument from the infinite universe of measurement instruments including health-status measures or health-related quality of life measures. List of Abbreviations: CAT, Computer Adaptive Testing; CTS, Classification, Terminology and Standards; DAR, Disability and > Rehabilitation; DIMDI, Deutsches Institut fu¨r Medizinische Dokumentation und Information; FDRG, Functioning and Disability Reference Group; ICD, > International Classification of Diseases; ICF, International Classification of Functioning, Disability and Health; ICF-CY, International Classification of Functioning, Disability and Health for Children and Youth; ICIDH, > International Classification of Impairment, Disabilities and Handicaps; ICIDH-2, International Classification of Impairment, Disabilities and Handicaps-2; ILO, International Labour Organization; ISPRM, International Society of Physical and Rehabilitation Medicine; NACC, North American Collaboration Center; NHP, Nottingham health profile; OMERACT, Outcome Measures in Theumatoid Arthritis Clinical Trials; SF-36, Medical Outcomes Study 36-item short-form health survey; SNOMED, Systemized Nomenclature of Medicine; UN, United Nations; UNCESCO, United Nations Educational, Scientific and Cultural Organization; UNSTAT, United Nations Statistics Division; VAS, Visual Analog Scale; WHA, World Health Assembly; WHO FIC CC Network, World Health Organization Network of the Collaboration Centers for the Family of International Classifications; WHO, World Health Organization; WHODAS II, World Health Organization Disability Assessment Schedule II
1
Introduction
Functioning is the lived experience of people (Bickenbach et al., 1999). It is a universal human experience (Bickenbach et al., 1999), in which body, person and society are intertwined (WHO,
The International Classification of Functioning, Disability and Health
1
2001). Over the life span people may experience a variation in the level of functioning associated with congenital disorders, injuries, acute and chronic health conditions and ageing. The experience of a limitation of functioning or disability thus is part of the human condition (Bickenbach et al., 1999). According to estimates of the World Health Organization (WHO) at least 10% of the world’s population experience disability. This figure is increasing as a result of population growth, medical advances and the ageing process, but also due to malnutrition, war, violence, road-traffic, domestic and occupational injuries and other causes often related to poverty. With the International Classification of Functioning, Disability and Health (ICF) approved by the 54th World Health Assembly in 2001 (WHO, 2001), the WHO provides for the first time a universal and internationally accepted framework and classification (Stucki, 2005a). The ICF is a promising starting point for the integrative understanding of functioning, disability and health and the overcoming of Cartesian dualism of body and mind as well as both sociological and biomedical reductionism (Imrie, 2004). The objective of this chapter is to introduce the reader to the ICF. In the first section we review the history and development of the ICF and describe its structure and validity. In the second section we illustrate how to use the ICF for the classification and measurement of functioning. Finally, we discuss the current state of the implementation and application of the ICF.
2
The ICF
2.1
The ICF in the Historical Perspective
Clinicians have relied on classifications for the diagnosis of health conditions for over 100 years. The International Classification of Diseases (ICD) was first published as a classification of causes of death in 1898 (Hetzel, 1997). In the meantime the International Classification of Diseases is undergoing its 11th revision. The ICD was initially used for actuarial reasons to document death. It was later adopted for epidemiology and by public health to monitor health and interventions. Lately it was used for clinical purposes mainly driven by the need to classify diagnoses in the context of reimbursement systems including diagnostic related groups. By contrast, the first classification of disability, the International Classification of Impairment, Disabilities and Handicaps (ICIDH) was published in 1976 and released for trial purposes in 1980 only. The ICIDH together with models of the Institute of Medicine (Pope and Tarlov, 1991) which are based on Nagi’s model (Nagi, 1965) and the Quebec model (Fougeyrollas et al., 1998) have provided the basis for definitions of > rehabilitation (Stucki et al., 2007a), the development of rehabilitation practice and research (Stucki et al., 2007a), and legislation and policy-making (Stucki, 2005a). The ICIDH model represented a real breakthrough in that the WHO recognized that the > medical model and its associated International Classification of Diseases did not address non-fatal health outcomes. Particularly in Europe, there was considerable interest in the application of the ICIDH as a unifying framework for classifying the consequences of disease during the last 20 years of the twentieth century. E.g. the Council of Europe launched its Recommendation No. R (92) 6 on ‘‘a coherent policy for people with disabilities’’ based on the ICIDH and the Quebec model. However, the ICIDH, which was never approved by the World Health Assembly as an official WHO classification, did not find worldwide acceptance (Bickenbach et al., 1999). It was criticized by the disability community over time for the use of negative terminology, such as handicap, and for not explicitly recognizing the role of the environment in its model. In the
5
6
1
The International Classification of Functioning, Disability and Health
publication of the second edition of the ICIDH in 1993, WHO thus expressed its intention to embark in the development of a successor classification.
2.2
The ICF in the WHO and the UN Perspective
The endorsement of the ICF by the 54th World Health Assembly (WHA) in May 2001 mirrors an important shift in the priorities by the WHO. Although WHO has traditionally focused on infection control and mortality reduction WHO now also emphasizes the importance of reducing the burden associated with non-fatal health conditions. Recognizing the importance of functioning and disability as a major public health issue both in the developed and in the developing world, WHO has developed the ICF to provide a unified, international and standardized language for describing and classifying health and health-related domains and hence to provide a common framework for health outcome measurement. The ICF thus complements indicators that have traditionally focused on deaths and diseases. To complement mortality or diagnostic data on morbidity and diseases is important since they alone do not adequately capture health outcomes of individuals and populations (e.g., diagnosis alone does not explain what patients can do, what their prognosis is, what they need, and what ¨ stu¨n et al., 2004). treatment costs) (U The ICF, which is now coordinated by WHO’s ‘‘Classification, Terminology and Standards (CTS)’’ team serves as reference framework throughout WHO. Most importantly, the ICF is the reference framework of the Disabilities and Rehabilitation Team (DAR) under the Department of Violence and Injury Prevention and Disability. The WHA resolution 58.23 on ‘‘Disability, including prevention, management and rehabilitation’’ approved in May 2005 by the 58th World Health Assembly and coordinated by the DAR team thus recalls the ICF as its framework (Stucki et al., 2007a). As requested by the resolution, the WHO is currently developing a world report on disability and rehabilitation whose structure is based on the ICF framework. While the ICF has been developed by WHO, the specialty agency responsible for health within the United Nations (UN) system, the ICF has been accepted as one of the United Nations social classifications (WHO, 2001). It thus now serves as reference framework for the UN and its other specialty agencies including the United Nations Statistics Division (UNSTAT), the United Nations Educational, Scientific and Cultural Organization (UNESCO) and the International Labour Organization (ILO). Although the ICF is not explicitly mentioned, the understanding of functioning as a universal experience according to the ICF framework is the basis for the characterization of disability in the UN Convention on the Rights of Persons with Disabilities approved on 13 December 2006 at the United Nations Headquarters in New York. While the convention does not establish new human rights, it does define the obligations on states to promote, protect and ensure the rights of persons with disabilities. Most importantly, it sets out the many steps that states must take to create an enabling environment so that persons with disabilities can enjoy inclusion and equal > participation in society.
2.3
Development of the ICF
Coordinated by WHO’s secretariat for mental health and later by the secretariat responsible for classifications and terminology, the ICF was developed in a worldwide collaborative
The International Classification of Functioning, Disability and Health
1
process through the Network of Collaboration Centers for the Family of International Classifications, especially the NACC, the North American Collaboration Center. After three preliminary drafts and extensive international field testing including linguistic and cultural applicability research, the successor classification which was first tentatively named ICIDH-2, the International Classification of Functioning, Disability and Health (ICF) was finalized in 2000 (WHO, 2001). The ICF for Children and Youth, the ICF-CY, was finalized and officially launched in 2007. Until March 2008 the ICF has been translated in several languages. The ICF not only addresses Western concepts but has worldwide cultural applicability. The ICF follows the principle of a universal as opposed to a minority model. Accordingly, it covers the entire life span. It is integrative and not merely medical or social. Similarly, it addresses human functioning and not merely disability. It is multi-dimensional and interactive and rejects the linear linkage between health condition and functioning. It is also etiologically neutral which means functioning is understood descriptively and not caused by diagnosis. It adopts the parity approach which does not recognize an inherent distinction or asymmetry between mental and physical functioning. These principles address many of the criticisms of previous conceptual frameworks and integrate concepts established during the development of the Nagi model (Nagi, 1965) and the Institute of Medicine model of 1991 (Pope and Tarlov, 1991). Most importantly, the inclusion of environmental and personal factors together with the health condition reflects the integration of the two main conceptual paradigms that had been used previously to understand and explain functioning and disability, that is, the medical model and the > social model. The medical model views disability as a problem of the person caused directly by the disease, trauma or other health conditions and calls for individual medical care provided by health professionals. The treatment and management of disability aim at cure and target aspects intrinsic to the person, i.e. the body and its capacities, in order to achieve individual adjustment and behaviour change (Lemert, 1972). By contrast, the social model views disability as the result of social, cultural, and environmental barriers that permeate society. Thus, the management of disability requires social action, since it is the collective responsibility of society at large to make the environmental modifications necessary for the full participation of people with disabilities in all areas of social life (DeJong, 1993; Dixon et al., 2007). The ICF and its framework achieve a synthesis, thereby providing a coherent view of different perspectives of health (Bickenbach et al., 1999).
2.4
Up-Date and Revision Process
The ICF published in 2001 is a first version. Similar to the ICD it will undergo up-dates and ultimately a revision process. The up-date is prepared by WHO’s CTS team in collaboration with the relevant committees and the ‘‘Functioning and Disability Reference Group’’ (FDRG) of the Network of the Collaboration Centers for the Family of International Classifications (WHO FIC CC Network). The up-date will include information obtained in a wide range of testing and validation studies conducted in collaboration with FDRG and in the scientific community. Currently, FDRG is exploring the possibility and methodological approaches to develop a classification of personal factors. In the future the ICF may evolve in a classification which is based on an ontological approach similar to the approach taken by SNOMED (Systemized Nomenclature of Medicine) (Stucki and Grimby, 2004).
7
8
1 2.5
The International Classification of Functioning, Disability and Health
The Structure of the ICF
As shown in > Figure 1-1, the ICF consists of three key components. In short, the first component, > body functions and structures, refers to physiologic functions and anatomic . Figure 1-1 The model of functioning and disability on which the ICF is based. Functioning is an umbrella term for body functions and structures, activities and participation. Disability is an umbrella term for impaired body functions and activities limitation in activities and restriction in participation
parts, respectively; loss or deviations from normal body functions and structures are referred to as impairments. The second component, > activities, refers to task execution by the individual. ‘‘Activity limitations’’ are thus difficulties the individual may have in executing activities (Stucki, 2005a). The third component, > participation, refers to involvement in life situations. ‘‘Participation restrictions’’ are thus problems the individual may experience with such involvement (Stucki, 2005a). These three components are summarized under the umbrella terms > functioning and disability. They are related to and may interact with the health condition (e.g., disorder or disease) and personal and environmental factors. The components of body functions and structures, activities and participation, and environmental factors are classified based on ICF categories. It is conceivable, that a list of personal factors will be developed over the next years. The ICF contains a total of 1,424 meaningful and discrete or mutually exclusive categories. Taken together the ICF categories are cumulative exhaustive and hence cover the whole spectrum of the human experience. The categories are organized within a hierarchically nested structure with up to four different levels as shown in > Figure 1-2. The ICF categories are denoted by unique alphanumeric codes with which it is possible to classify functioning and disability, both on the individual and population level. An example of the hierarchically nested structure is as follow: ‘‘b1 Mental functions’’ (first/ chapter level); ‘‘b130 Energy and drive functions’’ (second level); and ‘‘b1301 Motivation’’ (third level). Based on the hierarchically nested structure of the ICF categories, a higher-level category shares the attributes of the lower-level categories to which it belongs. In our example the use of a higher numbered level category (b1301 Motivation) automatically implies that the lower numbered level category is applicable (b130 Energy and drive functions).
The International Classification of Functioning, Disability and Health
1
. Figure 1-2 The structure of the ICF and the distribution of the ICF’s 1,424 categories across its four components and four levels of hierarchy
Because the ICF categories are always accompanied by a short definition and inclusions and exclusions the information on aspects of functioning can be reported unambiguously. Examples of ICF categories, with their definitions, inclusions and exclusions are shown in > Table 1-1.
2.6
Validity of the ICF
A wide range of studies across world regions and user perspectives have been examined and have provided empirical and theoretical evidence supporting different aspects of the validity of the ICF framework. They include exhaustiveness or width, and precision or depth of the classification.
2.6.1
Exhaustiveness or Width
A classification needs to be exhaustive by its very nature. In relation to the ICF and its categories, exhaustiveness refers to the coverage of the complete spectrum of health and health-related domains that make up the human experience of functioning and disability, and the complete spectrum of environmental factors that influence that experience of functioning and disability. Exhaustiveness is thus closely related to the concept of width, which refers to the number of distinct health and health-related domains at the same level of specification included in the classification. Based on results of published studies, the ICF appears to fulfil the formal criteria of exhaustiveness, especially in relation to the bandwidth of covered domains. In this respect, the results of the studies conducted in the context of the ICF Core Set development (Cieza et al., 2004a; Stucki and Grimby, 2004; Grill et al., 2005a) (> Table 1-2) can be considered ‘‘proof of concept.’’ To the surprise of many clinicians and scientists involved, the ICF has been shown to be a highly comprehensive classification covering virtually all aspects of the patient experience. More specifically, the ICF has covered the spectrum of problems encountered in people with a wide range of conditions and along the continuum of care. Ongoing validation studies
9
10
1
The International Classification of Functioning, Disability and Health
. Table 1-1 Examples of ICF categories with their corresponding code, title, definition and inclusion and exclusion criteria Codea and title, definition, inclusions and exclusions b130 Energy and drive functions General mental functions of physiological and psychological mechanisms that cause the individual to move towards satisfying specific needs and general goals in a persistent manner Inclusions: functions of energy level, motivation, appetite, craving (including craving for substances that can be abused), and impulse control Exclusions: consciousness functions (b110); temperament and personality functions (b126); sleep functions (b134); psychomotor functions (b147); emotional functions (b152) b280 Sensation of pain Sensation of unpleasant feeling indicating potential or actual damage to some body structure Inclusions: sensations of generalized or localized pain, in one or more body part, pain in a dermatome, stabbing pain, burning pain, dull pain, aching pain; impairments such as myalgia, analgesia and hyperalgesia s730 Structure of upper extremity d450 Walking Moving along a surface on foot, step by step, so that one foot is always on the ground, such as when strolling, sauntering, walking forwards, backwards, or sideways Inclusions: walking short or long distances; walking on different surfaces; walking around obstacles Exclusions: transferring oneself (d420); moving around (d455) d920 Recreation and leisure Engaging in any form of play, recreational or leisure activity, such as informal or organized play and sports, programmes of physical fitness, relaxation, amusement or diversion, going to art galleries, museums, cinemas or theatres; engaging in crafts or hobbies, reading for enjoyment, playing musical instruments; sightseeing, tourism and traveling for pleasure Inclusions: play, sports, arts and culture, crafts, hobbies and socializing Exclusions: riding animals for transportation (d480); remunerative and non-remunerative work (d850 and d855); religion and spirituality (d930); political life and citizenship (d950) e1101 Drugs Any natural or human-made object or substance gathered, processed or manufactured for medicinal purposes, such as allopathic and naturopathic medication a The letter b of the ICF refers to body functions, s to body structures, d to activities and participation domains, and e to environmental factors
for the ICF Core Sets from the patient and health professional perspectives (> Table 1-2) have shown that the ICF broadly covers patient problems and aspects of functioning treated by occupational therapists, physiotherapists and psychologists e.g. in patients with rheumatoid arthritis. The results also show that health professionals from different professions differ greatly in their intervention goals, reflecting the importance of validating the ICF from the perspective of many different health professions.
Early postacute context
Grill et al., 2005a
Grill et al., 2005a
Musculoskeletal conditions
Cardiopulmonary conditions
Grill et al., 2005a
Grill et al., 2005a
Neurological conditions
Musculoskeletal conditions
Grill et al., 2005a
Grill et al., 2005a
Neurological conditions
Grill et al., 2005d
Grill et al., 2005b
Grill et al., 2005b
Grill et al., 2005b
Grill et al., 2005b
Acute context
Grill et al., 2005a
Preparatory phase
Scheuringer et al., 2005b
Scheuringer et al., 2005b
Scheuringer et al., 2005b
Literature review
Grill et al., 2005c
Grill et al., 2005c
Grill et al., 2005c
Grill et al., 2005c
Delphi method
Patient Expert perspective perspective
ICF data collection
Protocol paper
ICF Core Set
. Table 1-2 The ICF Core Set development
Boldt et al., 2005
Stier-Jarmer et al., 2005
Scheuringer et al., 2005a
Stoll et al., 2005
Ewert et al., 2005
Consensus conference
Focus groups or patient interviews
Patient perspective
Grill et al., 2006; Mueller et al., 2008
Grill et al., 2006; Mueller et al., 2008
Grill et al., 2006; Mueller et al., 2008
Mueller et al., 2008
Linking
Delphi method
Expert perspective
Validation phase
Nursing resources
Economic perspective The International Classification of Functioning, Disability and Health
1 11
Long term context
Grill et al., 2005a
Geriatric patients
Grill et al., 2005e
Weigl et al., Cieza et al., 2004 2004d Weigl et al., Stucki et al., 2004 2004a
Cieza et al., Ewert et al., Brockow 2004a 2004 et al., 2004a
Cieza et al., Ewert et al., Brockow 2004a 2004 et al., 2004a
Cieza et al., Ewert et al., Brockow et al., 2004a 2004a 2004
Cieza et al., Ewert et al., Cieza et al., 2004a 2004 2004e
Cieza et al., Ewert et al., Wolff et al., 2004a 2004 2004
Osteoarthritis
Osteoporosis
Rheumatoid arthritis
Chronic ischemic heart disease
Diabetes
Weigl et al., Ruof et al., 2004 2004
Weigl et al., Cieza et al., 2004 2004e
Weigl et al., Dreinho¨fer 2004 et al., 2004
Weigl et al., Cieza et al., 2004 2004c
Cieza et al., Ewert et al., Brockow 2004a 2004 et al., 2004a
Low back pain
Weigl et al., Cieza et al., 2004 2004b
Cieza et al., Ewert et al., Brockow 2004a 2004 et al., 2004a
Weigl et al., 2004
Grill et al., 2005f
Wildner et al., 2005
Consensus conference
Chronic widespread pain
Scheuringer et al., 2005b
Scheuringer et al., 2005b
Preparatory phase
Cieza et al., Ewert et al., 2004a 2004
Grill et al., 2005a
Protocol paper
Coenen et al., 2006; Stamm et al., 2005
Grill et al., 2006; Mueller et al., 2008
Grill et al., 2006; Mueller et al., 2008
Kirchberger et al., 2007
Validation phase
1
Cardiopulmonary conditions
. Table 1-2 (continued)
12 The International Classification of Functioning, Disability and Health
Cieza et al., Ewert et al., Geyh et al., 2004a 2004 2004a
Stroke
i.p.
Aringer i.p. et al., 2006
Kesselring i.p. et al., 2008
Tschiesner i.p. et al., 2007
Vieta et al., i.p. 2007
Systemic lupus erythematosus
Multiple sclerosis
Head and neck cancer
Bipolar disorders
i.p.
i.p.
i.p.
i.p.
i.p.
i.p.
i.p.
i.p.
i.p.
i.p.
i.p.
i.p.
Weigl et al., Geyh et al., 2004 2004b
Weigl et al., Brach et al., 2004 2004
The first rows of the table show the phases of the development of the ICF Core Sets. The numbers in the cells refer to the publications reporting on the results of the different phases for the corresponding health conditions. The abbreviation i.p. means in progress. Empty cells means that the corresponding study has not been performed
i.p.
i.p.
i.p.
i.p.
van Echteld i.p. et al., 2006
Spinal cord injury Bieringi.p. Sorensen et al., 2006
Ankylosing spondylitis
Stamm et al., 2007
Cieza et al., Ewert et al., Brockow 2004a 2004 et al., 2004c
Breast cancer
Psoriasis and psoriatic arthrits
Cieza et al., Ewert et al., Brockow Weigl et al., Cieza et al., 2004a 2004 et al., 2004b 2004 2004f
Depression
Weigl et al., Stucki et al., 2004 2004c
Cieza et al., Ewert et al., Wolff et al., 2004a 2004 2004
Obstructive pulmonary diseases
Weigl et al., Stucki et al., 2004 2004b
Cieza et al., Ewert et al., Wolff et al., 2004a 2004 2004
Obesity
The International Classification of Functioning, Disability and Health
1 13
14
1
The International Classification of Functioning, Disability and Health
Further proof to the comprehensiveness of the ICF is the finding that items of a wide range of measurement instruments (> Table 1-3) can be mapped to the ICF. Most importantly, the ICF broadly represents the contents of health-related quality of life measures.
2.6.2
Precision or Depth
The second consideration for a classification is its depth or precision. Depth or precision can be defined as the number of distinct levels of specification differentiated within a health or health-related domain. Ultimately, the proposal, as presented to the World Health Assembly in 2001, was arbitrary. However, there were guiding principles. Most importantly, the level of specification of ICF categories was established in relation to the human experience of people across a wide range of health conditions, along the continuum of care, along the life span and across the WHO regions. Since the ICF categories are intended to be discrete and meaningful elements, they reflect the intuitive level or the level of informed ‘‘lay experts’’ but not the level of ‘‘professional experts’’ in a specific area. Few studies have so far explicitly addressed this issue. A study that linked health-related quality of life measures to the ICF found that items with different content are linked to the same ICF category (Cieza and Stucki, 2005a). This can be seen as an indication that the ICF does not differentiate these categories adequately. One example is the category b152, Emotional functions. In a review of the items of the SF-36 (Medical Outcomes Study 36-Item Short-Form Health Survey) and the NHP (Nottingham Health Profile), different items of these instruments were linked to the same ICF category b152, even though they referred to different emotions. Based on this and other results, the most common emotional functions that could be specified in a future version of the ICF are: sadness, happiness, anxiety, and anger (Cieza and Stucki, 2005a).
2.6.3
ICF Framework
Jette has rightly argued that for ‘‘scientific investigation, a crucial aspect of any conceptual framework is its internal coherence and its ability to differentiate among concepts and categories within the framework (Kaplan, 1964). Without empirical differentiation, conceptual frameworks cannot be investigated and validated. One of the common criticisms of the original ICIDH was that it was difficult to ascertain the boundaries between the basic concepts, each lacked the clarity and distinctness necessary for useful empirical testing (Grimby et al., 1988; Dijkers et al., 2000; Gray and Hendershot, 2000; Johnston and Pollard, 2001). Thus, for the ICF to be useful as a framework for research, it is critical that the classification is clear about the phenomena it classifies with distinct and measurable definitions of each dimension. Without distinct and measurable dimensions, researchers will have trouble using the ICF for hypothesis development, study design and measurement construction’’ (Jette et al., 2003). Currently, only few published studies have empirically investigated the components of the ICF (Jette et al., 2003). A most important question with regard to the components of the ICF framework is the differentiation of activities and participation (Field and Jette, 2007), and, the differentiation between capacity and performance. In the development process of the ICF no clear distinction between the activities and participation component could be made in relation to specified sets of ICF categories. In the first version of the ICF, WHO thus suggest the further investigation of this issue and offers four possibilities to differentiate between these components. A first empirical study exploring the issue found, that there are distinct concepts conforming to the two ICF components activities and participation (Jette et al., 2003).
Long term context
Early postacute context
Context
Measurements/instruments
Stucki et al., 2006
Stamm et al., 2006b
Weigl et al., 2003 Sigl et al., 2006
Osteoarthritis
Osteoarthritis
Low back pain
North American Spine Society Lumbar Spine Outcome Assessment Instrument (NASS); Oswestry Low Back Disability Questionnaire (ODI); Roland-Morris Disability Questionnaire (RMQ)
Western Ontario and McMaster Universities (WOMAC) and LequesneAlgofunctional Indices
Health Assessment Questionnaire (HAQ); Australian/Canadian Osteoarthritis Hand Index (AUSCAN); Cochin scale; Functional Index of Hand OA (FIHOA); Score for Assessment and Qualification of Chronic Rheumatoid Affections of the Hands questionnaire (SACRAH); Arthritis Impact Measurement 2 Short Form questionnaire (AIMS2-SF)
Bariatric Analysis and Reporting Outcome System (BAROS); Bariatric Quality of Life Index (BQL); Lite, Impact of Weight on Quality of Life Questionnaire (IWQOL); LEWIN-TAG Questionnaire (LEWIN-TAG); Obesity Adjustment Survey-Short Form (OAS-SF); Obesity-Related Coping Questionnaire (OCQ); Obesity-Related Distress Questionnaire (ODQ); Obesity Eating Problems Scale (OE); Obesity-Related Problems Scale (OP); Obesity-Related Well-being Questionnaire (ORWELL); Short-Specific Quality of Life Scale (OSQOL); Obesity and Weight-Loss Quality of Life (OWLQOL); Weight-Related Symptom Measure (WRSM)
Grill et al., Functional Independence Measure (FIM), Functional Assessment Measure 2006 (FAM); Barthel Index (BI)
Reference
Obesity
Neurological conditions, Musculoskeletal conditions, Cardiopulmonary conditions, Geriatric patients
Health condition
. Table 1-3 Mapping of measurement instruments to the ICF
The International Classification of Functioning, Disability and Health
1 15
Generic
Sigl et al., 2005 Stucki et al., 2007
Ankylosing spondylitis
Chronic obstructive pulmonary diseases
St. George’s Respiratory Questionnaire (SGRQ); Chronic Respiratory Questionnaire, Standardized Version (CRQ-SAS); Pulmonary Functional Status & Dyspnea Questionnaire, Modified Version (PFSDQM); Pulmonary Functional Status Scale (PFSS); Breathing Problems Questionnaire (BPQ); Seattle Obstructive Lung Disease Questionnaire (SOLDQ); Quality of Life for Respiratory Illness Questionnaire (QOLRIQ); Airway Questionnaires 20 (AQ20); London Chest Activity of Daily Living Scale (LCADL); Maugeri Foundation Respiratory Failure Questionnaire (MRF28); Clinical COPD Questionnaire (CCQ)
Bath Ankylosing Functional Index (BASFI); Dougados Functional Index (DFI); Health Assessment Questionnaire modified for the spondylarthropathies HAQ-S); Revised Leeds Disability Questionnaire (RLDQ)
Stroke Impact Scale (SIS); Stroke-Specific Quality of Life Scale (SSQOL); Stroke and Aphasia Quality of Life Scale (SAQOL-39); Quality of Life Index – Stroke Version (QLI-SV); Stroke-Adapted Sickness Impact Profile-30 (SA-SIP30); Burden of Stroke Scale (BOSS); Quality of Life Instrument for Young Hemorrhagic Stroke Patients (HSQuale)
Quality of Life Questionnaire of the European Foundation for Osteoporosis (QUALEFFO-41); Osteoporosis Assessment Questionnaire (OPAQ 2.0); Osteoporosis Assessment Questionnaire Short Version (OPAQ-SV)
Measurements/instruments
Cieza and Medical Outcomes Study 36-Item Short-Form Health Survey (SF-36); Stucki Nottingham Health Profile (NHP); Quality of Life Index (QLI); World Health 2005 Organization Quality of Life Scale (WHOQOL-BREF); World Health Organisation Disability Assessment Shedule II (WHODASII); European Quality of Life Instrument (EQ-5D)
Geyh et al., 2007
Stroke
Different conditions
Borchers et al., 2005
Reference
Osteoporosis
Health condition
1
Context
. Table 1-3 (continued)
16 The International Classification of Functioning, Disability and Health
Different conditions
Stamm et al., 2004
Geyh et al., 2007
Canadian Occupational Performance Measure (COPM); Assessment of Motor and Process Skills (AMPS); Sequential Occupational Dexterity Assessment (SODA); Jebsen Taylor Hand Function Test (JT-HF); Moberg Picking Up Test (MPUT); Button Test (Button); Functional Dexterity Test (FDT)
Medical Outcomes Study 36-Item Short-Form Health Survey (SF-36); Reintegration to Normal Living Index (RNL); Sickness Impact Profile (SIP); European Quality of Life Instrument (EQ-5D); LHS London Handicap Scale (LHS); Nottingham Health Profile (NHP); Dartmouth COOP Charts (COOP); 15-Dimensional Measure of Health Related Quality of Life Test (15-D); Assessment of Life Habits (LIFE-H); Assessment of Quality of Life (AQoL); Craig Handicap Assessment and Reporting Technique (CHART); Health Utilities Index Mark II (HUI II); Health Status Questionnaire (HSQ); Lancashire Quality of Life Profile (LQLP); Quality of Life Index (QLI); World Health Organization Quality of Life Scale (WHOQOL)
The fourth column shows the instruments which that have been linked to the ICF and from which a content comparison has been performed. The results have been reported in the paper referenced in the third column
Occupational context
Different conditions
The International Classification of Functioning, Disability and Health
1 17
18
1
The International Classification of Functioning, Disability and Health
Another approach to study the validity of the ICF framework is its reflection from the perspective of theoretical or professional models in relation to functioning. e.g. occupational therapy models which focus on occupations and activities of daily living in the context of the environment which can be expected to be closely related to the ICF. In a paper exploring the link of conceptual occupational therapy models to the ICF, the majority of the concepts from three conceptual occupational therapy models could be linked to the ICF (Stamm et al., 2006a). The ICF also proved to be useful as a framework for comparing the similarities and differences of the three conceptual occupational therapy models. The findings of the study also demonstrated that there are strong conceptual connections between the ICF and occupational therapy models, which encourage occupational therapists to use the ICF in their practice (Stamm et al., 2006a).
3
Implementation of the ICF
With the approval of the ICF the WHO member states are now called upon to implement the ICF in the health, education, labour and social sector. The implementation of the ICF is coordinated by the Network of the Collaboration Centers for the Family of International Classifications (WHO FIC CC Network) and its Functioning and Disability Reference Group (FDRG) in which members of WHO and members of WHO Collaborating Centers from all WHO world regions are represented. The mandate of FDRG is to promote the use of the ICF and improve international comparability of functioning and disability data by establishing standardised procedures and implementation guidelines for different applications of the ICF. The ICF has also found immediate interest in the health sciences and particularly rehabilitation (Stucki et al., 2002; Stucki, 2005a). By 2008 there have been over 600 ICF related publications reflecting the interest, relevance, and impact of its application in health research worldwide. The ICF itself has become the focus of interest of scholars worldwide. It has e.g. been critically discussed in a number of papers in recent reports by the Institute of Medicine on the future of disability in America (Field and Jette, 2007). Closely ICF related applications include the use of the ICF for the classification and measurement of functioning as presented in the previous section. It is important to recall that the ICF is relevant for all medical disciplines and allied professional groups. In physiotherapy and occupational therapy, many curricula are now already based on or have integrated the ICF (Allan et al., 2006). Also, following reports on the application of the ICF in rehabilitation (Steiner et al., 2002; Stucki et al., 2002; Rentsch et al., 2003; Cieza and Stucki, 2006a), there are now also reports on the application of the ICF in other medical specialties in which rehabilitation is a major health strategy including psychiatry and rheumatology (Cieza and Stucki, 2006a). OMERACT (Outcome Measures in Rheumatoid Arthritis Clinical Trials), an international group committed to the standardization of outcome measures in rheumatology now uses the ICF as their reference framework (Stucki et al., 2007b).
4
ICF-based Classification and Measurement of Functioning
4.1
ICF Categories: Building Blocks and Reference Units
The ICF categories are the discrete, meaningful, universally shared and understood elements which allow users to comprehensively classify and measure functioning of individuals and
The International Classification of Functioning, Disability and Health
1
populations. They are thus the building blocks for the construction of ICF-based practical tools such as the ICF Checklist (WHO, 2003) and the ICF Core Sets (Cieza et al., 2004a; Stucki and Grimby, 2004; Grill et al., 2005a; Stucki et al., 2005b) as well as clinical measurement instruments such as the ICF Core Set Index currently under development for Ankylosing Spondylitis (Cieza and Stucki, 2008) and self-reported measurement instruments such as the ¨ stu¨n, 2000). WHODAS-II (Epping-Jordan and U While ICF-based practical tools such as the ICF Core Sets allow the classification of functioning states, clinical and self-reported measurement instruments allow the measurement and hence the estimation of functioning status or aspects of it in relation to specific purposes. Vice versa, the ICF categories serve as meaningful and universal reference units for reporting and communicating results of measurements of aspects of functioning made with any measurement instrument from the infinite universe of measurement instruments (Stucki, 2005a). In this context it is important to recall the difference between the mutually exclusive and discrete elements of a classification such as the ICF categories versus measurement items or simply items e.g. of self-reported health status measures. As meaningful and universally shared elements, ICF categories represent constructs while items as indicators of constructs are used to estimate the variation in a construct, e.g. an ICF category. As shown in a following paragraph, there are e.g. many items used in a wide range of self-reported health status measures which can serve as indicators to estimate the level of the ICF category b130 Energy and drive functions (Cieza et al., 2008b) (> Figure 1-3).
4.2
ICF-based Practical Tools: ICF Checklist and ICF Core Sets
To implement the ICF in clinical medicine, service provision and policy, practical tools need to be developed (Stucki et al., 2002). In this context it is important to recall that the ICF has been developed as a reference classification and is not intended to be a practical tool. To address the needs of prospective users, the FDRG of the WHO FIC CC Network collaborates with international organizations in official relation with WHO including ISPRM and a wide range of partners in the development of ICF-based practical tools including the ICF Core Sets. The main challenge to the application of the ICF is the size of the classification system with ¨ stu¨n, the leader of WHO’s CTS team has pointed out that ‘‘a clinician its 1,424 categories. Dr. U cannot easily take the main volume of the ICF and consistently apply it to his or her patients. In daily practice, clinicians will only need a fraction of the categories found in the ICF.’’ Therefore, ‘‘to be useful, practical ICF-based tools need to be tailored to the need of the prospective users without forgoing the information needed for health statistics and health reporting.’’
4.2.1
ICF Checklist
The ICF Checklist is a 12-page, ‘‘short’’ version of the ICF with 125 second-level categories. All information from written records, primary respondent, other informants, and direct observation can be used (http://www3.who.int/icf/checklist/icf-checklist.pdf). It takes around 1h to complete but may take much longer in patients with multiple impairments, activity limitations, and participation restrictions. It has been applied in a wide range of surveys and in studies in the process of developing ICF Core Sets (> Table 1-2).
19
20
1
The International Classification of Functioning, Disability and Health
. Figure 1-3 Rasch scale for measurement items mapped to the ICF category b130 Energy and drive. The x- and the y-axes represent the ICF category interval scale of the continuum energy and drive, with values ranging from 0 to 100. Not all values from 0 to 100 are represented on the y-axis because of space constraints. The 16-items in order of difficulty from the easiest item (bottom) to the most difficult item (top) are presented on the y-axis. The value corresponding to the position of the items is presented next to them. The position of the thresholds of the response options of the items are represented by the bars in the diagram. The different grey tones represent the different response options for each individual item. The vertical arrows represent the position of each of the response options of the ICF Qualifier. RAQoL rheumatoid arthritis quality of life questionnaire; HAQ the health assessment questionnaire; SF-36 the medical outcomes study short form 36; EQ-5D the European Quality of Life instrument; MFI the multidimensional fatigue inventory; CES-D the Center for Epidemiological Studies Depression Scale
4.2.2
ICF Core Sets
4.2.2.1
The ICF Core Set Project
The goal of the ICF Core Set project is to systematically develop parsimonious and hence practical sets of ICF categories for clinical practice, service provision and research and to link the ICF to health conditions as coded with the ICD (Stucki and Grimby, 2004; Stucki et al., 2005b). The ICF Core Sets serve first as practical tools for the documentation of functioning and second as international reference standards for the reporting of functioning (Stucki, 2005a) irrespective of which measurement instruments were used. They are also the starting point for the development of clinical and self-reported measurement instruments (Grill and Stucki, 2008; Cieza and Stucki, 2008a; Cieza et al., 2008c).
The International Classification of Functioning, Disability and Health
1
The ICF Core Set Project is a joint project of the ICF Research Branch of the WHO FIC CC Germany (DIMDI = Deutsches Institut fu¨r Medizinische Dokumentation und Information) at the Institute for Health and Rehabilitation Sciences at the Ludwig-Maximilian-University in Munich, Germany (http://www.ICF-research-branch.org), together with WHO’s CTS team, the International Society of Physical Medicine an Rehabilitation (ISPRM) and a large number of partner organizations and associated institutions as well as committed clinicians and scientists (Stucki and Grimby, 2004; Stucki et al., 2005b). 4.2.2.2
Conceptual Approach
The conceptual approach for the development of the ICF Core Sets was derived from two perspectives: (1) the perspective of people who share the experience of the same condition (e.g. multiple sclerosis) or condition group (e.g. neurological conditions) and (2) the perspective of the health service context along the continuum of care and the life span. 4.2.2.3
ICF Core Sets for the Acute Hospital and (Early) Post-acute Rehabilitation Facilities
The ICF Core Sets for the acute hospital including the ICF Core Sets for neurological, cardiopulmonary and musculoskeletal conditions are intended for use by physicians, nurses, therapists and other health professionals not specialized in rehabilitation care provision (Grill et al., 2005a; Stucki et al., 2005b). By contrast, the ICF Core Sets for (early) post-acute rehabilitation facilities including the ICF Core Sets for neurological, cardiopulmonary and musculoskeletal conditions as well as the ICF Core Set for geriatric patients are intended for use by physicians, nurses, therapists and other health professionals specialized in rehabilitation or geriatric care provision (Grill et al., 2005a; Stucki et al., 2005b). The use of the term early indicates the ‘‘early’’ part of rehabilitation where patients have both, medical needs requiring hospital care, and rehabilitation needs. 4.2.2.4
ICF Core Sets for Chronic Conditions
The ICF Core Sets for chronic conditions are intended for use in the community (Cieza et al., 2004a; Stucki and Grimby, 2004). For each chronic health condition, both a Brief ICF Core Set and a Comprehensive ICF Core Set have been developed. While the ICF Core Sets serve as practical tools for single encounters, minimum data sets for the reporting of clinical and epidemiological studies and health statistics, the Comprehensive ICF Core Sets are intended for use in multidisciplinary settings. 4.2.2.5
Generic ICF Core Set
While the condition and context-oriented ICF Core Sets are useful when classifying functioning for patients with specific health problems in specific health care situations, a parsimonious set of categories is needed to be able to assess and compare functioning across conditions and > contextual factors. The Generic ICF Core Set is currently being developed in an iterative process involving a number of criteria and methodological approaches. A first study in this process examined the explanatory power of determined ICF categories in relation to external standards across the 12 chronic conditions for which condition-specific ICF Core Sets have already been developed (Cieza et al., 2006b). The categories identified as candidate categories from this study are shown in > Table 1-4.
21
22
1
The International Classification of Functioning, Disability and Health
. Table 1-4 ICF categories identified as candidate ICF categories for the Generic ICF Core Set (Cieza et al., 2006b) ICF component
Candidate ICF categories for generic ICF Core Sets
Body functions
b130 Energy and drive functions b152 Emotional functions b230 Vestibular functions b280 Sensation of pain b730 Muscle power functions
Activities and participation
d450 Walking d620 Acquisition of goods and services d640 Doing housework d660 Assisting others d850 Remunerative employment d920 Recreation and leisure
Environmental factors
e450 Individual attitudes of health professionals e580 Health services, systems and policies
. Figure 1-4 Illustration of the process to develop ICF Core Sets
4.2.2.6
Development Process
While there are some singularities in the process of developing ICF Core Sets in relation to the context for which they are being developed, the development as illustrated in > Figure 1-4 involves an international consensus process based on evidence gathered in a preparatory phase and an international testing and validation phase in the six WHO world regions (Africa, the Americas, the Eastern Mediterranean, Europe, South-East Asian, and the Western Pacific) (Cieza et al., 2004a). The preparatory phase consists of: (1) an empirical data collection, based on the ICF, reflecting the perspective and the situation of the patient (2) an expert survey using the
The International Classification of Functioning, Disability and Health
1
Delphi method, (3) a systematic review on outcomes used in observational and experimental clinical studies, which also represents the view of experts. Additionally, for ICF Core Sets now in the preparatory phase (4) a qualitative study using focus group or patient interviews, representing the view of patients complement the methods. The results of the preparatory studies are presented at a consensus conference. They represent the starting point for a structured decision-making and consensus process in which clinicians and health professionals, experts in the field for which the specific ICF Core Set is to be developed, participate. Finally, the ICF Core Sets are tested and validated in an international effort in a wide range of contexts.
4.3
Mapping the World of Measures to the ICF
4.3.1
Applications
Since the ICF is the universal and standardized language to describe and report functioning and health, users need to be able to map the world of measures to the ICF. The qualitative > mapping of measurement instruments to the ICF relies on linkage rules (Cieza et al., 2005b). The quantitative mapping relies on transformations using the Rasch-model (Cieza et al., 2008b). Qualitative mapping is applied for the content comparison of measurement instruments e.g. when studying their comparative content validity. The ICF-based comparison of measurement instruments can therefore assist researchers and clinicians to identify and select a most suited measurement instrument for a specified purpose. ICF-based comparisons also enable researchers to ensure that all ICF categories of a suitable ICF Core Set are covered by candidate measurement instruments and hence to report functioning according to international standards (Stucki, 2005a) as described in the last section of this chapter. > Table 1-3 lists studies which have compared most widely used measurement instruments for specified health conditions as well as a comparison of generic health status measures. Qualitative in combination with quantitative mapping is used for the identification of items addressing the construct covered by a specified ICF category and the construction of Rasch scales to estimate the level of functioning for this category. As we will describe in more detail in the following paragraph this involves the identification of items from measurement instruments which address the construct of a specified ICF category within their scope. Another example of qualitative combined with quantitative mapping is the transformation of information from electronic records (Mayo et al., 2004).
4.3.2
Linkage Methodology
The linking methodology consists of two main steps. The first step refers to the identification of concepts within the health-related information to be translated to the ICF. The second step refers to linking those concepts to the ICF. 4.3.3.1
Step one, Identification of Concepts
The first step, the identification of concepts, varies slightly depending on the origin of the information that is to be translated. In health-status questionnaires, the concepts refer to the
23
24
1
The International Classification of Functioning, Disability and Health
different contents addressed in each of its items. A single item may contain more than one concept. For example, item 8 of the SF-36 ‘‘During the past four weeks, how much did pain interfere with your normal work (including both work outside the home and housework)’’ contains three different concepts ‘‘pain,’’ ‘‘work outside the home,’’ and ‘‘housework.’’ In qualitative data collection with open-ended questions in focus groups, patient interviews or e-mail surveys, the process of identification of concepts is similar to the process followed with questionnaires. However, while in questionnaires the concepts are identified within items, in qualitative data the concepts are identified within ‘‘meaning units.’’ A meaning unit is defined as a specific unit of text of either a few words or a few sentences with a common theme (Karlsson, 1993). A meaning unit division does not follow linguistic grammatical rules. Rather, the text is divided wherever the researcher discerns a shift in meaning (Kvale, 1996). > Table 1-5 presents an example of meaning units identified in an extract of the information collected in a focus group. When linking clinical assessments, concepts refer to the aims with which a clinical assessment was performed. For example, when pulse rate is assessed to measure ‘‘exercise tolerance,’’ this aim is considered the meaningful concept of the clinical assessment ‘‘heart rate.’’ However, when pulse rate is assessed to measure ‘‘heart rate’’ and ‘‘heart rhythm’’ these two aims are considered the meaningful concepts addressed in the same clinical assessment ‘‘heart rate.’’ 4.3.3.2
Step Two, Linking of Concepts to the ICF
When linking clinical interventions, the concepts also refer to the aims with which an intervention was applied. For example, nurses mobilize their patients with different aims, for example, ‘‘mobility improvement’’ or ‘‘prevention of skin ulcer.’’ Thus, ‘‘mobility improvement’’ or ‘‘prevention of skin ulcer’’ is identified as concept for the intervention ‘‘mobilization’’ depending on the aim with which the intervention was performed. After the concepts have been identified, the second step involves the linking of those concepts to the ICF according to ten rules (Cieza et al., 2005b). The most relevant and obvious rule states that concepts must be linked to the ICF category or categories which most precisely represent them. An example of the linkage of concepts to the ICF is shown in > Table 1-5. Both steps of the linking methodology should always be performed by two trained health professionals independently of each other. Thus, after the second step, two independent results of the linking process exist. These results are compared. The reliability of the linking process is evaluated by calculating kappa coefficients (Cohen, 1960) and nonparametric bootstrapped confidence intervals (Efron, 1982) based on the two independent linking results in order to indicate the degree of agreement between the two health professionals. Disagreement regarding the ICF categories selected during the linking process is resolved by structured discussion and an informed decision by a third expert. The result of applying the linking methodology is a list of ICF categories that is equivalent in content to the original health-related information.
4.4
ICF-based Measurement of Functioning
4.4.1
Measuring a Single ICF Category
WHO prosposes the so-called generic qualifier scale that goes from 0 (no problem) to 4 (complete problem) to rate the magnitude or the severity of the problem in each of the ICF categories (> Table 1-6). Considering the > ICF Qualifier, there are, in principle, two
The International Classification of Functioning, Disability and Health
1
. Table 1-5 Illustration of the linkage procedure (Cieza and Stucki, 2008a) with parts of a conversation recorded during a focus group interview. The information has been divided into meaning units, concepts have been identified within the meaning units and they have been linked to ICF categories ID
Transcription divided according to meaning units
Identified concepts
ICF categories
Question by researcher: If you think about your body and mind, what does not work the way it is supposed to? 2
2
My nails break more. I used to have long, strong nails, but now they break easily. Also, my thumbnails split quickly My hands; they are not painful but I have no power. Things often drop
– Breaking nails
b860 Functions of nails
– Thumbnails split
b860 Functions of nails
– No power in hands
b7300 Power of isolated muscles and muscle groups
– Things drop
d440 Fine hand use
– Nails are not strong
b860 Functions of nail
1
For the past couple of years I’ve noticed that my nails are not strong
3
I have always had bad nails. That’s why I can’t – Hair falling out b850 Functions of hair judge whether they’ve become worse. But my due to medication hair has been falling out. Could be due to the medication. It’s hard to say. It’s awful e1101 Drugs
4
I haven’t lost any hair, but I stopped dyeing it. – Stopping dyeing d5202 Caring for hair I thought that, since I already have to take hair such strong medication, I should do without hair dye and let the natural color grow in again – Strong medication (+) without hair loss
e1101 Drugs
[. . . .]
approaches to measure a specified ICF category, i.e. to quantify the extent of variation therein. The first is to use the ICF Qualifier as a rating scale ranging from 0–4. The second is to use information obtained with a clinical test or a patient-oriented instrument and to transform this information into the ICF Qualifier taking into account the percentage values provided by WHO. 4.4.4.1
Direct Coding of the ICF Qualifier
With this approach a physician or health professional integrates all accessible and suitable information from the patient’s history, clinical and technical exams to code a specified category according to established coding guidelines (Reed et al., 2005). To ensure quality in a specific setting, it is advisable to regularly assess the reliability of coding (Grill et al., 2007). > Figure 1-5 shows a simple and informative graphical approach to assess the inter-observer
25
26
1
The International Classification of Functioning, Disability and Health
. Table 1-6 ICF Qualifier with percentage values provided by the WHO ICF Qualifiera
Percentage of problem
0 – NO problem (none, absent, negligible,. . .)
0–4
1 – MILD problem (slight, low,. . .)
5–24
2 – MODERATE problem (medium, fair. . .)
25–49
3 – SEVERE problem (high, extreme,. . .)
50–95
4 – COMPLETE problem (total,. . .)
96–100
a
‘‘Having a problem may mean an impairment, a limitation, a restriction or a barrier, depending on the construct,’’ i.e. depending on whether we are classifying body functions and structures (impairments), activity and participation (limitations or restrictions) or environmental factors (barriers or facilitators)
reliability of ICF Qualifier codes (Grill et al., 2007). The rating of certain ICF categories may be facilitated by complementary instructions provided in addition to the descriptions of the ICF categories as provided in the ICF reference material. > Table 1-7 shows an additional instruction developed by the American Psychological Association (2007) (APA, 2007) for the ICF category b130 Energy and drive functions for which the original description in the ICF reference material is shown in > Table 1-6. Similar instructions have been developed by the American Psychological Association for a large number of ICF categories (APA, 2007). 4.4.4.2
Transformation of Information Obtained with a Clinical Test or a Patient-oriented Instrument
With the second approach, the ICF Qualifier serves as a reference scale. The results from a clinical test or a patient-oriented measurement instrument are transformed into the ICF Qualifier. For many ICF categories there are suitable clinical tests which include standardized expert and technical examinations or patient-oriented measurement instruments which include patient and proxy-reported, self-administered or interview-administered questionnaires which are routinely used in clinical practice or for research purposes. In this case, information already available can be transformed to report the results in the standard language of the ICF. Since the ICF Qualifier is a rating scale for which WHO has provided percentage values as a reference (> Table 1-6), transformation to the ICF Qualifier is straightforward in the case of interval-scaled clinical tests or patient-oriented instruments, which comprehensively and uniquely cover the content of a respective ICF category. For example, the Visual Analog Scale (VAS) to assess pain can be used to address the ICF category b280 Sensation of pain. The values of VAS-Pain can be transformed into an ICF Qualifier in a straightforward manner, since it represents a 100 millimeter (mm) interval scale marked as ‘‘no pain’’ at one end and as ‘‘worst pain’’ at the other (Kvale, 1996). Considering the percentage values of the ICF Qualifier in > Table 1-6, a person marking a level of pain between 0 (zero) and 4 mm would receive the qualifier 0 in the ICF category b280 Sensation of pain, between 5 and 24 mm the Qualifier 1, between 25 and 49 mm the Qualifier 2, between 50 and 95 mm the Qualifier 3, and between 96 and 100 mm the Qualifier 4. In the case where there are no readily available clinical tests or patient-oriented instruments with interval-scale properties that can be used to assess a specified ICF category one
The International Classification of Functioning, Disability and Health
1
. Figure 1-5 Bangdiwala observer agreement chart for the ICF category d430. The chart is a square whose edges are determined by sample size. The edges of the black squares show the number of patients who got identical ratings from both observers. The large bright rectangle shows the maximum possible agreement, given the marginal totals. Partial agreement is showed by including a weighted contribution from off-diagonal cells, here represented by hatching. One observer’s ratings would differ systematically from the other observer’s ratings if all black squares were above or below the diagonal
may consider the construction of an ICF category interval scale using parts of clinical test batteries or selected items of patient-oriented measurement instruments that cover a specified ICF category. > Figure 1-3 illustrates the construction of an interval reference scale using the Rasch model to estimate the level of functioning for b130 Energy and drive functions (Cieza et al., 2008b). Sixteen of the 19 items linked from three instruments did fit the Rasch model and could be integrated in an ICF category interval scale. Based on this principle, clinicians can estimate the level of b130 Energy and drive functions by adding the responses to the 16 items. In clinical practice, one would obviously need only a subset of possibly five items to reliably estimate the level of functioning in b130 Energy and drive functions. Alternatively one may increase efficiency by using computer adaptive testing (CAT). Whatever method is used, the obtained raw scores can then be transformed into the ICF Qualifier which serves as a reference scale. A major advantage of the second approach is that the original format of the items used to construct the ICF category interval scale remains unchanged. Thus, it is possible to use the information provided by items within the context of their original instruments and, at the same time, within the context of the ICF. This application can be extremely useful, given the increasing use of the ICF and the ICF Qualifier as references when documenting and reporting functioning and disability (Jette, 2006).
27
28
1
The International Classification of Functioning, Disability and Health
. Table 1-7 Additional instructions for ICF categories illustrated with the example b130 Energy and drive functions b130: Energy and drive functions General mental functions of physiological and psychological mechanisms that cause the individual to move towards satisfying specific needs and general goals in a persistent manner Inclusions: functions of energy level, motivation, appetite, craving (including craving for substances that can be abused), and impulse control Exclusions: consciousness functions (b110); temperament and personality functions (b126); sleep functions (b134); psychomotor functions (b147); emotional functions (b152) Additional Information This code includes general behavioral tendencies including Energy level b1300 and Motivation b1301 to move toward goals. It also includes the constructs of Appetite b1302 and Craving b1303, which may be general tendencies or relate to specific substances or behaviors (e.g., psychoactive substances, food, gambling). In addition, this code includes Impulse control b1304, which may refer to impulses in general or relate to more specific impulses to engage in particular behaviors. This code and its subcodes should be used only to refer to characteristics or behaviors that are consistent or occur frequently over time, not to single behaviors or transitory states These codes may be useful in a variety of settings. Motivation, craving, and impulse control are often a part of motivational assessment in relation to substance abuse treatment or other treatments that have the goal of reducing, avoiding, or abstaining from particular behaviors (e.g., substance use, overeating, gambling). In such cases, impairments related to these factors may be a part of the disorder. Energy level and motivation may also be important in cases of CNS injury or disease (e.g., stroke), where concerns related to ‘‘lack of initiation’’ or ‘‘mental fatigue’’ may be present, and in patients with psychological disorders such as depression and bipolar disorder. Impairments in impulse control are by definition a part of substance abuse and impulse-control disorders, and may also be a central part of a variety of other psychological disorders including attention deficit hyperactivity disorder, conduct disorder, and bipolar disorder Generally, Energy level b1300 and Motivation b1301 should be reserved for cases in which abnormal levels or significant changes in energy level and motivation occur as a direct result of a disorder, disease process, or injury, or as an effect of treatment (e.g., decreased energy level is a side effect of some medications) Motivation is considered to be particularly important in relation to the success of treatment for many health conditions. However, caution should be exercised in assigning this code. Body functions are meant to be coded with the ICF to the extent that impairments are attributable to a health condition or health-related state, which will not be to the extent that high or low motivation is a general personality characteristic of the individual. This is not to say that it will not be highly relevant to treatment, only that it would correspond more closely in this case to what the ICF identifies as Personal Factors rather than to Body Functions. In addition, Motivation b1301 should not be used to describe an individual’s motivation to comply with a specific treatment, such as physical therapy in rehabilitation programs. Finally, lack of motivation may be used by health care personnel or others in the patient’s social environment as a pejorative explanation for a patient’s lack of progress in treatment, one that attributes the problem to the patient. It is important not to attribute lack of motivation to patients who are physically or mentally unable to perform particular tasks or actions or who are not receiving the most appropriate treatments to help them progress Case Examples Following a stroke, a 67-year-old woman has difficulty selecting or getting started on projects, and often complains of feeling ‘‘too tired’’ and ‘‘mentally worn out’’
The International Classification of Functioning, Disability and Health
. Table 1-7 (continued)
1
A 45-year-old man with an alcohol abuse disorder refuses all attempts at treatment, indicating that although he recognizes the negative consequences of substance use in his life, he is not willing to stop drinking Other codes within this section b1300: Energy level b1301: Motivation b1302: Appetite b1303: Craving b1304: Impulse control b1308: Energy and drive functions, other specified b1309: Energy and drive functions, unspecified
4.5
Measuring Across ICF Categories
4.5.1
Self-reported ICF-based Measurement Instruments
Based on the ICF, WHO has developed the > WHO Disability Assessment Schedule Version II (WHODAS II) (Po¨sl et al., 2007), a generic self-administered questionnaire used in adults >18 years of age which covers the ICF components activity and participation. It includes six domains: understanding and communicating, getting around, self care, getting along with others, household and work activities, and participation in society. It has been developed crossculturally and is applicable across the spectrum of cultural and educational backgrounds. In addition to self-report, an interviewer and proxy version is available. The time to complete the questionnaire for the 12-item version takes approximately 5 min and for the 36-item version about 20 min. The first study applying the WHODAS II in rehabilitation using a German version demonstrated that it is a useful instrument for measuring functioning and disability in patients with musculoskeletal diseases, internal diseases, stroke, breast cancer and depressive disorder (Po¨sl et al., 2007). The results of this study also support the reliability, validity, dimensionality, and responsiveness of the WHODAS II. However, for the domain household and work activities, a clear distinction between work activities versus household activities was apparent in musculoskeletal and internal conditions (Po¨sl et al., 2007). Therefore, one may in the future consider the separate scoring and reporting of these sub-domains. For specific conditions and/or settings one may want to use a specific measurement instrument. A suitable starting point for the development for such measurement instruments are the ICF Core Sets. The ICF Research Branch of the WHO FIC CC Germany at the University of Munich is thus cooperating with and supporting research groups in the process to develop self-reported questionnaires based on the ICF Core Sets (www.icf-research-branch.org). 4.5.2
ICF-based Clinical Measurement Instruments
Clinician’s ratings of the ICF Qualifier (> Table 1-6) across a number of ICF categories, e.g. across the categories of an ICF Core Set, can be reported in the form of a categorical profile. A categorical profile across a valid set of ICF categories such as an ICF Core Set provides an estimation of a persons functioning state. The functioning state is the central information for clinicians when planning and reporting the results of a health
29
1
ICF Qualifier range from 0 = no problem in the components of body functions (b), body structures (s), activity and participation (d) and from 4 = complete barrier to +4 = complete facillitator in the environmental factors. In personal factors, the sign + and - indicates to what extent a determined personal factor has a positive or negative influence on the individuals functioning. The symbol ü indicates that the treatment goals with their goal values have been achieved and the symbol - that they were not achieved
. Table 1-8 ICF-based assessment and evaluation including goal setting and goal achievement in a patient after Spinal Cord Injury. The functioning states at the start of rehabilitation and after 4 weeks are shown as categorical profiles based on expert ratings of the ICF Qualifier
30 The International Classification of Functioning, Disability and Health
The International Classification of Functioning, Disability and Health
1
care intervention. > Table 1-8 shows the example of functioning states at the start and the end of a rehabilitation program. The aggregation of information obtained from a categorical profile using the Rasch model results in a summary score (Grill and Stucki, 2008; Cieza and Stucki, 2008a). In the case of aggregation of information across a valid set of categories such an ICF Core Set, the summary score provides an estimation of a persons functioning status. If using an electronic clinical chart, the creation of a score from a categorical profile created based on an ICF Core Set does not require additional work. Functioning status information provides clinicians with an intuitive, overall understanding of a patient’s general level of functioning. It can be used by clinicians, service program providers and payers e.g. for the assignment of patients to suitable health service programs, to monitor and manage persons functioning along the continuum of care and across service program providers, to evaluate service programs, to predict resources and hence costs and to derive payment schemes. The principle of how to develop one- or multi-dimensional clinical measurement instruments based on clinicians ratings of ICF Core Sets has been recently demonstrated (Grill and Stucki, 2008; Cieza and Stucki, 2008a). It could also be demonstrated how to apply such scores across countries by adjusting for differential item function. It is thus possible to compare functioning status information across countries and world regions.
4.5
Conclusion
The ICF has become the universal and unifying framework for functioning, disability and health. As international standards, the ICF Core Sets are the practical tools for the classification and description of patients functioning states. They are the reference for the reporting of measurements made with a wide range of validated measurement instruments and they serve as starting point for the development of clinical measures or ICF Core Indices as well as selfreported instruments.
Summary Points The ICF is a multipurpose classification designed to serve various disciplines and different sectors. The ICF is the international classification for health and health-related states.
The ICF offers a conceptual framework for information that is applicable to personal health care, including prevention, health promotion, and many other fields. Its framework is the bio-psycho-social model of functioning and disability. The ICF achieve a synthesis, thereby providing a coherent view of different perspectives of health. The components of body functions and structures, activities and participation, and environmental factors are classified based on 1,425 ICF categories. The categories are organized within a hierarchically nested structure with up to four different levels and are denoted by unique alphanumeric codes. The WHO provides with the ICF for the first time a universal and internationally accepted framework and classification. It is a promising starting point for the integrative understanding of functioning, disability and health and the overcoming of Cartesian dualism of body and mind as well as both sociological and biomedical reductionism.
31
32
1
The International Classification of Functioning, Disability and Health
Implementation of the ICF includes e.g. its use as a screening tool for referral to rehabilitation services and for the rehabilitation management in rehabilitation facilities. Parallel to the official implementation activities the ICF has found immediate interest in the health sciences, particularly rehabilitation as well as in different scholars worldwide. The objectives of ICF Core Sets are to be a parsimonious and hence practical selection of ICF categories for clinical practice, service provision and research and to link the ICF to the International Classification of Diseases.
Acknowledgments The authors would like to thank Prof. Jerome Bickenbach and Dr. Somnath Chatterij from WHO for the inspiring discussions of the ICF and related concepts and Dr. Thomas Ewert, Gisela Immich and Susanne Stucki for their help in the preparation of the manuscript.
References Allan CM, Campbell WN, Guptill CA, Stephenson FF, Campbell KE. (2006). J Interprofessional Care. 20: 235–245. APA, American Pschological Association. (2007). Procedural Manual and Guide for Standardized Application of the International Classification of Functioning, Disability and Health (ICF), Field Trial Version from http://icf.apa.org Aringer M, Stamm TA, Pisetsky DS, Yarboro CH, Cieza A, Smolen JS, Stucki G. (2006). Lupus. 15: 248–253. ¨ stu¨n TB. (1999). Bickenbach JE, Chatterji S, Badley EM, U Soc Sci Med. 48: 1173–1187. Biering-Sorensen F, Scheuringer M, Baumberger M, Charlifue SW, Post MW, Montero F, Kostanjsek N, Stucki G. (2006). Spinal Cord. 44: 541–546. Boldt C, Brach M, Grill E, Berthou A, Meister K, Scheuringer M, Stucki G. (2005). Disabil Rehabil. 27: 431–436. Brach M, Cieza A, Stucki G, Fu¨ssl M, Cole A, Ellerin BE, Fialka-Moser V, Kostanjsek N, Melvin J. (2004). J Rehabil Med Suppl. 44: 121–127. Brockow T, Cieza A, Kuhlow H, Sigl T, Franke T, Harder M, Stucki G. (2004a). J Rehab Med Suppl. 44: 30–36. Brockow T, Wohlfahrt K, Hillert A, Geyh S, Weigl M, Franke T, Resch KL, Cieza A. (2004b). J Rehab Med Suppl. 44: 49–55. Brockow T, Duddeck K, Geyh S, Schwarzkopf SR, Weigl M, Franke T, Brach M. (2004c). J Rehab Med Suppl. 44: 43–48. Borchers M, Cieza A, Sigl T, Kollerits B, Kostanjsek N, Stucki G. (2005). Clin Rheumatol. 24: 139–144. Coenen M, Cieza A, Stamm TA, Amann E, Kollerits B, Stucki G. (2006). Arthritis Res Ther. 8: R84.
¨ stu¨n TB, Chatterji S, Kostanjsek N, Cieza A, Ewert T, U Stucki G. (2004a). J Rehabil Med. 44: 9–11. Cieza A, Stucki G, Weigl M, Kullmann L, Stoll T, Kamen L, Kostanjsek N, Walsh N. (2004b). J Rehabil Med Suppl. 44: 63–68. Cieza A, Stucki G, Weigl M, Disler P, Ja¨ckel W, van der Linden S, Kostanjsek N, de Bie R. (2004c). J Rehabil Med Suppl. 44: 69–74. Cieza A, Schwarzkopf SR, Sigl T, Stucki G, Melvin J, Stoll T, Woolf A, Kostanjsek N, Walsh N. (2004d). J Rehabil Med Suppl. 44: 81–86. Cieza A, Stucki A, Geyh S, Berteanu M, Quittan M, Simon A, Kostanjsek N, Stucki G, Walsh N. (2004e). J Rehabil Med Suppl. 44: 94–99. Cieza A, Chatterji S, Andersen C, Cantista P, Herceg M, Melvin J, Stucki G, de Bie R. (2004f). J Rehabil Med Suppl. 44: 128–134. Cieza A, Geyh S, Chatterji S, Kostanjsek N, Ustun BT, Stucki G. (2006b). BMC Med Res Methodol. 6: 36. ¨ stu¨n TB, Cieza A, Geyh S, Chatterji S, Kostanjsek N, U Stucki G. (2005b). J Rehabil Med. 37: 212–218. Cieza A, Hilfiker R, Boonen A, Chatterji S, Konstanjsek N, ¨ stu¨n TB, Stucki G. (2008b). Clin J Epidemiol. U in press. ¨ stu¨n Cieza A, Hilfiker R, Chatterji S, Konstanjsek N, U TB, Stucki G. (2008c). Clin J Epidemiol. in revision. Cieza A, Stucki G. (2005a). Qual Life Res. 14: 1225–1237. Cieza A, Stucki G. (2006a). In: Bartlett SJ, Bingham CO, Maricic MJ, Iversen MD, Ruffing V (eds.) Clinical Care in the Rheumatic Diseases. Amer College of Rheumatology, Atlanta, pp. 79–87. Cieza A, Hilfiker R, Boonen A, van der Heijde D, Braun J, Stucki G. (2008). Disabil Rehabil. 12: 1–10. [Epub ahead of print]
The International Classification of Functioning, Disability and Health Cohen J. (1960). Educ Psychol Meas. 20: 37–46. DeJong G. (1993). Arch Phys Med Rehabil. 74: 1017–1024. Dijkers MP, Whiteneck G, El-Jaroudi R. (2000). Arch Phys Med Rehabil. 81: S63–S80. Dixon D, Pollard B, Johnston M. (2007). Pain. 130: 249–253. Dreinho¨fer K, Stucki G, Ewert T, Huber E, Ebenbichler G, Gutenbrunner CH, Kostanjsek N, Cieza A. (2004). J Rehabil Med Suppl. 44: 75–80. Efron B. (1982). The Jackknife, the Bootstrap, and Other Resampling Plans. Society for Industrial and Applied Mathematics, Philadelphia. ¨ stu¨n TB. (2000). WHO Mental Epping-Jordan JA, U Health Bull. 6: 5–6. Ewert T, Grill E, Bartholomeyczik S, Finger M, Mokrusch T, Kostanjsek N, Stucki G. (2005). Disabil Rehabil. 27: 367–373. Ewert T, Fuessl M, Cieza A, Andersen C, Chatterji S, Kostanjsek N, Stucki G. (2004). J Rehab Med Suppl. 44: 22–29. Field ML, Jette AM. (2007). The Future of Disability in America. National Academy. Washington DC. Fougeyrollas P, Noreau L, Bergeron H, Cloutier R, Dion SA, St-Michel G. (1998). Int J Rehabil Res. 21: 127–141. Gray DB, Hendershot GE. (2000). Arch Phys Med Rehabil. 81: S10–S14. Geyh S, Kurt T, Brockow T, Cieza A, Ewert T, Omar Z, Resch KL. (2004a). J Rehab Med Suppl. 44: 56–62. Geyh S, Cieza A, Schouten J, Dickson H, Frommelt P, Omar Z, Kostanjsek N, Ring H, Stucki G. (2004b). J Rehabil Med Suppl. 44: 135–141. Geyh S, Cieza A, Kollerits B, Grimby G, Stucki G. (2007). Qual Life Res. 16: 833–851. Grill E, Ewert T, Chatterji S, Kostanjsek N, Stucki G. (2005a). Disabil Rehabil. 27: 361–366. Grill E, Huber EO, Stucki G, Herceq M, Fialka-Moser V, Quittan M. (2005b). Disabil Rehabil. 27: 447–458. Grill E, Quittan M, Huber OE, Boldt C, Stucki G. (2005c). Disabil Rehabil. 27: 437–445. Grill E, Lipp B, Boldt C, Stucki G, Koenig E. (2005d). Disabil Rehabil. 27: 459–465. Grill E, Stucki G, Boldt C, Joisten S, Swoboda W. (2005e). Disabil Rehabil. 27: 467–473. Grill E, Hermes R, Swoboda W, Uzarewicz C, Kostanjsek N, Stucki G. (2005f). Disabil Rehabil. 27: 411–417. Grill E, Stucki G, Scheuringer M, Melvin J. (2006). Am J Phys Med Rehabil. 85: 640–649. Grill E, Mansmann U, Cieza A, Stucki G. (2007). J Rehabil Med. 39: 71–76. Grill E, Stucki G. (2008). Clin J Epidemiol. Jul 9 [Epub ahead of print] Grimby G, Finnstam J, Jette A. (1988). Scand J Rehabil Med. 20: 93–98.
1
Hetzel AM. (1997). History and Organization of the Vital Statistics System. National Center for Health Statistics, Hyattsville. Imrie R. (2004). Soc Health Illn. 26: 287–305. Jette AM. (2006). Phys Ther. 86: 726–734. Jette AM, Haley SM, Kooyoomjian JT. (2003). J Rehabil Med. 35: 145–149. Johnston M, Pollard B. (2001). Soc Sci Med. 53: 1261–1273. Kaplan A. (1964). The Conduct of Inquiry: Methodology for Behavioral Science. Chandler, San Francisco. Karlsson G. (1993). Psychological Qualitative Research from a Phenomenological Perspective. Almqvist and Wiksell International, Stockholm. Kesselring J, Coenen M, Cieza A, Thompson A, Kostanjsek N, Stucki G. (2008). Mult Scler. 14: 252–254. Kirchberger I, Glaessel A, Stucki G, Cieza A. (2007). Phys Ther. 87: 368–384. Kvale S. (1996). Interviews: An Introduction to Qualitative Research Interviewing. Sage, Thousand Oaks. Lemert EM. (1972). Human Deviance, Social Problems, and Social Control. Prentice-Hall, Englewood Cliffs. Mayo NE, Poissant L, Ahmed S, Finch L, Higgins J, Salbach NM, Soicher J, Jaglal S. (2004). Am Med Inform Assoc. 11: 514–522. Mueller M, Boldt C, Grill E, Strobl R, Stucki G. (2008). BMC Nurs. 7: 3. Nagi SZ. (1965). Some Conceptual Issues in Disability and Rehabilitation. American Sociological Association, Washington DC. Pope AMP, Tarlov AR. (1991). Disability in America: Toward a National Agenda for Prevention. National Academy, Washington DC. Po¨sl M, Cieza A, Stucki G. (2007). Qual Life Res. 16: 1521–1531. Reed GM, Lux JB, Bufka LF, Trask C, Peterson DB, Stark S, Threats TT, Jacobson JW, Hawley JA. (2005). Rehabil Psychol. 50: 122–131. Rentsch H, Bucher P, Dommen N, Wolf C, Hefti H, Fluri E, Wenger U, Wa¨lti C, Boyer I. (2003). Disabil Rehabil. 25: 411–421. Ruof J, Cieza A, Wolff B, Angst F, Ergeletzis D, Omar Z, Kostanjsek N, Stucki G. (2004). J Rehabil Med Suppl. 44: 100–106. Scheuringer M, Stucki G, Huber EO, Brach M, Schwarzkopf SR, Kostanjsek N, Stoll T. (2005a). Disabil Rehabil. 27: 405–410. Scheuringer M, Grill E, Boldt C, Muellner P, Mittrach R, Stucki G. (2005b). Disabil Rehabil. 27: 419–429. Sigl T, Cieza A, Brockow T, Chatterji S, Kostanjsek N, Stucki G. (2006). J Clin Pain. 22: 147–153. Sigl T, Cieza A, van der Heijde D, Stucki G. (2005). Ann Rheum Dis. 64: 1576–1618.
33
34
1
The International Classification of Functioning, Disability and Health
Stamm T, Cieza A, Machold KP, Smolen JS, Stucki G. (2004). Arthritis Rheum. 51: 917–924. Stamm T, Cieza A, Coenen M, Machold KP, Nell VP, Smolen JS, Stucki G. (2005). Arthritis Rheum. 53: 431–439. Stamm T, Cieza A, Machold K, Smolen JS, Stucki G. (2006a). Austr Occup Ther J. 53: 9–17. Stamm T, Geyh S, Cieza A, Machold K, Kollerits B, Kloppenburg M, Smolen J, Stucki G. (2006b). Rheumatology. 45: 1534–1541. Stamm TA, Nell V, Mathis M, Coenen M, Aletaha D, Cieza A, Stucki G, Taylor W, Smolen JS, Machold KP. (2007). Arthritis Rheum. 57: 487–494. Steiner WA, Ryser L, Huber E, Uebelhart D, Aeschlimann A, Stucki G. (2002). Phys Ther. 82: 1098–1107. Stier-Jarmer M, Grill E, Ewert T, Bartholomeyczik S, Finger M, Mokrusch T, Kostanjsek N, Stucki G. (2005). Disabil Rehabil. 27: 389–395. Stoll T, Brach M, Huber EO, Scheuringer M, Schwarzkopf SR, Kostanjsek N, Stucki G. (2005). Disabil Rehabil. 27: 381–387. Stucki G. (2005a). Am J Phys Med Rehabil. 84: 733–740. Stucki A, Borchers M, Stucki G, Cieza A, Amann E, Ruof J. (2006). Int J Obes. 30: 1791–1799. Stucki G, Cieza A, Geyh S, Battistella L, Lloyd J, Symmons S, Kostanjsek N, Schouten J. (2004a). J Rehabil Med Suppl. 44: 87–93. Stucki A, Daansen P, Fuessl M, Cieza A, Huber E, Atkinson R, Kostanjsek N, Stucki G, Ruof J. (2004b). J Rehabil Med Suppl. 44: 107–113. Stucki A, Stoll T, Cieza A, Weigl M, Giardini A, Wever D, Kostanjsek N, Stucki G. (2004c). J Rehabil Med Suppl. 44: 114–120. Stucki G, Boonen A, Tugwell P, Cieza A, Boers M. (2007b). J Rheumatol. 34: 600–606. Stucki G, Cieza A, Melvin J. (2007a). J Rehabil Med. 39: 279–285.
Stucki G, Ewert T, Cieza A. (2002). Disabil Rehabil. 24: 932–938. Stucki G, Grimby G. (2004). J Rehabil Med. 44: 5–6. ¨ stu¨n TB, Melvin J. (2005b). Disabil Rehabil. Stucki G, U 27: 349–352. Stucki A, Stucki G, Cieza A, Schuurmans M, Kostanjsek N, Ruof J. (2007). Respir Med. 101: 1113–1122. Tschiesner U, Cieza A, Rogers SN, Piccirillo J, Funk G, Stucki G, Berghaus A. (2007). Eur Arch Otorhinolaryngol. 264: 1215–1222. ¨ stu¨n TB, Chatterji S, Kostanjsek N. (2004). J Rehabil U Med. 44: 7–8. Weigl M, Cieza A, Andersen C, Kollerits B, Amann E, Stucki G. (2004). J Rehab Med Suppl. 44: 12–21. Weigl M, Cieza A, Harder M, Geyh S, Amann E, Kostanjsek N, Stucki G. (2003). Osteoarthritis Cartilage. 11: 519–523. WHO (World Health Organization). (2001). International Classification of Functioning, Disability and Health: ICF. WHO, Geneva. WHO (World Health Organization). (2003). ICF Checklist. Version 2.1a, Clinician Form for International Classification of Functioning, Disability and Health from http://www3.who.int/icf/checklist/icf-checklist. pdf Wildner M, Quittan M, Portenier L, Wilke S, Boldt C, Stucki G, Kostanjsek N, Grill E. (2005). Disabil Rehabil. 27: 397–404. Wolff B, Cieza A, Parentin A, Rauch A, Sigl T, Brockow T, Stucki A. (2004). J Rehab Med Suppl. 44: 37–42. van Echteld I, Cieza A, Boonen A, Stucki G, Zochling J, Braun J, van der Heijde D. (2006). J Rheumatol. 33: 2475–2483. Vieta E, Cieza A, Stucki G, Chatterji S, Nieto M, SanchezMoreno J, Jaeger J, Grunze H, Ayuso-Mateos JL. (2007). Bipolar Disord. 9: 16–24.
2 The Keele Assessment of Participation R. Wilkie 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
2
Developing a Conceptual Model of Participation Restriction . . . . . . . . . . . . . . . . . . . . 39
3 3.1 3.2 3.3 3.4
Developing an Instrument and Measurement Model to Measure Participation Restriction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 Item Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 Number of Items . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 Scoring of Each Item . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 Instrument Score . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
4
Psychometric Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
5 5.1 5.2
Pre-Testing Interview Stage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 Cognitive Interviews . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 Qualitative Interviews . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
6
Further Development of Instrument . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
7 7.1 7.1.1 7.1.2 7.1.3
Pilot Questionnaire Stage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 Pilot Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 Performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 Validity of the Four Filter Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 Discriminant Validity (Conceptual Discrimination – Does Frequency Matter?) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 Repeatability of the KAP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 Validity of the KAP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 Convergent Validity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 Discriminant Validity (Item Discrimination) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 Pilot Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 Validity of the Four Filter Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 Discriminant Validity (Conceptual Discrimination) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 Repeatability of the KAP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 Convergent and Discriminant Validity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
7.1.4 7.1.5 7.1.6 7.1.7 7.2 7.2.1 7.2.2 7.2.3 7.2.4
#
Springer Science+Business Media LLC 2010 (USA)
36
2
The Keele Assessment of Participation
8 8.1
Overview of Psychometric Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 Discriminant Validity (Item Discrimination) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
9
Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
The Keele Assessment of Participation
2
Abstract: The World Health Organization has proposed > participation restriction to reflect the societal consequences of health conditions. Participation restriction can be multiply determined and represents the outcome of an individual’s ongoing interaction with their environment and with society; it emphasizes the independent causal role of environmental factors in determining functioning and refers to what is actually performed by the individual in real life situations. It is appealing concept for those interested in older populations and the impact of chronic diseases because even when health conditions and activity limitations persist, there may still be potential to maintain participation. Despite its importance, participation restriction is inconsistently represented or absent from the content of many health status instruments and has not been clearly measured in population studies. This chapter describes the development of the Keele Assessment of Participation (KAP), a self-complete instrument designed to provide estimates of > person-perceived participation restriction in population-based surveys. It specifies a > conceptual model of participation restriction (based on the individual’s perception of > performance of tasks), outlines how the instrument was developed (with the aim of being short and simple for application in epidemiological studies) and how its scale scores were decided (to allow differentiation between large groups in the population). It also describes, and reports the results of the > pre-pilot studies and pilot studies designed to investigate its measurement properties. The KAP performed adequately in > validity and > reliability tests and can be considered as an instrument that is likely to detect and provide sensible estimates of participation restriction in postal surveys. List of Abbreviations: 95% CI, 95% confidence intervals; ICF, international classification of functioning, disability and health; IPA, impact on participation and autonomy; k, > kappa; kw , weighted kappa; KAP, keele assessment of participation; n, number; Norstop, North Staffordshire osteoarthritis project; RNL, reintegration to normal living; WHO, World Health Organization
1
Introduction
The World Health Organization (WHO) have endorsed the International Classification of Functioning, Disability and Health (ICF) (> Figure 2‐1) as a framework to describe the health state of a person at a particular point in time (World Health Organization, 2001). The framework provides a basis for understanding and studying health-related functioning and the consequences of health conditions. It incorporates the > biopsychosocial model, a synthesis of the medical and social approaches, and emphasizes the multi-dimensional nature of health-related functioning and consequences through interplay of the body, the person and broader social and environmental factors. Focusing on the consequences of health conditions will enhance the understanding of their impact in the general population and of their health and policy implications (Ebrahim, 1997). The ICF has three levels of negative consequences of health conditions: anatomical/ physiological level – impairment (e.g., pain, high blood pressure); individual level – activity limitation (e.g., difficulty picking up objects); societal level – participation restriction (e.g., difficulty getting around or with work) (World Health Organization, 2001). Each level is distinct and the WHO advocates data to be collected independently for each level to explore the associations and causal links between them. Participation restriction is an appealing concept for those interested in older populations and the impact of chronic diseases because
37
38
2
The Keele Assessment of Participation
. Figure 2‐1 The International Classification of Functioning, Disability and Health International Classification of Functioning, Disability and Health (WHO, 2001). The ICF framework organizes information into two parts: (1) functioning and disability and (2) contextual factors. Functioning and disability serve as umbrella terms for normal and abnormal functioning at body, individual and societal levels. Functioning refers to body structure and function, activities and participation. Disability refers to impairment, activity limitation and participation restriction. Environmental and personal factors may influence all three levels of functioning
even when health conditions and activity limitations persist, there may still be the potential to maintain participation (Harwood et al., 1998). For example, people with knee pain and difficulty standing for long periods continue to work and socialize with friends. There is a growing interest and acknowledgement of the influence of environmental factors on functioning, which may explain the gap between individual capacity and performance in activities of daily life (Grimby and Smedby, 2001; Verbrugge, 1994) (e.g., people may be able to walk on a treadmill, but find walking in a town centre difficult due to uneven pavements and obstacles). Social, economic and political factors are increasingly recognized as contributors to health (Simeonsson et al., 2000). The social context shapes the impact of a health condition on a person’s life and may be more concerning to individuals than impairments and activity limitations. For example, being unable to maintain relationships with friends may be an equal source of suffering as the experience of pain itself. Population-based studies have tended to focus on impairment and activity limitation and we know least about participation restriction, which cannot be inferred from these (World Health Organization, 2001). Information on participation restriction is highly fragmented and has been measured inconsistently with researchers using different terminology and different conceptual models. There are no published instruments based exclusively on the concept of participation restriction despite it being argued as a principal component of health-related quality of life (Fransen et al., 2002; Wilkie et al., 2004). Most health measurement instruments do not separate participation restriction from activity limitation (Chapireau and Colvez, 1998; Weigl et al., 2003). The latter measures only an individual’s capacity to execute a task regardless of the specific context of their situation and social surroundings. To exclusively measure the occurrence of participation restriction in population samples a new instrument was required.
The Keele Assessment of Participation
2
This chapter consists of two sections. The first describes the development of the Keele Assessment of Participation (KAP). This instrument was developed by the author and colleagues (Dr George Peat, Dr Elaine Thomas, Dr Helen Hooper, Professor Peter Croft) as part of the North Staffordshire Osteoarthritis Project (Norstop), a general investigation of the health of older people, which is based in the Arthritis Research campaign National Primary Care Centre, Keele University, United Kingdom. One of the key objectives of Norstop was to enhance the epidemiology of the consequences of health conditions in community-dwelling adults. There was a perceived need to investigate participation restriction to provide a fuller account of the impact of health conditions in the community. The KAP was developed specifically for population based postal surveys and prior to the Norstop survey, the challenges for measurement were how to develop the concept, operationalize and quantify participation restriction; the KAP was developed as a self-complete instrument to provide estimates of person-perceived participation restriction in population-based surveys. The approach may miss some specific details and distinctiveness of participation restriction in individuals, since it cannot be specific for all and may not capture relevant information for clinical or rehabilitation purposes. This section includes
Development of a conceptual model of participation restriction Instrument construction Proposal of a > measurement model The second section describes the > psychometric testing of the instrument for its ability to measure participation restriction in population studies. This chapter draws heavily on a shorter account which is published in peer review journal (Wilkie et al., 2005) and on a PhD (Wilkie R, A study of participation restriction and joint pain in community-dwelling older adults, Keele University, 2006).
2
Developing a Conceptual Model of Participation Restriction
The WHO defines participation restriction as ‘‘problems an individual may experience in involvement in life situations’’ (World Health Organization, 2001). This description is broad and is intended to provide a basis for tailoring and developing to meet specific uses (Stucki et al., 2002). Hence, the first stage in developing an instrument to measure participation restriction was to propose a conceptual model. The following propositions were made to direct instrument development: i. Participation restriction is a performance-based, context-dependent, phenomenon Performance is about what people do. It is therefore about functioning in an individual’s real-life environment: the ‘‘lived experience’’ that will be influenced by facilitators and barriers in the environment (World Health Organization, 2001). Indeed at one point in their guidelines the WHO declares performance as the only qualifier that they regard as appropriate for measuring participation restriction. In characterizing participation restriction by performance, instrument scores can be interpreted as restrictions in the performance of everyday tasks that involve interaction with the environment and society. For example, taking children to play in the park or working as a shop assistant. ii. Participation restriction is most meaningfully understood as a person-perceived phenomenon requiring the judgment of the individual
39
40
2
The Keele Assessment of Participation
The WHO defines participation restriction from a > societal perspective if there is difference between observed functioning and what would be expected of an individual who did not have the same or similar health condition, or would be regarded as normal functioning for someone of that age, sex and culture (World Health Organization, 2001). Although useful for epidemiological studies in which application of a defined standard can allow comparison between different groups, studies of handicap found it difficult to define what was ‘‘normal’’ because of the variability of individual roles and needs (Carr, 1999; Harwood et al., 1998). The same problem arises for participation restriction. Comparing individuals with a reference standard of what is proposed as normal for society or study population, may be based on an erroneous assumption that participation restriction occurs when individuals deviate from ‘‘norms.’’ Instead it is better to view normality as an individually defined concept. An alternative approach to measuring participation restriction is to allow individuals themselves to judge whether or not their participation is maintained or falls below what they expect. The experience of participation restriction is specific to each person due to the variability of roles and influencing factors (Carr, 1996; Ueda and Okawa, 2003). That judgment may still reflect perceived societal norms but is also likely to depend on the different roles that individuals fulfill, their standards, needs, aspirations, and expectations. Environmental and personal factors would still be expected to influence participation restriction measured from this perspective. iii. Participation restriction is not a fixed state but may be experienced intermittently within a given time period Participation restriction is a dynamic concept. The nature of ongoing interaction of the individual with the physical, social and psychological environment will not be static but is likely to change with time and setting. Factors particular to the individual, and facilitators and barriers to participation will vary. For example, the symptoms of joint pain and the presence of someone to assist participation may change with time and setting. The key dimensions of participation restriction are how and when performance in life situations occurs. iv. There are an infinite number of life situations that can be organized under the WHO domains for the measurement of participation restriction; domains are conceptually mutually exclusive The WHO offers nine domains as a method of organizing the infinite number of life situations where participation restriction can occur. Each domain is a practical, meaningful set of actions, tasks and life situations. They are conceptually different and mutually exclusive, suggesting that each form of participation restriction may be determined by a unique group of factors. The absence of a single concept of participation restriction is implicit in the four alternative methods for allocating domains and sub-domains to either participation or activity limitation in the ICF (World Health Organization, 2001). Criteria (i.e., does the functional task involve interaction with factors external to the individual (e.g., type of environment, the assistance of others, use of aids) or a non-standardized environment (e.g., public park) were applied to each item in the nine domains to identify which domains measured participation restriction. Six domains were identified as relating to participation restriction; namely mobility, self-care, domestic life, interpersonal interaction, major life and community, social and civic life (> Table 2‐1). The nature and extent of participation restriction will be considered at domain level, which requires responders to make an ‘‘internal calculation’’ of performance in a number of functional activities of a particular domain.
The Keele Assessment of Participation
2
. Table 2-1 A priori designation of ICF domains and subdomains to activity limitation or participation restriction ICF Domain
Activity limitation
Participation restriction
1. Learning and applying knowledge
●
2. General tasks and demands
●
3. Communication
●
4. Mobility
●a
●b
5. Self-care
●
●d
c
6. Domestic life
●
7. Interpersonal interaction
●
8. Major life
●
9. Community, social, and civic life
●
Each task within the nine domains were reviewed for whether they captured activity limitation or participation restriction. In domains 1 to 3 all tasks were deemed to measure activity limitation. In domains 4 and 5 some items were deemed to capture activity limitation and others captured participation restriction. In domains 6 to 9 all tasks were deemed to measure participation restriction a Mobility subdomains: Changing and maintaining body position (4.10–4.19); Walking (4.50) b Mobility subdomains: Transferring oneself (4.20); Carrying, moving and handling objects (4.30–4.49); Moving around (4.55–4.69); Moving around using equipment (4.70–4.89) c Self-care subdomains: Eating (5.50); Drinking (5.60) d Self-care subdomains: Washing oneself (5.10); Caring for body parts (5.20); Toileting (5.30); Dressing (5.40); Looking after one’s health (5.70)
v. Participation restriction across multiple domains is summative with restriction in more domains indicating greater participation restriction. Prior to any empirical testing, each item is considered to be equal. Increasing numbers of domains where participation restriction occurs indicates more aspects of life in which participation restriction is experienced.
3
Developing an Instrument and Measurement Model to Measure Participation Restriction
The instrument and measurement model were designed to be consistent with the specified conceptual model of participation restriction (i.e., performance in life situations as perceived by the individual). Where possible, empirical evidence was used to inform choices in the development of the questionnaire and measurement model. Items were designed to be short, specific and consistent to enhance response and completion rates (McColl et al., 2001). Questionnaire development drew on previously published recommendations on item wording and response formats (McColl et al., 2001; Moser and Kalton, 1971; Sudman and Bradburn, 1974). The measurement model outlining issues of scaling and scoring of this measure was developed with particular attention to the needs of an epidemiological study of participation restriction in population studies.
41
42
2 3.1
The Keele Assessment of Participation
Item Development
The phrase ‘‘I have’’ was used to measure performance. The phrase ‘‘I wanted’’ was used to indicate individual perception and judgment, and the phrase ‘‘as and when I wanted’’ was used to denote how and when participation restriction tasks were performed. The term ‘‘restriction’’ was avoided to maintain the neutral language that characterizes the ICF. The items contained a recall period set at 4 weeks, which was considered to be a convenient time unit for participants to refer to, with the aim of preventing recall bias due to memory failure and ‘‘telescoping’’ (Abramson and Abramson, 1999). Each item was ‘‘closed’’ and invited responses on an adjective ordinal scale. A nonoverlapping five point adjectival ordinal scale was devised, to capture the dynamic nature of participation restriction and comprehensively presented proportions of time. Adjectives for response categories were chosen to facilitate consistent interpretation between responders (e.g., All the time, Most of the time ….. None of the time).
3.2
Number of Items
Items were designed to measure domains in a way that requires responders to make an aggregate judgment on the basis of the life situations within the domain that are relevant and meaningful to them. Domain titles were reviewed for their clarity and whether they represented a collection of similar functions for responders to interpret and relate to when responding. Three domains (mobility, domestic life and major life) were broken down to present a more interpretable set of tasks. This resulted in a total of 11 items (> Table 2‐2). The items were presented in the measurement instrument in the same order as the ICF domains.
3.3
Scoring of Each Item
Individual item scoring was based on a pragmatic assumption of when participation restriction occurs. The adjectival ordinal scale of each item is dichotomized to indicate participation and participation restriction (i.e., 0-0-1-1-1) (> Figure 2‐2). The scale was dichotomized on the basis of a plausible boundary between participation and participation restriction. Participation that occurs ‘‘as and when you want’’ only some of the time or less is regarded as restricted.
3.4
Instrument Score
A simple conceptual and arithmetical approach was taken to scoring which makes no assumptions about individual items in terms of their importance, other than that they are all equally contributing to the score (Streiner and Norman, 2003). The instrument score is a simple count of the number of items of restriction indicated by each responder, up to a total of eleven. Increasing numbers of items with participation restriction indicate more aspects of life where individuals perceive their performance to be restricted. This approach does not consider the potential weighting of the total score towards domains with more than one item (e.g., there are three items for domestic life and one for self-care).
The Keele Assessment of Participation
2
. Table 2-2 KAP items measuring participation restriction in ICF domains ICF domains a
KAP items
4. Mobility 4.1–4.29 Changing and maintaining body parts 4.3–4.49 Carrying, moving and handling body parts 4.5–4.69 Walking and moving 4.7–4.99 Moving around using transport
Mobility within the home (Item 1) Mobility outside the home (Item 2)
5. Self-Care 5.1–5.29 Washing, caring for body parts 5.3–5.69 Toileting, Dressing 5.7–5.99 Looking after one’s health
Self-Care (Item 3)
6. Domestic Lifea 6.1–6.29 Acquisition of resources 6.3–6.49 Household tasks 6.5–6.99 Caring for household objects and others
Looking after the home (Item 4) Looking after things (Item 5) Looking after dependants (Item 6)
7. Interpersonal interaction 7.1–7.29 General interpersonal interaction 7.3–7.99 Personal interpersonal relationships
Interpersonal interaction (Item 7)
8. Major Life 8.6–8.79 Economic life 8.4–8.59 Work and employment 8.1–8.39 Education 9. Community, Social and Civic life 9.1–9.19 Community life 9.2–9.29 Recreation and leisure 9.3–9.99 Religion and spirituality
Managing money (Item 8) Work (Item 9) Education (Item 10) Social Life (Item 11)
a
For mobility and domestic life aspects of the main sub-domains feature in the two and three items, respectively, designed to measure these domains. KAP items were designed to measure domains in a way that requires responders to make an aggregate judgment on the basis of the life situations within the domain that are relevant and meaningful to them. Domain titles were reviewed for their clarity and whether they represented a collection of similar functions for responders to interpret and relate to when responding. Three domains (mobility, domestic life and major life) were broken down to present a more interpretable set of tasks. This resulted in a total of 11 items
Estimates of prevalence can be produced for:
Participation restriction in each of the 11 different aspects of life (i.e., the proportion of responders in each aspect of life who do not participate ‘‘as and when they want,’’ ‘‘some of the time or less.’’ Participation restriction in at least one item (i.e., the proportion of responders who indicate that they perceive their performance in at least one item/aspect of life to be restricted ‘‘as and when they want’’ at least ‘‘some of the time’’). Participation restriction in multiple aspects of life (1–3, 4–6, 7–11 areas) could also be calculated – i.e., a simple count of the number of items in which restriction occurs.
43
44
2
The Keele Assessment of Participation
. Figure 2‐2 Scale scoring and dichotomy for KAP items. Individual item scoring was based on a pragmatic assumption of when participation restriction occurs. The adjectival ordinal scale of each item is dichotomized to indicate participation and participation restriction (i.e., 0-0-1-1-1). The scale was dichotomized on the basis of a plausible boundary between participation and participation restriction. Participation that occurs ‘‘as and when you want’’ only some of the time or less is regarded as restricted
4
Psychometric Testing
To examine the ability of the Keele Assessment of Participation to measure person-perceived participation restriction in population studies, and to inform the interpretation of scores, the instrument was examined against the relevant review criteria proposed by the Scientific Advisory Committee of the Medical Outcomes Trust (Lohr et al., 1996) in pre-pilot and pilot studies. A number of methods were included to specifically examine each attribute (> Figure 2‐3).
5
Pre-Testing Interview Stage
The objectives of the cognitive and qualitative interviews were to examine face > content validity and > responder burden.
5.1
> validity,
Cognitive Interviews
Two separate purposive convenience samples of individuals aged 50 years and over were selected for > cognitive interviews to represent older people with a range of participation restriction. One sample consisted of patients receiving treatment for a range of conditions on a rheumatology ward (n = 8) and the other sample consisted of healthy volunteers with no joint or current health problems (n = 3). Participants completed the draft instrument and were observed for any difficulties encountered during this process, which was timed. Semi-structured interviews asked questions related to face validity, responder burden and content validity. All participants in the cognitive interviews reported the questions to be easy to understand and complete, and relevant for the assessment of tasks of daily life. The mean time for completion was 3 min (range: 2–4) with no difficulties observed. No additional domains were suggested for inclusion.
The Keele Assessment of Participation
2
. Figure 2‐3 Flowchart of the methods for examining the measurement properties of the Keele Assessment of Participation. Pre-pilot and pilot studies for the main study were used to examine the psychometric properties of the KAP. This figure sets out what was examined at each stage
5.2
Qualitative Interviews
Qualitative interviews were conducted in a separate sample of participants (n = 4), purposively sampled from a rheumatology ward to represent a range of experiences of health conditions and of past or present functional restrictions, to generate narrative accounts of the impact and experience of living with these. The interviewer was provided with the World Health Organization definition of the domains covered by the questionnaire. From these, stem questions for the interview were generated, which allowed the participants to respond openly about their experiences in relation to each domain. Participants were then invited to complete the KAP and were asked if the questionnaire covered all areas of life and if any additional questions should be included. The interviews were tape-recorded and transcribed, and anonymous data analyzed focusing on face and content validity. Face validity was assessed from participants’ views as they completed the questionnaire, and their opinions gathered on the relevance of the KAP to their problems and on whether it did reflect their restrictions. To examine content validity, interview transcriptions were analyzed for descriptions of functional restrictions. These were then compared to the answers provided by the same participant in filling out their KAP questionnaire. The participants found the items to be relevant, allowed them to convey the problems they were having, and were presented in an acceptable way. All functional restrictions discussed in the interviews could be mapped to a domain of the KAP. However the number of functional restrictions discussed in the interviews differed from the number indicated in the KAP: higher at interview for three participants, higher on the questionnaire for the fourth. Reasons for the
45
46
2
The Keele Assessment of Participation
discrepancies were identified by examining the interview transcripts. For example, participant 1 reported being unable to climb steps or stairs, but did not record restriction of mobility within the home on the KAP because she used a stair lift.
6
Further Development of Instrument
In both interview studies, a number of participants expressed that for four domains (looking after dependents, work, education and social activities) they either chose not to participate in them or that the items were not relevant. Filter questions, with responses ‘‘yes’’/‘‘no,’’ were added to these items to indicate if responders had, for example, dependents to look after, or if they chose to participate in work, education or social activities. If responders indicated ‘‘yes,’’ they would be asked to complete the relevant participation item. If they indicated ‘‘no’’ they would be asked to go to the next question and were scored as not having participation restriction for that item. The modified instrument (> Figure 2‐4) with filter questions was included in the ‘‘Pilot questionnaire stage’’ to assess the KAP’s repeatability and validity.
7
Pilot Questionnaire Stage
In addition to investigating how the instrument would perform in a population survey, the objectives of the pilot study were to examine repeatability and > construct validity.
7.1
Pilot Methods
The KAP was included in a survey instrument mailed to a random sample of 1,461 adults aged 50 years and over drawn from the registered population of one general practice belonging to the North Staffordshire Primary Care Research Consortium.
7.1.1
Performance
To measure the performance of the KAP in a population survey, completion rates and missing data were calculated as proportions of the total number of responders. The distributions of responses for each KAP item and the number of restricted items were calculated for all responders.
7.1.2
Validity of the Four Filter Questions
Filter items were added to the instrument to identify people who could not be restricted because they did not participate in that area of life. To assess that this assumption was correct, each filter question was matched with a frequency question. For each filter question the proportion of those responding that they did not choose to participate in that particular area of life and also indicated that the frequency of participation was ‘‘no days’’ in the corresponding frequency question was calculated. This was taken as an indication of those correctly classified as not having participation restriction in each filter question.
The Keele Assessment of Participation
2
. Figure 2‐4 Keele Assessment of Participation. This is the Keele Assessment of Participation (KAP), an 11 item self-complete instrument designed to measure person-perceived participation. The KAP can be considered as an instrument that is likely to detect and provide sensible estimates of participation restriction in postal surveys of a population of adults aged 50 and over
47
48
2 7.1.3
The Keele Assessment of Participation
Discriminant Validity (Conceptual Discrimination – Does Frequency Matter?)
KAP items were developed to capture the individual’s perspective of their participation and this viewpoint may be independent of how often or how many times someone participates in an area of life. Additional items were included in the questionnaire that assessed the same dimensions as the KAP but from a frequency perspective, i.e., ‘‘On how many days in the past month have you been out for a walk’’ with response options ‘‘All days’’ to ‘‘No days.’’ There were 23 pairs of corresponding items. Responses to KAP items were compared with the corresponding frequency items in order to establish whether the KAP measured participation restriction differently to a normative population standard (i.e., from a societal perspective). For each frequency item, restriction was defined as reporting participation less frequently than the modal value for the whole sample. Actual agreement for each corresponding pair (%) was calculated. Responders to the questionnaire, who gave permission for further contact, were randomly designated to two groups and sent further questionnaires, each group addressing a separate issue: (1) the repeatability of the KAP, and (2) validity of the KAP compared to items measuring participation derived from other instruments.
7.1.4
Repeatability of the KAP
Responses to the first and second completion of the KAP were compared for both the main and filter questions. Repeatability was calculated by actual agreement (% of responders with agreement) and agreement beyond chance (kappa (k) for dichotomous variables; linearly weighted kappa [kw] for analysis of the five response options) and described using cut-offs suggested by Landis & Koch (Landis and Koch, 1977). The difference in the prevalence of any participation restriction between the first and second mailings was calculated with 95% confidence intervals (using Confidence Interval Analysis for Windows) to determine if there was a systematic difference in prevalence between the two mailings. Repeatability was assessed for (1) the original five response options for each item, (2) the dichotomized response for each item, and (3) the categories of ‘‘None’’ and ‘‘Any’’ participation restriction.
7.1.5
Validity of the KAP
The Impact on Participation and Autonomy (IPA) (Cardol et al., 2001) and the Reintegration to Normal Living (RNL) (Wood-Dauphinee et al., 1988) were used to examine convergent and discriminant validity when compared to the KAP. The IPA was developed as a self-complete questionnaire focusing on perceived and experienced participation restriction, and has been administered previously as a postal questionnaire. The RNL assesses global function and measures both the patients’ perceptions of their own capabilities and objective indicators of physical, social, and psychologic performance. It was developed as a tool to monitor progress during the rehabilitation process and to predict future outcomes, and has recently been used in a postal survey (Harker et al., 2002). They were chosen as they were known to contain specific items that measure participation restriction. The response scales of the IPA and the RNL were dichotomized a priori to indicate participation restriction for comparison with the KAP. Responses of seven or below on the visual analogue scale of RNL items and ‘‘moderate,’’
The Keele Assessment of Participation
2
‘‘poor’’ or ‘‘very poor’’ on the IPA item scale were considered to measure participation restriction and correspond to the KAP dichotomy.
7.1.6
Convergent Validity
Items from the RNL and IPA were matched to similar items on the KAP, based on domain content. For example, the self-care items of the RNL and the IPA were compared with the selfcare item of the KAP. There were ten pairs of items from the RNL and the KAP and 24 pairs from the IPA and the KAP for analysis. Actual agreement (%) for each corresponding pair between KAP and RNL, and KAP and IPA, was calculated.
7.1.7
Discriminant Validity (Item Discrimination)
Discrimination was investigated by comparing KAP items with RNL and IPA items from noncorresponding domains to establish the ability of KAP items to measure participation restriction specific to one domain. For example, the KAP item for mobility within the home was compared with IPA and RNL items which captured every other domain. This produced 60 pairs of noncorresponding items between KAP and RNL, allowing seven KAP items to be examined, and 160 between KAP and IPA, allowing eight KAP items to be examined. Agreement for noncorresponding pairs was calculated and compared with the levels of agreement for corresponding pairs. If KAP items measured participation restriction specific to one domain, the agreement for the corresponding pairs would be higher than for the non-corresponding pairs.
7.2
Pilot Results
A total of 1,117 completed questionnaires were received (adjusted response rate 71.7%). The mean completion of KAP items was 98.2% (range: 97.0–99.5%). The distribution of participation restriction for individual KAP items ranged from 4% (work) to 18% (mobility outside the home) (> Table 2‐3). Fifty three percent of responders had no restriction at all (> Figure 2‐5).
7.2.1
Validity of the Four Filter Questions
The proportion of responders who indicated ‘‘no’’ in a filter question and also indicated that they did not participate in the corresponding task (or frequency item) ranged from 64% (‘‘Do you have any relatives, or other people who depend on you?’’) to 98% (‘‘Do you choose to take part in education or training?’’) (> Table 2‐4).
7.2.2
Discriminant Validity (Conceptual Discrimination)
There was no clear association between person-perceived participation restriction and the frequency with which people participated in these domains. Mean percentage agreement, where responders indicated participation or participation restriction in both matched items for 23 matched pairs, was 47.6% (range: 24–87%). Three pairs of corresponding items had moderately high observed agreement (mobility within the home (87%), education (79%) and looking after dependents (71%)). The lowest observed agreement was for pairs of items referring to looking after the home (24%).
49
50
2
The Keele Assessment of Participation
. Table 2‐3 Prevalence of participation restriction in each of the 11 areas of life Area of life
Prevalence of restriction (%)
Mobility within the home
9
Mobility outside the home
18
Self-care
6
Looking after the home
11
Looking after belongings
10
Looking after dependants
14
Interpersonal interaction
10
Managing money
12
Work
4
Education
4
Social activities
14
The prevalence of restriction in each domain is described as the proportion of the total population. The prevalence of restriction was highest for mobility outside the home (18%) and lowest for work and education (4% for each)
7.2.3
Repeatability of the KAP
Out of a possible 314, 196 completed repeat questionnaires were received (adjusted response of 62.4%). For the four filter questions, the mean observed agreement was 87.5% (range: 84–92%) and chance-corrected agreement ranged from moderate (‘‘Do you have any relatives, or other people, who depend on you?’’ k = 0.54; 95% CI: 0.41, 0.67) to substantial (‘‘Do you choose to take part in paid or voluntary work?’’ k = 0.70; 95% CI: 0.57, 0.83). The mean observed agreement for the five response options to each participation item was 75.1% (range: 68–83%) and chance-corrected agreement ranged from slight (kw = 0.34; 95% CI: 0.09, 0.59) to moderate (kw = 0.64; 95% CI: 0.54, 0.74) (> Table 2‐5). Better repeatability was seen for the dichotomized response (mean agreement = 90.4%; range: 85.3–94.4%) (> Table 2‐5) with chance-corrected agreement ranging from slight (k = 0.20, 95% CI: 0.04, 0.44) to substantial (k = 0.71; 95% CI: 0.57, 0.85). The actual agreement for the categories of ‘‘None’’ and ‘‘Any’’ restriction was 71.6% and chance corrected agreement gave a kappa of 0.42 (95% CI: 0.27, 0.57). There was a systematic difference in the prevalence of ‘‘Any’’ participation restriction between the first and second administrations with the prevalence for the first mailing higher than for the second (46.7% cf 37.3%; % difference: 9.4; 95% confidence interval: 1.4%, 17.3%).
7.2.4
Convergent and Discriminant Validity
Analysis was performed for 102 responders who completed both KAP and RNL and for a separate 104 responders who completed KAP and IPA.
The Keele Assessment of Participation
2
. Figure 2‐5 Distribution of the number of restricted items. Multi-domain scores are calculated only for responders who have completed all 11 items. The distribution of the number of items where restriction was indicated was examined to determine the ability of the KAP to explain the full range of amounts of participation restriction in the general population. Responders indicated restrictions in a varying number of items. A ceiling effect was observed with over half of the responders indicating that they were not restricted in any items. The majority of those with restrictions indicated 1–3 restrictions, with a small minority indicating restriction in a substantial number of aspects of life. The distribution of the number of restrictions is similar to that expected for a general population sample (i.e., mostly consisting of people who do not suffer from health conditions or functional restrictions)
Mean percentage agreement for the ten pairs of corresponding items (convergent validity) between KAP and RNL was 79.3% (range: 72–84%). Mean percentage agreement for 23 pairs of corresponding items between KAP and IPA was 87.7% (range: 74–97%). Mean agreement for 60 pairs of non-corresponding items (discriminant validity) between the KAP and the RNL was 76.0% (range: 57–89%). The mean agreement for 160 pairs of noncorresponding items between the KAP and the IPA was 82.8% (range: 66–97%).
51
52
2
The Keele Assessment of Participation
. Table 2‐4 The number of responders who gave a negative answer to the filter question and the number and proportion of those responders who indicated that they do not participate in the matched task Total no. stating ‘‘no’’
No. (%) indicating no days of participation in that area
6. have any relatives or other people who depend on you?
726
467a (64)
9. choose to take part in paid or voluntary work
710
552b (78)
10. choose to take part in education
900
878c (98)
11. choose to take part in social activities
331
253d (77)
331
307e (93)
Filter item Do you
To examine the validity of the filter questions the number and proportion of responders who said that they choose not to participate in an area of life and did not participate is in the right hand column. The highest level of agreement occurred for education: 878 of 900 (98%) responders indicated that they did not choose to take part in education and did not participate in work or training activities. The lowest level of agreement was for looking after dependents; 467 of 726 (64%) responders indicated that they did not have any dependants and did not look after others. Key for superscript: Denotes the number of people who indicated in the past month their frequency was no days in the comparable tasks of a looking after others b going out to work c go on an education or training course d go to a club, church or social event e play a sport
8
Overview of Psychometric Testing
Although performed with small samples and with limited analyses, the pre-pilot tests suggested that the KAP had sufficient levels of face and content validity; all items were considered acceptable and relevant and no additional domains had appeared which were not already covered by the KAP. The fact that each KAP item requires an overall judgment on a number of tasks was illuminated by the finding that participants reported some specific functional restrictions without indicating participation restriction in the corresponding item because the restriction was not considered important or severe enough to influence participation. The minimal responder burden and high completion rates support the potential usefulness of KAP for epidemiological research. The frequency distribution of restricted items indicated a wide range of participation restriction in the general population, but also highlighted a ceiling effect with over half of the responders having no restricted items at all. The filter questions were included to establish when responders could not be restricted because they chose not to participate in those items. The a priori assumption for the filter questions (people who choose not to participate in a domain would not take part in such tasks and therefore could not be ‘‘restricted’’) was confirmed on analysis for most responders. There were, however,
The Keele Assessment of Participation
2
. Table 2‐5 Test retest repeatability for the main component of KAP items Prevalence of participation restriction
Dichotomized categories (participation/ participation restriction) Observed agreement No. (%)
Baseline No. (%)
Repeat No. (%)
Mobility within the home
24 (12.3%)
19 (9.7%) 184 (94.4)
Mobility outside the home
39 (20.0%) 37 (19.0%) 177 (90.8)
Self care
Kappa score k (95%CI)
Five response options Observed agreement Kappa score k No. (%) (95%CI)
0.71 (0.57,0.85)
154 (79.0)
0.63 (0.53,0.73)
0.71 (0.57,0.85)
133 (68.2)
0.64 (0.54,0.74)
11 (5.6%)
9 (4.6%) 183 (93.9)
0.37 (0.23,0.51)
160 (82.1)
0.45 (0.33,0.55)
Looking after 21 (11.1%) the home
14 (7.4%) 167 (88.8)
0.34 (0.20,0.48)
136 (72.0)
0.46 (0.35,0.57)
9 (4.7%) 176 (92.6)
0.26 (0.12,0.40)
148 (77.9)
0.42 (0.30,0.54)
4 (2.1%)
36 (92.3)
0.36 (0.05,0.67)
27 (69.2)
0.34 (0.09,0.59)
Interpersonal interaction
23 (12.0%) 20 (10.5%) 176 (92.2)
0.61 (0.46,0.74)
138 (71.9)
0.51 (0.41,0.61)
Managing money
29 (15.3%) 19 (10.0%) 162 (85.3)
0.34 (0.19,0.47)
149 (78.0)
0.37 (0.25,0.49)
Looking after belongings
11 (5.8%)
Looking after dependants
4 (2.1%)
Work Education Social activities
11 (5.8%)
4 (2.1%)
50 (88.3)
0.20 ( 0.04,0.44)
46 (76.7)
0.39 (0.20, 0.58)
7 (3.7%)
6 (3.2%)
21 (87.5)
0.59 (0.19,0.99)
20 (83.3)
0.56 (0.22, 0.90)
24 (12.8%) 22 (11.7%) 110 (88.7)
0.54 (0.36,0.72)
86 (68.8)
0.52 (0.39, 0.65)
The reliability of each KAP item was examined using the test-retest method of repeatability and summarized using actual agreement and agreement beyond chance (Kappa). It is difficult to interpret the levels of chance-corrected agreement for individual items due to the effects of low prevalence. For example, the mean level of actual agreement for the dichotomous classification of items was 90.4% (range: 85.3–94.4%), and the range of levels of chance-corrected agreement was 0.20 (slight) to 0.71 (substantial (Landis and Koch, 1977)). The prevalence of participation restriction for some items was so low (e.g., 4% for work) that the possible agreement above chance can only be small due to a high expected agreement. In these cases it is difficult even to achieve moderate kappa values. However levels of actual agreement were good. As expected agreement levels were higher for the dichotomy of participation/participation restriction than for individual response options. k kappa; kw linearly weighted kappa; CI confidence interval
some responders who stated they did not have dependents or chose not to take part in work, education or social activities, but who also indicated that they did participate in a related frequency task and could therefore potentially be restricted. This occurred more so for those who did not have dependents but who looked after others on at least a few days. This questions the ability of this
53
54
2
The Keele Assessment of Participation
filter question to include those who are involved in looking after relatives or others as part of a caring role within domestic life and indicates a need for re-phrasing. It also suggests that capturing the ‘‘choice’’ to participate is difficult, and highlights the difficulty in capturing functioning in aspects of life such as work. Participation restriction may be under-estimated as a result, however the size of any error is likely to be modest, and most responders were consistent (at least 64% for each item). The study of discriminant validity confirmed that the individuals’ perception does not simply reflect the frequency of participation restriction. There was a relatively high observed agreement within some of the matched pairs, suggesting that participation frequency may be more influential on the perception of participation restriction for some tasks than others. The analysis did not consider the effects of age and gender on the matched pairs, where frequency may have a greater bearing on perceived restriction for items for some age groups or for men or women. However, in general the results support the conceptual distinction between a person-perceived participation restriction and a restriction indicated by less frequent participation measured against a normal population standard. The latter would result in a proportion of individuals who perceive no participation restriction, being classed as restricted. Conversely, frequent performance of certain tasks does not mean that individuals will inevitably perceive themselves as not being restricted in that domain. Person-perceived participation restriction as measured by KAP was not completely stable over time. There were 71.6% of responders who consistently indicated participation (‘‘none’’) or participation restriction (‘‘any’’) on the two occasions separated by 4 weeks. The levels of actual agreement for the filter questions and the dichotomous classification of participation restriction were also reasonably high. Other levels of repeatability were influenced by measurement characteristics (low prevalence for some items (e.g., work 4%)) and by systematic differences in the prevalence of participation restriction between the questionnaire mailings both of which will tend to reduce levels of agreement and kappa scores (Hoehler, 2000). The systematic differences suggest that perceived participation restriction can vary over short periods of time. This may be attributable to the instability of the person’s perceived participation restriction and the possibility that their perception at one time point may be influenced by that at another. In this study, reassessment of perceived participation may have led to a more optimistic conclusion and resulted in a reduced prevalence of participation restriction. It may also be that more of the group as a whole were experiencing unusually high restriction on the first occasion, but regressed to their normal level by the second time point. The use of early responders who may have returned their questionnaires quickly perhaps due to the perceived relevance of the questionnaire, may have given rise to this. The KAP had high levels of agreement with items from the Reintegration to Normal Living Index and the Impact of Participation and Autonomy, two instruments that consist of a high proportion of items which measure person-perceived participation restriction and can be considered to measure a similar concept. This suggests that person-perceived participation restriction is being measured by KAP items. The levels of agreement were higher than the those reported by McDowell & Newell (McDowell and Newell, 1996) who reviewed a large number of health outcome measures and found that tests of convergent validity generally demonstrate low levels of agreement typically falling between 0.40 and 0.60. The interview content demonstrated that participants considered expectations, aspirations, needs and contextual factors, such as the presence of
The Keele Assessment of Participation
2
carers and the availability of mobility aids, when rating perceived participation. This is consistent with the conceptual model for the KAP.
8.1
Discriminant Validity (Item Discrimination)
Evidence of responders discriminating between aspects of life was obtained during the cognitive and qualitative interviews. However these abstract distinctions were not demonstrated empirically by the study of item discrimination, since agreement between corresponding items on the different scales was little different to that between non-corresponding items. This suggests that participation restriction may not occur specifically in individual aspects of life, but may overlap into different aspects of life with patterns of co-occurrence. Although the hypothesis of discordance was flawed, the study suggests that the same characteristics that influence participation restriction in one aspect of life also influence participation restriction in others. This apparent lack of discrimination should not affect the use of the KAP or the inclusion in it of questions about different aspects of life.
9
Conclusions
The KAP was developed to specifically measure participation restriction in population surveys. The rationale has been presented for measuring participation restriction from the perspective of the individual and with reference to the performance of tasks. The instrument is intended to measure participation restriction comprehensively and consists of eleven items, each representing a different aspect of life. The measurement model of the KAP allows prevalence estimates of participation restriction in any, multiple, and each different aspect of life. It was designed to be a short and simple questionnaire for application in epidemiological studies and to describe, and discriminate between large groups in the population. This approach may miss some specific details and distinctiveness of participation restriction in individuals, since it cannot be specific for all and may not capture relevant information for clinical or rehabilitation purposes. The pilot studies have tested a number of attributes that are traditionally considered when reviewing the quality of health outcome instruments. The KAP has performed adequately in validity and reliability tests and can be considered as an instrument that is likely to detect and provide sensible estimates of participation restriction in postal surveys of a population of adults aged 50 and over.
Summary Points Participation restriction is an appealing concept for those interested in older populations and the impact of chronic diseases because even when health conditions and activity limitations persist, there may still be the potential to maintain participation. Population-based studies have tended to focus on impairment and activity limitation and we know least about participation restriction, which cannot be inferred from these.
55
56
2
The Keele Assessment of Participation
Participation restriction is a performance-based, context-dependent, phenomenon, most meaningfully understood when perceived by individuals, which requires their own judgment.
The Keele Assessment of Participation was developed as a self-complete instrument to
provide estimates of person-perceived participation restriction in population-based surveys. It consists of 11 items that comprehensively capture participation as proposed by the World Health organization in the International Classification of Functioning. The Keele Assessment of Participation performed adequately in validity and reliability tests and can be considered as an instrument that is likely to detect and provide sensible estimates of participation restriction in postal surveys of a population of adults aged 50 and over.
Acknowledgements The study in which this work was undertaken was supported financially by a Program Grant awarded by the Medical Research Council, UK (grant code: G9900220) and by funding secured from the North Staffordshire Primary Care R&D Consortium for NHS service support costs. The author would like to thank Dr George Peat, Dr Elaine Thomas, Dr Helen Hooper and Professor Peter Croft for their input in developing the KAP. Also the administrative and health informatics staff at Keele University’s Primary Care Sciences Research Centre and the doctors, staff and patients of the participating general practice and the rheumatology wards of the Haywood Hospital, North Staffordshire.
References Abramson ZH, Abramson JH. (1999). Survey Methods in Community Medicine: Epidemiological Research, Programme Evaluation, Clinical Trials. Churchill Livingstone, Edinburgh. Cardol M, de Haan RJ, de Jong BA, van den Bos GAM, de Groot IJM. (2001). Arch Phys Med Rehabil. 82: 210–216. Carr AJ. (1996). Br J Rheumatol. 35: 921–932. Carr AJ. (1999). Osteoarthritis Cartilage. 7: 230–238. Chapireau F, Colvez A. (1998). Soc Sci Med. 47: 59–66. Ebrahim S. (1997). J Epidemiol Commun Health. 51: 469–471. Fransen J, Uebelhart D, Stucki G, Langenegger T, Seitz M, Michel BA. (2002). Ann Rheum Dis. 61: 225–231. Grimby G. Smedby B. (2001). J Rehabil Med. 33: 193–194. Harker WF, Dawson DR, Boschen KA, Stuss DT. (2002). Int J Rehabil Res. 25: 93–102. Harwood RH, Prince M, Mann A, Ebrahim S. (1998). Int J Epidemiol. 27: 261–268. Hoehler FK. (2000). J Clin Epidemiol. 53: 499–503.
Landis JR, Koch GG. (1977). Biometrics. 33: 159–174. Lohr KN, Aaronson NK, Alonso J, Burnam MA, Patrick DL, Perrin EB, Roberts JS. (1996). Clin Ther. 18: 979–992. McColl E, Jacoby A, Thomas L, Soutter J, Banford C, Steen N, Thomas R, Harvey E, Garratt A, Bond J. (2001). Health Technol Assess. 5(31): 43–101. McDowell I, Newell C. (1996). Measuring Health. A Guide to Rating Scales and Questionnaires, 2nd ed. Oxford University Press, Oxford. Moser CA, Kalton G. (1971). Survey Methods in Social Investigation, 2nd ed. Dartmouth Publishing, Aldershot. Simeonsson RJ, Lollar D, Hollowell J, Adams M. (2000). J Clin Epi. 53: 113–124. Streiner DL, Norman GR. (2003). Health Measurement Scales: A Practical Guide to Their Development and Use, 3rd ed. Oxford University Press, Oxford. Stucki G, Cieza A, Ewert T, Kostanjsek N, Chatterji S, Bedirhan Ustun T. (2002). Disabil Rehabil. 24: 281–282.
The Keele Assessment of Participation Sudman S, Bradburn N. (1974). Response Effects in Surveys: A Review and Synthesis. Aldine, Chicago. Ueda S, Okawa Y. (2003). Disabil Rehabil. 25: 596–601. Verbrugge LM, Jette A. (1994). Soc Sci Med. 38: 1–14. Weigl M, Cieza A, Harder M, Geyh S, Amann E, Kostanjsk N, Stucki G. (2003). Osteoarthritis Cartilage. 11(7): 519–523. Wilkie R, Peat G, Thomas E, Croft PR. (2004). Arthritis Rheum. 51: 755–762.
2
Wilkie R, Peat G, Thomas E, Hooper H, Croft PR. (2005). Qual Life Res. 14: 1889–1899. Wood-Dauphinee SL, Opzoomer MA, Williams JI, Marchand B, Spitzer WO. (1988). Arch Phys Med Rehabil. 69: 583–590. World Health Organization. (2001). International Classification of Functioning, Disability and Health. World Health Organization, Geneva.
57
3 The Global Person Generated Index F. Martin . L. Camfield . D. Ruta 1 1.1 1.2 1.3
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 Overview of Quality of Life Measures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 The Importance of Different Areas of Life . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 Individualized Measures of Quality of Life . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
2 The Patient Generated Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 2.1 Uses of the PGI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 3 3.1 3.2 3.3 3.4 3.5 3.6 3.7
The Global Person Generated Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 Development of the Pilot Instrument . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65 Sampling and Data Collection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65 Quantitative Validation: Content Validity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66 Quantitative Validation: Criterion Validity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66 Quantitative Validation: Construct Validity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 Qualitative Validation: Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 Qualitative Validation: Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
4 4.1 4.2 4.3 4.4
Conclusions Regarding the Usage of PGI and GPGI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 Comparing Scores Between Individuals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 Weighting QOL Scores by Importance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 Advantages of the Measures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70 Disadvantages of the Measures’ Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
#
Springer Science+Business Media LLC 2010 (USA)
60
3
The Global Person Generated Index
Abstract: The entry briefly describes the value of > individualized quality of life measures and the philosophy behind them and outlines the main measures in this field. It then focuses on the > Patient Generated Index (PGI) (Ruta et al., 1994) and explains its purpose and methodology. Three different forms of administration are explored and its use in a variety of healthcare settings is described. Finally, its psychometric properties are summarized. The > Global Person Generated Index (GPGI) (Ruta et al., 2004a) is then introduced, which is an open-ended measure with a simple method of administration, appropriate for use in developing countries. The piloting and administration of the GPGI in three developing country settings is described in detail to illustrate the challenges of measuring quality of life in these contexts. The validation process is described, focusing particularly on the innovative technique of > qualitative validation, which used semi-structured interviews collected on the same occasion to assess the accuracy with which the measure captured the respondent’s world view. Some response errors were identified, which required additional administrator training, and the value of a brief accompanying interview are discussed. The measure’s advantages are summarized, namely that is flexible and person-centered, relatively quick to administer, generates new information, works across cultures, and provides a direct measure of the gap theory of quality of life (Calman, 1984). The entry concludes that the common criticism of individualized measures that respondents are not rating the same dimensions evaporates when we consider that respondents are rating the results of their judgments of what is important to quality of life: these judgments are directly comparable (Parducci, 1995, p. 29). Therefore they provide a valid measure of people’s perceptions of the quality of their life as a whole across time. List of Abbreviations: ADDQOL, Audit of Diabetes Dependent QOL; GPGI, Global Person Generated Index; MacDQOL, individualized measure of the impact of macular degeneration on quality of life; PGI, Patient Generated Index; QOL, quality of life; SEIQOL, > Schedule for Evaluation of Individual Quality of Life; SPSS, Statistical Package for the Social Sciences (quantitative data analysis software); SWLS, Satisfaction with Life Scale; WeD, Wellbeing in Developing Countries ESRC Research Group (see http://www.welldev.org.uk/)
1
Introduction
1.1
Overview of Quality of Life Measures
Most measures of quality of life (QOL) take the form of questionnaires where items are generated by reviews of the literature, consultation with ‘‘experts,’’ and qualitative research and are rated using Likert scales (e.g., the MOS 36-item Short-Form Health Survey (SF-36), Ware and Sherbourne, 1992). These approaches assume that every item is of relevance to every individual, and further that each item is of equal importance to the respondent, as items are rarely weighted. Items are chosen for inclusion in QOL measures based on an analysis of how they contribute statistically to a model of QOL (e.g., the Personal Wellbeing Index (PWI), International Wellbeing Group, 2006; World Health Organization Quality of Life Measure (WHOQOL), Skevington, 1999). This assumes that following statistical procedures will ensure the inclusion of items that are important to everyone. However, the degree to which items are equally important to everyone must be considered.
The Global Person Generated Index
1.2
3
The Importance of Different Areas of Life
The idea that the importance of different aspects of life varies both across and within individual life courses and cultures is highly intuitive and supporting evidence exists. Indeed, the rationale for developing disease, condition, population, or country-specific QOL measures is that different aspects of life are relevant and important and require different item content (e.g., Tovbin et al., 2003). For example, even within a sample of ‘‘healthy’’ adults in the UK, there were differences by age in the areas participants nominated as important to their life: older participants mentioned health more often than younger participants, whilst mentioning money was more common in the younger (Bowling, 1995). Furthermore, not all of the life areas nominated were covered by commonly used QOL scales, e.g., SF-36 does not explicitly cover sleep, financial management, material welfare, sexual functioning, communication, education, independence or religion (ibid). Differences in importance are also seen in relation to health. For example, Weitzenkamp et al. (2000) suggested that participants with spinal cord injuries ranked components of QOL in a different way to non-injured participants, placing greater emphasis on relationships with relatives, learning, creative expression, reading and other sedentary leisure activities. While the study’s conclusions are limited by the fact that the participants with spinal cord injuries were British, while the non-disabled participants were American, this illustrates another important area of difference. Further support comes from Hensel et al. (2002) who found that people with intellectual disabilities rated five out of eight domains of the same measure as significantly more important than non-disabled participants. This would not be expected if one believes in the universality of importance ratings. Here, importance is shown to be variable and must therefore be measured. Differing importance of the different facets has implications for the accuracy of the overall measurement of quality of life. For example, one person may rate their physical health as low, but not important, whereas another person may rate it equally low, but very important, so its absence may have a more negative impact on their quality of life. Importance is therefore a central issue in accurate measurement of quality of life. An area that is poorly rated, but not seen as important may have a different effect on QOL from a poorly rated area that is highly important (see Ferrans and Frisch, 2004). This suggests the importance of some degree of individualization in assessing the true impact of a shortfall in a particular life domain on a person’s quality of life.
1.3
Individualized Measures of Quality of Life
The underlying philosophy of the individualized approach to measuring QOL has been outlined. Some measures take a partial approach to individualization where respondents rate common items, but are able to exclude irrelevant items by scoring some areas as ‘‘not applicable.’’ An extension of this approach allows weighting of the remaining items by importance, e.g., the ‘‘Audit of Diabetes Dependent QOL’’ or ‘‘ADDQOL’’ (Bradley et al., 1999) and the measure of the impact of macular disease on QOL (‘‘MacDQOL,’’ Mitchell and Bradley, 2004). Fully individualized measures of QOL, where the respondent nominates, scores and weights all elements of the rating include the ‘‘Schedule for the Evaluation of Individual Quality of Life’’ or ‘‘SEIQOL’’ (McGee et al., 1991; O’Boyle, 1995) and the ‘‘Patient
61
62
3
The Global Person Generated Index
Generated Index’’ or ‘‘PGI’’ (Ruta et al., 1994). Both instruments ask respondents to nominate personally important areas of life (step one in the PGI), to indicate their current satisfaction with these areas (step two), and to rate their importance (step three). Importance rating is carried out by allocating a limited number of points, which encourages respondents to indicate priorities within the domains. The SEIQOL appears to have never been used without an accompanying interview, partly due to the complexity of its scoring system. However, the PGI has been used in postal and self-report studies (Lindblad et al., 2002; Tully and Cantrill, 2002). Furthermore, unlike the SEIQOL, the PGI was originally designed to focus on the impact of a specific health condition on QOL.
2
The Patient Generated Index
The ‘‘Patient Generated Index’’ or ‘‘PGI’’ has three steps, as shown in > Figure 3-1. In the first step respondents are asked to nominate up to five important areas of their life affected by the specified health condition (e.g., ‘‘your arthritis’’). The second step asks respondents to rate each nominated area on a scale of 0–10 (previous versions used 0–60 or 0–100, but 0–10 was considered more accessible and psychometrically credible, see Cummins and Gullone, 2000).
. Figure 3-1 The Patient Generated Index – the measure (*see Martin et al. (2007) for a review of the different versions of this measure)
The Global Person Generated Index
3
The lowest scale point is anchored with the statement ‘‘the worst you can imagine’’ and highest point is anchored with the statement ‘‘exactly as you would like to be.’’ In some versions, participants are also asked to rate a sixth area, meant to represent ‘‘all other areas of life not already mentioned.’’ In later versions of the PGI both a sixth and seventh area are used to differentiate between ‘‘other health-related areas’’ not already nominated and ‘‘other nonhealth related areas.’’ In the third and final step, respondents are asked to spend points (different versions have a different number of points to spend) to indicate the relative importance of each life area for overall QOL. The scores from step two are multiplied by the step three weights. This weighted score is then multiplied by the proportion of points allocated to each area. This is then summed to generate a single index score, reported as a percentage. When used longitudinally, three formats of administration are possible. In the ‘‘blind’’ format, the areas previously nominated as important in PGI stage one (i.e., at baseline) are not made available to the respondent. In the ‘‘open’’ format, the areas nominated previously are shown to the respondent; in both blind and open formats respondents can add, remove or substitute important areas of life. While changes in areas nominated over time are overt and explicit in the ‘‘open’’ format, these could also be due to forgetting or misremembering previous areas in the ‘‘blind’’ format. Conversely, in the ‘‘closed’’ format the areas nominated previously are rated a second time with no option for change.
2.1
Uses of the PGI
The PGI has been adapted to different contexts and respondents, for example, older adults (Dempster and Donnelly, 2000), new mothers (Symon et al., 2003) and in a community nursing setting (Griffiths et al., 2000). The structure of the measure also evolved as step two changed from 0 to 100 to an 11 point scale, and methods for step three ranged from spending £100 to between 12 and 60 points (Martin et al., 2007). The PGI has also been used across a wide range of clinical conditions, from lower limb amputees (Callaghan and Condie, 2003) to atopic dermatitis (Herd et al., 1997). More recently, the PGI was used to assess ‘‘> response shift’’ (Sprangers and Schwartz, 1999), namely a change in the meaning of one’s evaluation of a construct as a result of a change in one’s internal standards of measurement, a change in one’s values, or a change in one’s definition of the construct (Ahmed et al., 2005). A full review of the measure’s psychometric properties is given by Martin et al. (2007), which shows that whilst validity and reliability for group comparisons appear sound, responsiveness to change is complex. The PGI continues to be used in a variety of contexts, with appropriate adaptation, and most recently, a non-health related version of the measure has been developed to establish whether Calman’s (1984) definition of QOL as the gap between expectations and experience applies outside healthcare.
3
The Global Person Generated Index
The ‘‘Global Person Generated Index’’ or ‘‘GPGI’’ is then an extension of the PGI, which moves the focus from health and healthcare settings to encompass a more holistic understanding of quality of life (Ruta et al., 2004b). It was designed to combine the open-endedness of the SEIQOL (McGee et al., 1991) with a simple administration method appropriate to developing country contexts. This will enable its use as project planning or evaluation tool, although there are some issues to address, as described in the final section. > Figure 3-2 shows
63
3
. Figure 3-2 The Global Person Generated Index (the psychometric properties of this scale continue to be researched. The psychometric properties of the English version of this scale are yet to be investigated)
64 The Global Person Generated Index
The Global Person Generated Index
3
the current English version. The GPGI was piloted and administered in Bangladesh, Thailand, and Ethiopia during the first phase of the exploratory quality of life work conducted by the Wellbeing in Developing Countries (WeD) ESRC Research Group (see Camfield and Ruta, 2007; Jongudomkarn and Camfield, 2006; Ruta et al., 2004b, 2006) and has subsequently been further developed by WeD affiliated researchers working in these countries.
3.1
Development of the Pilot Instrument
WeD fieldworkers in each of the three countries were given a free hand to interpret and translate a culturally relevant version of the GPGI. This included translations of: instructions to nominate important life areas; the wording and construction of Likert statements and scales; and the method of ‘‘spending’’ points. Back and re-translation methods were used. A purposive sample of 36 (17 female, 19 male) respondents was identified by local fieldworkers in seven towns and villages. A qualitative content analysis was undertaken to explore the elicited life areas, and the validity of rating and weighting methods. Although a range of life areas were generated, with similarities and differences observed between countries, economic ‘‘development’’ focused areas were frequently mentioned in all field sites. In Bangladesh, an alternative form of wording for step two was introduced. In Bangla the words ‘‘good’’ (bhalo) and ‘‘bad’’ (kharap) are vague descriptors. Therefore a seven point scale was constructed around the notions of satisfaction (shontushtho) and dissatisfaction (oshontushtho). In Ethiopia it was difficult to find a word equating with ‘‘important’’ in the context of the GPGI. Eventually the word ‘‘wesagn,’’ literally meaning crucial or needed, was used. It was also decided to give respondents ten ten-cent pieces and to ask them to place the coins in the boxes on the questionnaire when ‘‘spending’’ points in step three of the GPGI. In Thailand, three methods were tried for step two: moons (pie-charts, originally depicted as full-empty, later light-dark), smiley faces (happy-sad) and numbers (labeled from good to bad), however, after mixed results it was agreed that any method could be used, if it was sufficiently well explained. In Thailand three methods were also tried for step three: spending Thai Baht, placing stones and making tallies. The coins were the most successful as they could be explained in terms of shopping in the market or ‘‘making merit’’ (donating to a Buddhist shrine). The stones were the least successful as people didn’t know how to value them. Following the pre-pilot analysis, a final questionnaire wording was agreed for the pilot instrument in each country. All countries decided to reduce the original 10 point scale used in step two to a seven point (i.e., a 0–6) scale, and to use the method of spending ten coins to assign importance weights in step three.
3.2
Sampling and Data Collection
The GPGI was administered as part of WeD’s exploratory research into Quality of Life, which took place in rural, peri-urban, and urban sites in Bangladesh, Ethiopia, and Thailand. The fieldwork involved semi-structured interviews, which were used to interpret and validate the GPGI, focus group discussions, and the piloting of other measures, such as the ‘‘Satisfaction with Life Scale’’ (Diener et al., 1985). In Ethiopia, it was carried out by local researchers, the majority of whom had spent at least 1 year attached to the site, which enabled them to build a good rapport with the inhabitants. The researchers received full training in the
65
66
3
The Global Person Generated Index
methods used, and the majority had also participated in the GPGI piloting. As the GPGI was interview-administered, the response rate was 100%, with the exception of Bangladesh where four female respondents (two of whom were illiterate) chose only to respond to the semistructured interviews. 242 GPGIs were administered during the main fieldwork (120 in Ethiopia, 102 in Thailand, and 22 in Bangladesh), using a purposive sample to capture differences in age, socio-economic status and religion, as well as location.
3.3
Quantitative Validation: Content Validity
Content validity, that is the extent to which a measure assesses content relevant to the underlying construct, was assessed in two ways. First, a frequency analysis was undertaken of the areas mentioned in respondents’ GPGI questionnaires in step 1. This analysis was undertaken separately for each of the three countries. Three commonly mentioned areas emerged: 1. Indicators of material well-being, which included income, assets, crops, property, land, livestock, job, debts and agriculture 2. Health, including health of the family and 3. Family or Children, which included children’s education Therefore a second method of assessing content validity involved correlating respondents’ ratings on a 0–6 scale as mentioned in step two of the GPGI, with scores derived from the semi-structured interviews (described in full in Camfield 4 Ruta, 2007). Material well-being as indicated in the semi-structured interview shows a moderate but statistically significant correlation (0.286) with material well-being as indicated in step two of the GPGI. Interestingly the material well-being score from the semi-structured interview shows a higher correlation with family/children (0.395) as indicated in the GPGI. The health score as indicated on the semi-structured interview shows a strong and statistically significant correlation (0.584) with health as measured by the GPGI. The family/children score on the semi-structured interview shows a moderate and significant correlation (0.361) with family/children as indicated in the GPGI. As with the score for material well-being, the family/children score from the semi-structured interviews shows a moderate but significant correlation (0.232) with the material well-being score as indicated on the GPGI.
3.4
Quantitative Validation: Criterion Validity
Assessing criterion validity (i.e., the extent to which a new measure correlates with established measures of the concept under study) is problematic in the absence of a gold standard measure of individual quality of life in developing countries (Cummins, 2007). However in Ethiopia it was possible to correlate GPGI scores with scores on the five items of the Satisfaction with Life Scale (SWLS), an established measure of a related concept. The GPGI shows weak to moderate but statistically significant correlations with four items ranging from 0.202 to 0.351. No correlation is observed between the GPGI and the item ‘‘I would change nothing in my life.’’ The SWLS item that is conceptually closest to the GPGI, ‘‘My life is close to my ideal,’’ shows the strongest correlation.
The Global Person Generated Index
3.5
3
Quantitative Validation: Construct Validity
Construct validity has been defined as ‘‘the extent to which a new measure is related to specified variables in accordance with an established theory or hypothetical construct’’ (Streiner and Norman, 1995). Two of the authors have recently proposed a general theory that both defines individual quality of life and explains its relationship to key causal determinants (Ruta et al., 2006). According to the theory, indicators of material well-being and indicators of health demonstrate a positive linear relationship with individual quality of life until a certain level is reached – described here as the ‘‘basic capability threshold.’’ Beyond this basic capability threshold, further marginal increases in levels of material well-being and health will give rise to rapidly diminishing marginal increases in quality of life. Accordingly the following validity hypotheses were tested: 1. GPGI scores show a positive linear or curvilinear relationship with material well-being and health scores (including health of family members) derived from semi-structured interviews Yes, slight curvilinearity at higher levels 2. Poor respondents in Bangladesh and Southern Thailand have lower quality of life scores than rich respondents Yes (55.7% vs. 65.4%, p < 0.01) 3. Rural respondents from Ethiopia have a lower quality of life than urban respondents Yes (55.9% vs. 65.1%, p < 0.05) Further construct validity tests of the relationship between GPGI Index scores and other respondent characteristics such as education were conducted; for example, there was a significant difference in GPGI scores between respondents who had completed further or higher education and everyone else (67.8% vs. 58%, p < 0.05). SPSS stepwise regression was used with specified independent variables to model the relationship between the GPGI and the variables identified above. This demonstrated that only material wellbeing (as indicated in respondents’ semi structured interviews) and ‘‘country’’ remained in the model. Together they were able to explain over 21% of the variation in respondents’ GPGI scores. Of 40 respondents explicitly mentioning religion as an important area in their GPGI, 34 were from Ethiopia. We were therefore able to test the hypothesis that Ethiopians who mentioned religion in their GPGI (n = 34) had higher quality of life scores than Ethiopians not mentioning religion in their GPGI (n = 82). While those mentioning religion had a slightly higher mean score (61.9% vs. 59.2%), this was not significant.
3.6
Qualitative Validation: Method
The qualitative validation used the data from the semi-structured interviews to explore content validity, aiming to establish whether (1) the area nominated in the GPGI appeared as an area of importance in the semi-structured interview; (2) whether the way the area was discussed in the semi-structured interview suggested that the appropriate number of points had been allocated to indicate its importance; (3) whether the respondent’s satisfaction with this area appeared to be adequately represented by their GPGI score; (4) whether there were areas in the semi-structured interview that appeared to be even more important to the
67
68
3
The Global Person Generated Index
respondent, but were not nominated in the GPGI; and (5) whether the overall picture given by the semi-structured interview of the extent to which the respondent’s reality met their expectations corresponded with the total score given in the GPGI. Any problems with scoring or spending points were also noted (e.g., where the number of points spent failed to add up to ten), although these may have been due to administrator rather than respondent error. For example, some administrators appear to have asked respondents to rank the areas, rather than spend points, as this a method commonly used in participatory research.
3.7
Qualitative Validation: Results
In the Ethiopian sample, a content analysis of 21 responses to the three steps of the GPGI, and comparison with the content of semi-structured interviews, revealed a close correspondence between the two sources for eight respondents (four female and four male, aged from 19 to 76 years). In each case, the life areas nominated in the GPGI were mentioned several times in the corresponding semi-structured interview and few if any additional areas were emphasized within the semi-structured interview that did not appear in the GPGI. There appeared to be a good match between the extent to which respondents felt their reality met their expectations for each GPGI nominated area, as measured in step two of the GPGI, and the content of the semi-structured interview. The relative importance attached to each GPGI nominated life area, as indicated in the points spent in step three of the GPGI, was also consistent with the interview content. A further 16 responses were content analyzed for Thailand, and another 16 for Bangladesh, sampled according to gender, type of site, region (Thailand), and age (Bangladesh). In Thailand only five responses showed close correspondence, although the majority of discrepancies were minor. The picture was slightly better in Bangladesh where nine respondents showed close correspondence. The majority of the Ethiopian sample (13 of 21 respondents) did not show the same degree of consistency between the content of the GPGI and the semi-structured interview. Analysis suggested three broad levels of inconsistency: errors in GPGI completion that appear to arise from a basic lack of comprehension on the part of respondents or interviewers; minor inconsistencies between one GPGI nominated area and the GPGI; and major inconsistencies where discrepancies were identified between two or more life areas, or where the semistructured interview raised questions about the validity of a respondent’s overall quality of life index score as measured by the GPGI. The exercise generated responses consistent with those collected in participatory research in developing countries, particularly studies that focus on people’s conceptions of wellbeing or poverty (e.g., Narayan et al., 2000). The most mentioned area across the three countries in the WeD sample (n = 242) was Health, however, this was partly due to its popularity in Ethiopia which accounted for 62% of the responses. The second was Money, assets, which was also the main priority in Thailand, accounting for 17% of the responses (78% item frequency). Children was the third most mentioned area, and featured in every country’s ‘‘top five,’’ with Bangladesh also prioritizing Children’s education/future. The fourth was Home, which was very important in Thailand (11.5% of respondents), and fairly important in Ethiopia (5.5%). Fifth and sixth most mentioned areas were Employment and Family. Both of these areas were important in Thailand (to 9.8 and 9.3% of respondents respectively), with Employment
The Global Person Generated Index
3
also important in Ethiopia (third priority, 6.9% of respondents) and Family in Bangladesh (second priority, 7.4% of respondents). Where the GPGI is less successful is in capturing areas that are abstract, or personal, and thus difficult to capture in a few words, e.g., ‘‘own boredom and lack of fulfillment.’’ It also fails to represent potentially shameful areas such as debt or mental health problems, and is slightly biased towards the normative, i.e., areas that are universally acknowledged as important, rather than areas that are ‘‘just’’ important to the respondent.
4
Conclusions Regarding the Usage of PGI and GPGI
The conclusions concerning the PGI and GPGI measures are broken down into four areas: the issue of whether individuals’ scores can be compared; the challenges posed by weighting areas of life by their perceived importance; the advantages of the measures; and the disadvantages. It must be made clear at this point that the psychometric properties of the ‘‘Global Patient Generated Index’’ require further investigation and that the English version has not yet been validated.
4.1
Comparing Scores Between Individuals
The most common criticism of individualized scores is that inter-individual comparisons are impossible if respondents are not all rating the same items. Parducci responds to this in his discussion of the different nature of pleasant experiences where he acknowledges that while a cold drink on a hot day is dissimilar to a reciprocated love, both are the result of a ‘‘pleasantness judgment’’ and therefore comparable to some extent (Parducci, 1995, p. 29). In the same way, if one respondent to the PGI chooses to evaluate ‘‘my health,’’ ‘‘having a high income’’ and ‘‘being fashionable,’’ they are making a judgment about the state of the areas that are important to their QOL, which is comparable to another respondent’s evaluations. This is true even if the content is as different as ‘‘coping with their feelings of grief,’’ ‘‘having a nice home’’ and ‘‘staying in touch with their friends’’ as the overall evaluation is of the state of the areas that are important to QOL. Whether the lack of a comparison of ratings on the same domains is a problem depends upon the purpose of the QOL measure. If the ratings of cancer patients of their psychological and physical health are to be compared to the ratings of healthy patients on these domains, then clearly an individualized approach presents difficulties. However, if the interest is in global perceptions of quality of life, perhaps over time, then an individualized approach provides the necessary data.
4.2
Weighting QOL Scores by Importance
Whilst weighting QOL scores by the perceived importance of the area has been criticized on statistical grounds (e.g., Trauer and MacKinnon, 2001), comparisons of weighted and unweighted QOL scores on the ‘‘PGI’’ suggests that there is no evidence that weighting adversely affects reliability (Ruta et al., 1994). PGI scores derived solely from the ‘‘satisfaction’’ ratings (ratings of the current state of important areas of life compared to the participant’s ideal state) achieved a test-retest correlation of 0.75 (p < 0.001) and the importance weighted index scores
69
70
3
The Global Person Generated Index
achieved 0.70 (p < 0.001), in both cases using respondents indicating no change in health status over time. Similar results have been reported elsewhere, with a slight (although probably non-significant) increase in reliability when the > weighted scores were used (Macduff and Russell, 1998). As such, weighting by importance cannot be seen as a weakness of the measure in statistical terms.
4.3
Advantages of the Measures
The advantages of the recent expansion of the PGI into the GPGI, and the extension of its use to developing countries, are firstly that people like it. Both respondents and researchers enjoyed using the measure (e.g., Bevan et al., 2003) and praised the fact that it was flexible and personcentered and in many ways resembled the participatory methods more commonly used in development research (White and Pettit, 2005). Secondly, the measure can produce genuinely new information, especially if it is then used as the basis for further discussion (e.g., by asking respondents to reflect on GPGI areas and scores provided on a previous occasion). Thirdly, it provides a direct operationalisation of the popular gap theory of quality of life (Calman, 1984), which is experiencing a renaissance due to a growing interest in social comparison and adaptation on the part of economists. Fourthly, it works well cross-culturally, if sensitively translated and carefully piloted. It has also been with used with a more representative sample of the population than is normally the case for cross-cultural validation. Finally, even with the addition of an appraisal schedule or brief contextual interview, it is quicker to administer than a semi-structured interview, albeit more time-consuming than a conventional measure.
4.4
Disadvantages of the Measures’ Approach
The PGI and GPGI have some disadvantages, although arguably many of these apply to all attempts to capture people’s quality of life, whether quantitative or qualitative. Firstly, even categories as apparently straightforward as family are difficult to interpret accurately without an accompanying interview or appraisal schedule and for this reason it can appear to only elicit normative responses. Secondly, the measure is sensitive to framing and context; for example, the translation of important as ‘‘crucial’’ or ‘‘needed’’ in the Ethiopian pilot and the fact that much research in Ethiopia is conducted by Non-Governmental Organizations elicited what the team ironically called ‘‘development-related quality of life’’ (Bevan et al., 2003). Thirdly, even rating satisfaction proved challenging for some respondents. Therefore administration requires skilled and patient interviewers as the method of allotting points to indicate importance is difficult to explain, especially to older and less educated respondents. The standard sample for psychometric validation world-wide is psychology undergraduates; as such the challenges of using QOL measures with other populations are rarely seen during measurement validation. However, QOL measures are now being used with different populations for different purposes – necessitating their validation with these groups. Nonetheless, the GPGI proved popular and effective when used in developing country contexts and provides a means of evaluating development interventions. It raises a further question as to whether a new model of validation should be considered that expands conventional understandings of validity to foreground the accuracy with which the measure has represented the respondent’s world view.
The Global Person Generated Index
3
Summary Points Individualized measures assume that quality of life is determined by the gap between people’s expectations and experiences.
As ‘‘universal’’ items are not equally valued, individualized measures ask respondents to nominate the areas they consider important and assess their performance against their own standards. The PGI has been used in a range of healthcare settings and proved valid and reliable for group comparisons (Martin et al., 2007). It was developed into a measure of general quality of life, the GPGI, which has been successfully translated into more than five world languages.
References Ahmed SM, Mayo NE, Wood-Dauphinee S, Hanley JA, Cohen SR. (2005). Qual Life Res. 14: 2247–2257. Bevan P, Kebede K, Pankhurst A. (2003). A Report on a Very Informal Pilot of the Person Generated Index of Quality of Life in Ethiopia (unpublished report for the ‘‘Wellbeing in Developing Countries Research Group’’ at the University of Bath, UK). Bowling A. (1995). Soc Sci Med. 41(10): 1447–1462. Bradley C, Todd C, Gorton T, Symonds E, Martin A. (1999). Qual Life Res. 8: 79–91. Callaghan BG, Condie ME. (2003). Clin Rehabil. 17: 858–864. Calman K. (1984). J Med Ethics. 10: 124–127. Camfield L, Ruta D. (2007). ‘‘Translation in Not Enough": Using the Global Person Generated Index (GPGI) to Assess Individual Quality of Life in Bangladesh, Thailand, and Ethiopia. Quality of Life Research, 16 (6): 1039–1051. Cummins RA. (1995). Soc Indic Res. 35: 179–200. Cummins RA, Gullone E. (2000). Why we should not use 5-point Likert scales: The case for subjective quality of life measurement. In Proceedings of the Second International Conference on Quality of Life in Cities. National University of Singapore, Singapore, pp. 74–93. Dempster M, Donnelly M. (2000). Qual Life Res. 9: 369–375. Diener E, Emmons RA, Larsen RJ, Griffin. (1985). J Pers Assess. 49(1): 71–75. Ferrans CE, Frisch M. (2004). The Importance of Importance: Why Weight Life Satisfaction. Paper Presented at the Advancing Quality of Life in a Turbulent World, International Society for Quality of Life Studies, Philadelphia, PA.
Griffiths R, Jayasuriya R, Maitlandv H. (2000). Aust N Z J Public Health. 24(5): 529–535. Hensel E, Rose J, Kroese BS, Banks-Smith J. (2002). J Intellect Disabil Res. 46(2): 95–107. Herd RM, Tidman MJ, Ruta D, Hunter JAA. (1997). Br J Dermatol. 136: 502–507. International Wellbeing Group. (2006). Personal Wellbeing Index, 4th ed. Australian Centre on Quality of Life, Deakin University, Melbourne (http:// www.deakin.edu.au/research/acQOL/instruments/ wellbeing_index.htm.). Jongudomkarn D, Camfield L. (2006). Soc Indic Res. 78: 489–529. Lindblad AK, Ring L, Glimelius B, Hansson MG. (2002). Acta Oncologica. 41(6): 507–516. Macduff C, Russell E. (1998). Qual Life Res. 7: 761-769. Martin F, Camfield L, Rodham K, Kliempt P, Ruta D. (2007). Qual Life Res. 16(4): 705–715. McGee H, et al. (1991). Psychol Med. 21: 749–759. Mitchell J, Bradley C. (2004). Qual Life Res. 13: 1163–1175. Narayan DR, Chambers R, Shah MK, Petesch P. (2000). Voices of the Poor: Crying out for Change. Oxford University Press for the World Bank, New York. O’Boyle A, Browne J, Hickey A, et al. (1995). SEIQOL Administration Manual. Department of Psychology, Royal College of Surgeons of Ireland, Dublin. Parducci A. (1995). Happiness, Pleasure, and Judgment: The Contextual Theory and Its Applications. Lawrence Erlbaum Associates, Hove. Ruta D, Camfield L, Donaldson C. (2006). Sen and the Art of Quality of Life Maintenance: Towards a Working Definition of Quality of Life. Wellbeing in
71
72
3
The Global Person Generated Index
Developing Countries (WeD) Research Group Working Paper. Available at: http://www.bath.ac. uk/econ-dev/wellbeing/research/workingpaperpdf/ wed12.pdf. Ruta DA, Camfield L, Martin F. (2004a). Qual Life Res. 13: 1545. Ruta D, Garratt AM, Leng M, Russell IT, MacDonald LM. (1994). Med Care. 32(11): 1109–1126. Ruta D, Martin F, Camfield L, Devine J, Bevan P. (2004b). Assessing Individual Quality of Life in Developing Countries: Piloting a Global PGI in Ethiopia and Bangladesh. Paper presented at the International Society for Quality of Life Research: Harmonizing International Health-Related Quality of Life (HRQOL) Research, Hong Kong. Skevington SM, Bradshaw J, Saxena S. (1999). Soc Sci Med. 48: 473–487. Sprangers MA, Schwartz CE. (1999). Soc Sci Med. 48: 1507–1515.
Streiner D, Norman G. (1995). Health Measurement Scales: A Practical Guide to Their Development and Use. Oxford University Press, Oxford. Symon A, MacKay A, Ruta D. (2003). J Adv Nurs. 42(1): 21–29. Tovbin D, Gidron Y, Jean T, Granovsky R, Schnieder A. (2003). Qual Life Res. 12: 709–717. Trauer T, MacKinnon A. (2004). Qual Life Res. 10(7): 579–585. Tully MP, Cantrill JA. (2002). J Health Serv Res Policy. 7(2): 81–89. Ware JE, Sherbourne CD. (1992). Med Care. 30: 473–483. Weitzenkamp DA, Gerhart KA, Charlifue SW, Glass CA, Whiteneck GG. (2000). Br J Health Psychol. 5: 57–69. White S, Pettit J. (2005). Participatory approaches and the measurement of human wellbeing. In: McGillivray M (ed.) Measuring Human Wellbeing. WIDER/Oxford University Press, Oxford, UK.
4 The Total Illness Burden Index S. Greenfield . J. Billimek . S. H. Kaplan 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
2 2.1 2.2 2.3
Structure and Scoring of the TIBI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 Constructing the Dimensions of the TIBI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 Computing Raw Dimension Scores . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 The Global Severity of Illness Measure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
3
The Predictive Ability of the TIBI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
4
Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
#
Springer Science+Business Media LLC 2010 (USA)
74
4
The Total Illness Burden Index
Abstract: The Total Illness Burden Index is a comprehensive summary measure of case mix or > severity of illness that aggregates all of the patients’ conditions, problems, and diseases, weighting them by their severity. The TIBI offers advantages over other methods of case mix measurement because it (1) is based on clinical information beyond the simple presence or absence of specific diagnoses, (2) is scored using a psychometric approach that avoids reliance on any single variable, (3) predicts health related quality of life and mortality events five or more years in the future, (4) is feasible, utilizing patient report over administrative data while requiring less than 15 minutes to complete, and (5) can be adapted based on the clinical context to assess > comorbidity with respect to a selected index condition, or to assess total morbidity from all of the patient’s conditions. These features allow the measure to be applied in clinical settings among patients facing real-time treatment decisions and for risk adjustment in research on a wide range of medical outcomes. List of Abbreviations: ANOVA, analysis of variance; APACHE II, Acute Physiology and Chronic Health Evaluation II; CP, cardiopulmonary; HRQOL, health-related quality of life; SF-36, Medical Outcomes Study Short Form, a measure of health-related quality of life; TIBI, Total Illness Burden Index; TIBI-CaP, Prostate cancer version of the Total Illness Burden Index
1
Introduction
The Total Illness Burden Index (Greenfield et al., 1995; Litwin; et al., 2007) is a comprehensive summary measure of severity of illness that aggregates all of the patients’ conditions, problems, and diseases, weighting them by their severity. The instrument is a reliable predictor of health-related quality of life (HRQOL) (Greenfield et al., 1995; Litwin; et al., 2007) and mortality (Greenfield et al., 1995; Litwin et al., 2007) that can be used both in clinical practice, primarily office-based practice, and in medical outcomes research. The definitions of severity of illness, appropriate data sources for its determination, and the analytic methods for assembling severity of illness measures used in health outcomes research have varied widely (Iezzoni, 2003). ‘‘case mix’’, ‘‘patient mix’’, ‘‘disease severity assessment’’, ‘‘risk adjustment’’ and ‘‘comorbidity assessment’’ all have been used to describe measures that array patients on a continuum of extent of total disease burden. Most of these measures have been developed to assess differences in utilization of health care services, or differences in mortality (Fortin et al., 2006; Hudon et al., 2005; Perkins et al., 2004; Tooth et al., 2008). Most are based on physician-reported diagnosis from claims databases or from medical records, and include only selected major diseases or conditions (Crabtree et al., 2000). Several diagnosis-based measures have been developed for reimbursement purposes in outpatient settings (Charlson et al., 1987; Pearte et al., 2006; Perkins et al., 2004), but these measures often do not represent the severity and therefore the prognosis of the condition, and further, are often not readily available at the time of decision-making. Summary measures of patient-reported HRQOL have been shown to predict mortality but are based on variables with which clinicians are less comfortable than the review of systems variables that they routinely collect (Min et al., 2007; Saliba et al., 2001). Most available measures of comorbidities are only available after treatment and have been used primarily to adjust group differences (Charlson et al., 1987; Greenfield et al., 1993).
The Total Illness Burden Index
4
In response to these shortcomings, the TIBI was developed as a patient-reported measure of comorbidity based on items akin to the traditional review of systems that patients regularly respond to when queried by their physicians. The instrument represents an approach to case mix or severity of illness measurement that is unique in five major ways. First, the TIBI is based on clinical information (symptoms and events) beyond diagnosis, addresses not only the presence but the severity of the patient’s conditions, independent of the many specific diagnoses a patient may have. In this regard, it closer to the APACHE II (Knaus et al., 1985) than to diagnosis-based instruments; however, in contrast to APACHE, it does not rely on a small number of non-specific variables, such as hemoglobin, representing end stage systemic breakdown. Second, it is scored and summarized using a psychometric approach that avoids reliance on any single variable, because single variables are inherently unreliable. It is constructed to reflect empirical confirmation of clinically defined scales, with special attention to analytic techniques that compensate for missing data, unreliability of clinical information, and interactions between variables. Third, it is aimed not at a short term time span, such as one month or one year, but on the more distant future, three to five years or longer. Fourth, the measure is feasible, although more time consuming and expensive than those indexes derived from administrative data, and the data can be collected in an office practice, without access to information that needs to be gathered from the medical record or administrative data sources. Fifth, it is flexible in its use. Dimensions of illness can be added or removed based on the context. In addition, although it is most often used to measure comorbidity in relation to an index condition, when that index condition is included among the dimensions, it becomes a measure of total morbidity, or ‘‘> multimorbidity’’. In this chapter, we describe the structure and scoring of the TIBI, and provide data to support its reliability and validity. Then we will return to the characteristics described above so that a user may be able to compare instruments and approaches.
2
Structure and Scoring of the TIBI
The TIBI was developed as a battery of over 100 items reflecting illness severity for 15 ‘‘dimensions’’, or sets of diseases or conditions, grouped by body system to assess comorbidity in type 2 diabetes patients (Greenfield et al., 1995). From this initial battery, we can modify the instrument to fit the clinical context, as we have done, for example, to assess for prostate cancer 11 dimensions of comorbidity with 84 patient reported items (Litwin et al., 2007; Stier et al., 1999). Below, we describe how the dimensions were (1) constructed, (2) scored and (3) aggregated to compute a global severity of illness score.
2.1
Constructing the Dimensions of the TIBI
Construction of the TIBI began with physicians, working independently, specifying the relevant variables (mostly symptoms, with some diagnoses) that should be included in a questionnaire to reflect the most common diseases, disease manifestations, or conditions within each body system that would be likely to affect patients’ functional outcomes or mortality over the next 3–5 years. Item content was revised based on physician consensus and on survey research principles. Each dimension was then scored on a clinical basis, and modified after the data were collected. For example, the items for chronic lung disease
75
76
4
The Total Illness Burden Index
include: orthopnea requiring extra pillows, frequency of shortness of breath while performing various activities, amount of cough and volume of sputum production, and numbers of episodes of bronchitis and flu in the past 6 months. For different uses, we have added, deleted, or combined dimensions to fit the medical context. However, the items reflecting the traditional review of systems variables, within each dimension, are stable and basic to the predictive accuracy of the instrument. First, dimensions reflecting the index disease should be deleted so that the instrument only includes true comorbid conditions. Second, dimensions reflecting previously unmeasured comorbidities (e.g., conditions that were the index disease in prior versions) can be added. Finally, the instrument can be shortened by deleting whole dimensions based on clinical input, low relevance, or poor scaling with outcome variables of interest, or by aggregating dimensions for which the symptoms overlap, when supported by data analysis. In any given situation, deletion or modification of dimensions could be justified by using analyses which removed a dimension and then recalculated the explained variance in one or more key dependent variables with and without that dimension. As an example, we modified the original TIBI to produce a version measuring 11 dimensions of comorbidity in prostate cancer patients. This version, called the TIBI-CaP was created by adding and deleting selected items to represent prostate cancer as the index condition based on results from a pilot study (Stier et al., 1999). Changes in the constituent dimensions of the two versions are outlined in > Table 4-1. Renal disease and genitourinary problems were dropped from the TIBI-CaP because they were closely related to the index condition of prostate cancer. Diabetes, on the other hand, which was the index condition in the original TIBI was now added as a possible comorbid condition in the TIBI-CaP. A dimension for other cancers, excluding prostate, was also added
. Table 4-1 Individual Body System Disease Measures of the original TIBI and the TIBI-CaP The Original TIBI (type 2 diabetes)
TIBI-CaP (prostate cancer)
Hearing problems
Hearing problems
Eye and vision conditions
Eye and vision conditions
Hypertension
Hypertension
Foot disease
Foot disease
Atherosclerotic heart disease
Atherosclerotic heart disease
Stroke and neurological disease
Stroke and neurological disease
Arthritis
Arthritis
Renal disease
Other cancers (excluding prostate)
Genitourinary problems
Diabetes
Chronic lung disease
Cardiopulmonary disease
Congestive heart failure Gastrointestinal autonomic neuropathy Lower gastrointestinal disease Upper gastrointestinal disease Nonspecific bowel disease
Gastrointestinal conditions
The Total Illness Burden Index
4
because cancers frequently spread to multiple body systems. Because shortness of breath was a symptom shared by chronic lung disease and congestive heart failure, these two dimensions were combined to form a single dimension called cardiopulmonary disease. Similarly, items from gastrointestinal autonomic neuropathy, lower gastrointestinal disease, upper gastrointestinal disease and nonspecific bowel disease were combined into a single dimension of gastrointestinal conditions. In addition to allowing researchers to define comorbidity in the most clinically relevant way in terms of an index condition, the ability to modify the instrument improves its feasibility by allowing unneeded items and dimensions to be pared away. It is usually estimated in survey research that a person can respond to about six items per minute, so that it takes 12 or so minutes for a patient to complete the TIBI-CaP. Because prior analyses have shown that each of the dimensions in the versions studied so far has correlates closely to various dimensions of HRQOL measured with the Medical Outcomes Study Short Form 36 (SF-36; Stewart et al., 1988), it may be hard to reduce the instrument to less than 50 or so items and retain its strong predictive capacity.
2.2
Computing Raw Dimension Scores
Physicians, working as a group, classified patient responses across items within each question (i.e., disease manifestations, symptoms) into multiple ordinal scale points reflecting severity levels. The scoring system for the cardiopulmonary (CP) dimension of the TIBI-CaP is presented in > Table 4-2 as an example. The CP dimension score can increase by one point if a patient has ever been told by a physician that they have had one or more of the following conditions: emphysema, chronic bronchitis, or asthma. An additional one to three points can be added based on the frequency of any of the following symptoms: pneumonia, bronchitis (for which the patient took antibiotics) or flu (with coughing). Another point is added if the patient uses extra pillows at night because of breathing problems. One or two additional points can be added based on the reported volume of sputum coughed up in a typical day. A report of frequent wheezing adds another point to the CP score. And finally, the frequency and causes of shortness of breath can be interpreted to add another one, three or five points to the CP score. Using this scoring scheme, a patient’s CP dimension score can range between zero and thirteen. For CP and the other dimensions, there were multiple ways to reach each level, thus minimizing misclassification due to missing or unreliable responses. Using this technique, physicians created clinically defined severity scales using diagnoses and symptoms for each of the different diseases or conditions, considered separately.
2.3
The Global Severity of Illness Measure
We then constructed a composite, global, or summary measure by aggregating the individual dimensions of the TIBI into a single score. First, clinicians ranked each disease or condition grouped by body system as having minimal, moderate, or severe impact on functional outcomes in an ambulatory population. For the TIBI-CaP, those diseases or conditions clinically defined as having minimal negative impact on functional outcomes in prostate cancer patients
77
78
4
The Total Illness Burden Index
. Table 4-2 Computing the raw score for the Cardiopulmonary (CP) Dimension of the TIBI-CaP
The Total Illness Burden Index
4
. Table 4-2 (continued)
included vision, hearing loss, diabetes, and hypertension. Diseases or conditions classified as having moderate negative impact on functional outcomes included foot disease and arthritis. Other cancers, GI conditions, atherosclerotic heart disease and cardiopulmonary disease were classified as having the greatest negative impact on functional outcomes. In order to weight the dimensions differentially, those considered to have the greatest clinical impact on illness burden were stratified, based on clinical judgement, into 4 severity levels (0–3 points), those with intermediate impact into 3 severity levels (0–2 points), and those with the least impact into 2 severity levels (0–1). The raw dimension scores were then recoded into these severity levels using clinically defined cutpoints. The severity level cutpoints for each dimension were then validated using analysis of variance (ANOVA) to compare scores on the Physical Functioning and Role Physical subscales of the SF-36 across the severity levels of each dimension. For every dimension of the original TIBI and the TIBI-CaP, patients falling into higher levels of severity reported worse HRQOL at the time the TIBI was administered. This same association between severity levels and HRQOL persisted for patients completing the SF-36 six months after completing the TIBI-CaP (> Table 4-3). In addition to validating the clinically defined cutpoints for the severity levels, these analyses support the modular structure of the instrument because each dimension independently predicts HRQOL. The severity levels for all the dimensions were summed to create the global score for the TIBI. The TIBI-CaP global score, for example, could theoretically range from 0 to 23, but the actual scores measured in our sample ranged from 0–18 with a mean of 3.5 (SD=2.6) and a median of 3. The amount of missing data was small – no items had more than 10% missing, and 78 of the 84 items had less than 4% missing. For those dimensions with 1 or more items missing, the dimension score was computed on the completed items based on the scoring rules. For presentation purposes, we collapse the global TIBI scores into five levels with 3-point intervals from least severe (scores of 0–2) to most severe (scores of > 12).
79
80
4
The Total Illness Burden Index
. Table 4-3 SF-36 Health-related quality of life scale scores at 6-month follow up by severity levels of the dimensions of the TIBI-CaP, mean (SD)* N
Physical Functioning Mean (SD)
Role Physical Mean (SD)
0
1961
86 (18)
80 (34)
1
490
72 (25)
58 (42)
2
291
62 (28)
48 (43)
3
66
50 (31)
30 (40)
0
1312
83 (22)
78 (36)
1
1439
79 (23)
71 (39)
2
162
68 (28)
49 (44)
3
87
57 (28)
35 (39)
0
2606
81 (23)
74 (38)
1
221
75 (25)
62 (43)
2
120
65 (28)
51 (44)
3
23
41 (24)
27 (39)
0
2321
83 (22)
76 (37)
1
454
72 (26)
60 (42)
2
105
64 (29)
42 (43)
3
30
60 (29)
33 (43)
0
2559
81 (23)
74 (38)
1
241
75 (23)
61 (42)
2
95
70 (27)
50 (44)
3
34
60 (31)
45 (46)
0
1689
86 (20)
81 (34)
1
1191
73 (25)
61 (42)
2
94
54 (30)
35 (42)
0
2390
83 (21)
76 (36)
1
309
72 (24)
58 (43)
2
209
55 (29)
39 (42)
0
1743
84 (21)
78 (36)
1
1230
73 (26)
63 (42)
TIBI-CaP Dimension Points Cardiopulmonary
Atherosclerotic heart disease
Stroke/neurologic
GI conditions
Other cancer
Arthritis/joints
Feet
Eyes/Vision
4
The Total Illness Burden Index
. Table 4-3 (continued) N
Physical Functioning Mean (SD)
Role Physical Mean (SD)
0
1510
84 (22)
79 (35)
1
1458
75 (25)
64 (42)
0
2773
80 (23)
73 (39)
1
227
70 (27)
57 (43)
0
2516
81 (22)
74 (38)
1
445
70 (27)
58 (43)
TIBI-CaP Dimension Points Hearing
Hypertension
Diabetes
*p Table 4-4). Each increasing category of the aggregate TIBI-CaP had a correspondent statistically significant decrease in levels of physical function and role-physical scores compared to the adjacent category. With TIBI-CaP treated as a continuous variable, we examined its contribution to variation in the subdimensions of the SF-36, compared with sociodemographic characteristics alone (> Table 4-5). A model including the TIBI as a covariate alongside sociodemographic
81
82
4
The Total Illness Burden Index
. Table 4-4 SF-36 Health-related quality of life scale scores at 6-month follow up by severity levels of TIBI-CaP, mean (SD)* n
Physical Functioning Mean (SD)
Role Physical Mean (SD)
0–2
1074
91 (13)
89 (25)
3–5
1026
80 (21)
71 (38)
6–8
378
65 (26)
47 (42)
9–11
94
51 (27)
31 (39)
12+
33
35 (18)
10 (23)
EQ-5D) is one of the most commonly used generic questionnaires to measure health-related > quality of life (HRQOL). The conceptual basis of the EQ-5D is the holistic view of health, which includes the medical definition, as well as the fundamental importance of independent physical, emotional and social functioning. The concept of health in EQ-5D also encompasses both positive aspects (well-being) and negative aspects (illness). The EQ-5D is short, easy to use and flexible. It has been used successfully in several different settings (scientific trials, health policies, pharmacoeconomics, clinics, etc.). It consists of a questionnaire and a > visual analogue scale (EQ-VAS). The EQ-VAS is a self-rated health status using a VAS. The EQ-VAS records the subject’s perceptions of their own current overall health and can be used to monitor changes with time. The self-assessment questionnaire is self-reported description of the subject’s current health in 5 dimensions i.e., mobility, selfcare, usual activities, pain/discomfort and anxiety/depression. The subject is asked to grade their own current level of function in each dimension into one of three degrees of disability (severe, moderate or none). The combination of these with the conditions ‘‘death’’ and ‘‘unconscious’’ enables description of 245 different health states. Each health state can be ranked and transformed a single score called the utility. The utility score is an expression of the Quality Adjusted Life Years (QALY) and is commonly used to make evidence-based decisions in analyses of cost-effectiveness. Therefore, the EQ-5D can be used for health outcomes studies and economic analyses. The EQ-5D was initially developed for adults; a new version has recently been developed for children aged 8–18 years old (> EQ-5D-Y). List of Abbreviations: CF-EQ-5D, child friendly euroQOL five dimensions questionnaire; EQ-5D, euroQOL five dimensions questionnaire; EQ-5D-Y, euroQOL five dimensions questionnaire for young; HRQOL, health related quality of life; NICE, National Institute of Clinical Excellence; QALY, quality adjusted life years; SF-6D, 6-item short form health survey questionnaire; TTO, time-trade-Of; VAS, visual analogue scale
1
Introduction
Assessment of health care requires measurement and monitoring of health. However, healthrelated quality of life (HRQOL) is an abstract, subjective concept that is difficult to measure. Several questionnaires are available but these instruments are usually long, complex and disease-specific limiting their usefulness and comparability. Simple devices including a basic common core are needed to facilitate comparison of health care outcomes. The EuroQOL five dimension questionnaire (EQ-5D) is a simple self-administered instrument that assesses HRQOL. It assesses function in five socially relevant domains: 1. 2. 3. 4. 5.
Mobility Self-care Usual activities Pain-discomfort Anxiety-depression
It is accompanied by a Visual Analogue Scale (VAS) on which the subject is asked to provide a self-assessment of their own health in a range from 0 (worst imaginable health state) to 100 (best imaginable health state).
The EQ-5D Health-Related Quality of Life Questionnaire
5
The origins of the EQ-5D date back to May 1987 when the EuroQOL Group first met to consider the development of a standardized non-disease-specific instrument for describing and valuing HRQOL (EuroQOL Group, 1990). The Group originally consisted of a network of 23 multilingual, multidisciplinary researchers from five European countries (the UK, Finland, the Netherlands, Norway and Sweden). The aim of this group was to develop a tool that provided a simple, generic measure of HRQOL for clinical or economic analyses that enabled international comparisons. Their initial objectives and the outcomes of their meetings were described in their first publication (EuroQOL Group, 1990). The EQ-5D was initially developed simultaneously in Dutch, English, Finnish, Norwegian and Swedish. It is now widely used and has been validated in many other countries and has been translated into most major languages. For several years the term ‘‘EuroQOL’’ was synonymous with the instrument because the EuroQOL Group developed it. However, because the current version of the ‘‘EuroQOL Instrument’’ assesses five dimensions of function in adults it was actually called EQ-5D. To avoid confusion the term EQ-5D should be used when referring to this instrument. The EQ-5D is one of the most commonly used standardized HRQOL questionnaires available today and has been used to measure the cost-effectiveness of therapies for many diseases. It is short, flexible, easy to use and can generate a single total score based on socially relevant measures of HRQOL. This score is known as the utility score. The importance and popularity of the EQ-5D are demonstrated by the fact that over 1,000 scientific papers listed in the ISI Web of KnowledgeTM refer to the EQ-5D and the utility score. The (EuroQOL Group, 1990) initially developed an instrument that assessed six dimensions of function EQ-6D; (EuroQOL Group, 1990) and at one stage the addition of a seventh dimension was even suggested. The assessments of the sixth and seventh domains (social relationships and energy-tiredness) made little contribution to the assessment of health status and so were removed (Williams, 2005). The recent suggestion to add a new cognitive dimension was also rejected because the current version of the EQ-5D can adequately assess HRQOL in populations with cognitive impairments (Wolfs et al., 2007). Recently, the need to develop a self-administered tool to measure the HRQOL in children and adolescents prompted adaptation of the EQ-5D. The EuroQOL Group has developed and validated the EuroQOL five dimensions questionnaire for young people (EQ-5D-Y). This is a new version of the EQ-5D for use in populations of children and adolescents aged 8–18 years old. The first results have been presented at recent EuroQOL Scientific Meetings and several research papers are currently in press and publication is expected during the second half of 2008 (Gusi et al., 2009; Wille et al., 2006).
2
Definitions and Concepts
To administer a questionnaire appropriately and interpret the results correctly it is crucial to understand the concepts underlying its development. The specific-validity and conceptual basis of questionnaires are reflected in the definitions of the concepts and words included in it. The child-friendly task force of the EuroQOL Group recently presented a revised version of the concepts and definitions included in the standard adult version of the EQ-5D. These definitions were adapted to children and adolescents (Wille et al., 2006). These concepts and definitions are available on the EuroQOL web site (www.euroQOL.org).
89
90
5
The EQ-5D Health-Related Quality of Life Questionnaire
EQ-5D manages eight main concepts:
The subject’s health state at the time of assessment The best and worst imaginable health states The five dimensions of EQ-5D The EQ-5D is a generic instrument intended for use by different health professionals. The conceptual basis of the EQ-5D is therefore a holistic view of health, which includes the medical definition, as well as the fundamental importance of independent physical, emotional and social functioning. For example, the worst imaginable health state, in terms of capacity for independent function, is that which prevents an acceptable level of function in all aspects of life. The concept of health in EQ-5D also encompasses both positive aspects (well-being) and negative aspects (illness). Subjects are asked to grade their own level of function or disability in each dimension on the specific day that the questionnaire is administered into one of three levels;
No problems, no disability (level 1) Some problems, moderate disability (level 2) A lot of problems, severe disability (level 3) The levels of function in the mobility dimension, for example, are classified as:
level 1 – able to walk easily, both inside and out, unaided level 2 – able to walk with difficulty/require some assistance (sticks, crutch, analgesia) level 3 – extremely difficult to walk inside or out/confined to bed or chair These difficulties may be due to a chronic condition (e.g., confined to bed because of terminal cancer) or acute illness (e.g., confined to bed because of flu). Similarly, the other dimensions refer to independence in daily personal care and usual activities rather than skills that can be learnt (e.g., the ability of children to dress themselves). The pain/discomfort dimension refers to physical injury or discomfort (ache, breathlessness, itching, palpitations, tiredness, ringing in the ears, etc.). Psychological or mental sufferings are assessed in the fifth dimension (anxiety/depression). This is related to a broad concept of psychological disturbance covering clinical depression and anxiety, feeling gloomy, dejected, down, sad, or unhappy, etc.
3
Use of EQ-5D
The EQ-5D is intended to be self-completed and has been used in postal surveys, clinics and faceto-face interviews (Szende and Williams, 2004). For practical reasons postal surveys using EQ-5D have been used in large population studies whilst face-to-face interviews are used with inpatients, the elderly and the illiterate or to increase the accuracy of responses. Proxy versions (> proxy questionnaire) have also been used to obtain information from the caregivers of subjects who could not complete the questionnaire themselves (e.g., coma, dementia, young children, etc. (Hung et al., 2007; Jonsson et al., 2006; Matza et al., 2005; Sach and Barton, 2007). The EQ-5D includes five questions related to the five dimensions that are measured. The level of function in each dimension is classified into one of three degrees of disability, reflecting no disability (level 1), moderate disability (level 2) and severe disability (level 3). The yuxtaposition of the responses to the five questions generates a 5-digit descriptor ranging
The EQ-5D Health-Related Quality of Life Questionnaire
5
from 11111 for perfect health to 33333 for the worst possible state. This descriptor represents a unique heath state. The patient’s health state can therefore be classified into any one of 243 theoretically possible health states. The states of ‘‘unconscious’’ and ‘‘dead’’ were added in order to obtain a value set or ‘‘tariff’’ for evaluation of the EQ-5D generated health states. So there are in fact 245 possible health states or EQ-Index. The subjects are asked to rate their own health on a vertical 20 cm visual analogue scale (EQ-VAS) included with the EQ-5D. The scale ranges from 0 (worst imaginable health state) up to 100 (best imaginable health state). The inclusion of this scale essentially allows numerical assessment of the patient’s perception of their own general health and HRQOL using a selfcompleted instrument that could be conducted by postal survey (Brooks, 1996). The EQ-VAS is an economical and practical addition to the EQ-5D that provides a general and unique insight into patients’ subjective perception of their own overall current health. It can also be used like a barometer to monitor changes in HRQOL over time. The EQ-VAS values obtained from large population surveys were used to obtain the reference values related to the 245 possible health states in different countries. This set of reference values is labeled the VAS tariff. The reference values for the EQ-5D scores obtained in various countries can be found in Measuring self-reported population health: An international perspective based on EQ-5D (Szende and Williams, 2004). With the availability of more time and research funds, the value sets were redefined using the Time Trade-Off (TTO) and Standard-Gamble (SG) methods which reflect the social acceptability of having various health states. Time Trade-Off is most commonly used. The book EQ-5D Value Sets: Inventory, Comparative Review and User Guide (Szende et al., 2007) provides comparative reviews of the TTO and VAS value sets, a guide for users of EQ-5D value sets and inventories of TTO and VAS evaluation surveys. The application of TTO to EQ-5D results usually reflects the preference of having one health state over another during the next 4 or 10 years. The mathematical combination of these preferences gives a total score that links each health state with time. This total score ranges from 1 (fully functional quality of life) to 0 (death) but also allows negative scores (worse than death). This score is called utility is linked to the quality-adjusted life years (QALYs) and it is very useful for economic analyses and pharmacoeconomics (Devlin and Williams, 1999). To date, over 250 scientific papers have used the utility score in economic analyses (costutility analyses) to describe the health of a population, effects of interventions, cost-effectiveness analyses or make decisions etc. The use of EQ-5D in economics analyses (Betegon and Badia, 2006) is increasing. It is recommended by well-recognised guidelines of institutions such as the National Institute of Clinical Excellence (NICE) in the United Kingdom and the Washington Panel on Cost-Effectiveness in Health and Medicine in the United States.
4
Description of a Sample Population Using the EQ-5D
This section illustrates how to use EQ-5D data to describe a sample using data from the ‘‘Exercise looks after you’’ program funded by public Regional Government of Extremadura in Spain (Gusi et al., 2007). This program recruited people over 55 years old with various different diseases detected by the primary care team, including for example obesity, diabetes or moderate depression. Thus this data is not representative of the general population. There are several ways to present the data obtained with the EQ-5D but for simplicity we will use the same subdivisions used in the EQ-5D user guide (Oppe et al., 2007):
91
92
5
The EQ-5D Health-Related Quality of Life Questionnaire
1. The descriptive system as a health profile 2. The EQ VAS as a measure of overall self-rated health status 3. The descriptive system as a weighted index
4.1
The Descriptive System as a Health Profile
A simple way to describe a population using the EQ-5D is with a frequency table of the health states obtained. > Table 5-1 is an example of a frequency table of the health states obtained using the EQ-5D. Health states with a frequency under 10 have been removed because of the large number of different states obtained (80).
. Table 5-1 Frequency of EuroQOL 5-Dimension questionnaire (EQ-5D) Health states obtained in a sample from the Exercise Looks After You program Q-5D health states
n
%
11111
360
26.3
11112
93
6.8
11113
12
0.9
11121
237
17.3
11122
133
9.7
11123
33
2.4
11131
40
2.9
11132
37
2.7
11133
26
1.9
11211
11
0.8
11221
14
1.0
11222
22
1.6
21111
22
1.6
21121
55
4.0
21122
45
3.3
21131
12
0.9
21132
16
1.2
21133
10
0.7
21222
14
1.0
21232
12
0.9
22232
11
0.8
Others Total Missing
156
11.3
1,368
99.8
3
0.2
These data are from the Exercise Looks After You program and include elderly people with multiple diseases
15.0
25.0
Level 3
71–80
Age groups +80
Total
13.3
40.0 7.2
20.5
72.3
10.8
38.6
50.6
1.2
9.6
89.2
1.2
8.4
90.4
0
21.7
78.3
9.4
35.7
54.9
17.6
45.3
36.8
0.9
11.6
87.5
0
5.4
94.6
4.0
18.3
81.3
3.1
22.9
74.0
6.3
43.2
50.5
0
3.1
96.9
0
3.1
96.9
0
10.4
89.6
10.5
30.8
58.7
17.5
47.0
35.6
0.9
14.4
84.7
0.4
7.8
91.8
2.0
26.5
73.3
6.9
17.2
75.9
13.8
31.0
55.2
0
13.8
86.2
0
13.8
86.2
0
17.2
82.8
5.3
32.9
61.8
15.8
51.3
32.9
3.9
30.3
65.8
0
11.8
88.2
0
18.4
81.6
6.0
21.8
72.2
8.8
39.1
52.1
0.5
7.4
92.1
0.5
6.5
93.1
0
15.7
84.3
9.8
33.7
56.5
17.3
46.5
36.0
1.1
13.8
85.1
0.2
7.4
92.4
0.3
22.0
77.8
These data are from the Exercise Looks After You program and include elderly people with multiple diseases. Women report higher levels of disability than men. Reported levels of disability do not increase with the age because of the specific characteristics of this population. M males; F females; values expressed in %
37.5
Level 2
46.7
0
37.5
Level 3
48.3
25.0
1.7
36.7
Level 2
0
8.3
90.9
0
16.7
83.3
0
25.0
75.0
75.0
Level 3
61–70
F (n = 60) M (n = 83) F (n = 552) M (n = 96) F (n = 465) M (n = 29) F (n = 76) M (n = 216) F (n = 1,153)
Level 1
12.5
Level 2
0
87.5
Level 3
Level 1
0
Level 2
0
Level 3
100
12.5
Level 1
87.5
Level 2
M (n = 8)
Level 1
Sex
Anxiety/depression Level 1
Pain/discomfort
Usual activities
Self-care
Mobility
EQ-5D Dimension
51–60
. Table 5-2 EuroQOL 5-Dimension questionnaire (EQ-5D) levels of disability obtained in the sample from the Exercise Looks After You program (n = 1,371). Reported by age and sex
The EQ-5D Health-Related Quality of Life Questionnaire
5 93
94
5
The EQ-5D Health-Related Quality of Life Questionnaire
The descriptive system is usually used with a frequency table presenting the proportion of answers obtained in each dimension by levels or using dichotomized information structuring the answers by dimension. An example of a frequency table can be seen in > Table 5-2. In this example the table has been expanded to show the information by demographic groupings (age and sex). An example of dichotomized information can be seen in > Figure 5-1, where the data is structured by percentage of reported problems by dimension and sex.
4.2
The EQ VAS as a Measure of Overall Self-Rated Health Status
The data should be presented as a measure of the central tendency with a measure of dispersion (e.g., mean and standard deviation or median with 25th and 75th percentiles). This information can be divided by age subgroups, sex, illness, etc. > Figure 5-2 is an example of this. Normally the score decreases with the age and men have higher scores than women. In this sample men had higher scores than women but the scores did not decrease with the age.
4.3
The Descriptive System as a Weighted Index
The EQ-5D index can be presented in the same format as the VAS with a measure of central tendency and a measure of dispersion. For more precise valuation, a specific value set should be used for data from a specific country. These examples have used the Spanish value set (Badia et al., 1999; Badia et al., 2001b). > Figure 5-3 shows the VAS value set for the ‘‘Exercise looks
. Figure 5-1 EuroQOL 5-Dimensions questionnaire (EQ-5D) levels of reported problems in each dimension by sex. These data are from the Exercise Looks After You program and include people with multiple diseases so are not representative of the general population. Women reported higher levels of disability in all dimensions. Values are expressed as a percentage of reported problems
The EQ-5D Health-Related Quality of Life Questionnaire
5
. Figure 5-2 Visual Analogue Scale (VAS) scores from the EuroQOL 5-Dimensions questionnaire (EQ-5D) by age and sex. These data are from the Exercise Looks After You program and include people with multiple diseases so are not representative of the general population. Males report lower (better) scores than females. Values expressed as mean ± standard deviation
after you’’ population and > Figure 5-4 shows the TTO value set. > Figure 5-5 shows the hypothetical effect of a treatment on health status (these data are fictitious and for illustration purposes only).
5
Recent Developments in EQ-5D
The EuroQOL foundation recently conducted several new surveys for the further development of the EQ-5D questionnaire. Some of these surveys were used to evaluate a five level version of the EQ-5D and the new version of the EQ-5D for children (EQ-5D-Y).
5.1
The Five Level Version of the EQ-5D
Although the EQ-5D has good psychometric properties and is able to detect small changes in chronic diseases (Campbell et al., 2006; Gusi et al., 2006), there are only 243 possible health states. This is considered a disadvantage in comparison to other instruments like the Health Utilities Index Mark 2 & Mark 3, and the Short Form 6D (SF-6D) which have 24,000, 972,000 and 18,000 possible health states respectively (Janssen et al., 2008).
95
96
5
The EQ-5D Health-Related Quality of Life Questionnaire
. Figure 5-3 EurQOL 5-Dimensions questionnaire (EQ-5D) value set based in the Visual Analogue Scale (VAS) by age and sex. These data are from the Exercise Looks After You program and include people with multiple diseases so are not representative of the general population. Men reported lower (better) scores than women. Values expressed as mean
If the severity levels in each domain of the EQ-5D were expanded from 3 to 5, the number of possible health states would increase to 3,125. This would increase the descriptive capacity of the EQ-5D significantly. This would improve its ability to detect small changes in health states. Nevertheless, the currently available three level version is easy to use, functions well and provides the utility score. Some authors have performed preliminary research on the psychometric characteristics of a five level version of EQ-5D. (Janssen et al., 2008; Pickard et al., 2007a; Pickard et al., 2007b) and the compatibility of the utility scores obtained from the three and five level versions. Kind (2007) suggested some factors which need to be considered in the development of a new five level EQ-5D. Kind indicated that ‘‘the task of establishing value sets for EQ-5D health states is an essential complementary exercise.’’ To expand the value sets for a 5 level version would take time and the currently available, standardized three level version would be a useful reference. For further discussion of the advantages and disadvantages of varying the number of severity levels of the EQ-5D see Van Agt and Bonsel (2005).
5.2
The Version of EQ-5D for Children (EQ-5D-Y)
The EQ-5D was developed for use in adults. Although this instrument has been used by some researchers to study the HRQOL in children (Badia et al., 2001a; Polinder et al., 2005;
The EQ-5D Health-Related Quality of Life Questionnaire
5
. Figure 5-4 EurQOL 5-Dimensions questionnaire (EQ-5D) value set based in the Time Trade Off (TTO) by age and sex EQ-5D TTO value set by age and sex. These data are from the Exercise Looks After You program and include people with multiple diseases so are not representative of the general population. Men reported lower (better) scores than women. Values expressed as mean
Stolk et al., 2000), children under 16 years of age may have difficulty understanding the questionnaire and VAS. As a reuslt, Hennessy and Kind (2002) developed an experimental version of the EQ-5D for use with children. In 2006 the EuroQOL Group designated an international Task force with the objective to develop a specific international version of EQ-5D for young children & adolescents (8–18 years old). The version was initially named the Child Friendly EQ-5D (CF-EQ5D) but the name was later changed to the EQ-5D for Young (EQ-5D-Y). The development of this international version of the EQ-5D for the young required the cooperation of a multinational team who tested versions in five different languages (English, German, Italian, Spanish and Swedish). The EQ-5D-Y was initially produced in English. This English version was the starting point for the versions in other languages. This ensured the > conceptual equivalence of each version and enables international comparisons. The EQ-5D-Y classifies subjects into one of three levels of function (severe disability, moderate disability or no disability) in five dimensions: 1. 2. 3. 4. 5.
Walking about Look-after myself Doing usual activities Having pain or discomfort Feeling worried, sad or unhappy
The research papers presenting the validity and characteristic proprieties of the different language versions of the EQ-5D-Y will be published soon. For example a paper presenting the validity of the Spanish version is currently in press and due to be published in late 2008.
97
98
5
The EQ-5D Health-Related Quality of Life Questionnaire
. Figure 5-5 EuroQOL 5-Dimensions questionnaire (EQ-5D) value set based on Time Trade Off (TTO) before and after a hypothetical treatment. Values expressed as mean. The data show the effect of the treatment in the quality of life. The control group and intervention group increase their scores, however the treatment group scores are higher after treatment, so the treatment is effective. This data are fictitious and for illustration purposes only
Parental proxy-versions to measure the HRQOL in children under 8 years old are currently being developed.
Summary Points The EuroQOL Group developed the EQ-5D, one of the most widely used generic questionnaires to measure HRQOL.
The EQ-5D is characterized by its ease of use, brevity and flexibility in different settings (trials, health policies, pharmacoeconomics, clinics, etc.).
The EQ-5D is a generic instrument intended for use by different health professionals. Thus its conceptual basis incorporates the holistic nature of HRQOL which includes the medical interpretation as well as physical, emotional and social functioning. The concept of health in EQ-5D includes both negative aspects (illness) and positive aspects (well-being). Most of the concepts relate to the ability to function independently. EQ-5D classifies subjects into one of three levels of function in five dimensions: mobility, self-care, usual activities, pain/discomfort and anxiety/depression. The combination of scores for question (11111–33333) plus the states of death and unconscious offers a descriptive system with up to 245 different possible health states.
The EQ-5D Health-Related Quality of Life Questionnaire
5
Each health state could be transformed and ranked in a single score (utility) which is thought to correlate with Quality of Life Adjusted-Years (QALY).
EQ-5D was developed for adults and the EQ-5D-Y has been recently developed for children and adolescents aged 8–18 years old.
Well recognized institutions including NICE in the UK recommend the use of EQ-5D for monitoring HRQOL and in economic analyses.
References Badia X, Garcia-Hernandez G, Cobos N, Lopez-David C, Nocea G, Roset M. (2001a). Med Clin (Barc). 116: 565–572. Badia X, Roset M, Herdman M, Kind P. (2001b). Med Decis Making. 21: 7–16. Badia X, Roset M, Montserrat S, Herdman M, Segura A. (1999). Med Clin (Barc). 112 (Suppl. 1): 79–85. Betegon L, Badia X. (2006). In: Badia X (ed.) Review of the use of the EQ-5D in cost-utility analysis. Paper Presented at the 23rd Scientific Plenary Meeting of the EuroQOL Group. Health Economics & Outcomes Research IMS Health, Barcelona. Brooks R. (1996). Health Policy. 37: 53–72. Campbell H, Rivero-Arias O, Johnston K, Gray A, Fairbank J, Frost H. (2006). Spine. 31: 815–822. Devlin N, Williams A. (1999). N. Z. Med J. 112: 68–71. EuroQOL Group. (1990). Health Policy. 16: 199–208. Gusi N, Badia X, Herdman M, Olivares PR. (2009). Aten Primaria. (in press). Gusi N, Herrera E, Davila M, Campon JC. (2007). Development of the socio-sanitary program ‘‘exercise looks after you’’: phase I for elderly. Paper Presented at the Third Annual Meeting of HEPA Europe, Graz (Austria). Gusi N, Tomas-Carus P, Hakkinen A, Hakkinen K, OrtegaAlonso A. (2006). Arthritis Rheum. 55: 66–73. Hennessy S, Kind P. (2002). Measuring health status in children:developing and testing a child friendly version of EQ-5D. Paper Presented at the EuroQOL Plenary Meeting. University of New York, New York. Hung SY, Pickard AS, Witt WP, Lambert BL. (2007). J Clin Epidemiol. 60: 963–970. Janssen MF, Birnie E, Bonsel GJ. (2008). Qual Life Res. 17(3): 463–473. Jonsson L, Andreasen N, Kilander L, Soininen H, Waldemar G, NygaardH, Winblad B, Jonhagen ME, Hallikainen M, Wimo A. (2006). Alzheimer Dis Assoc Disord. 20: 49–55. Kind P. (2007). Med Care. 45: 809–811.
Matza LS, Secnik K, Mannix S, Sallee FR. (2005). Pharmacoeconomics. 23: 777–790. Oppe M, Rabin R, & De Charro F. (2007). EQ-5D User Guide version 1.0. EuroQOL Group. Pickard AS, De Leon MC, Kohlmann T, Cella D, Rosenbloom S. (2007a). Med Care. 45: 259–263. Pickard AS, Kohlmann T, Janssen MF, Bonsel G, Rosenbloom S, Cella D. (2007b). Med Care. 45: 812–819. Polinder S, Meerding WJ, Toet H, Mulder S, Essink-Bot ML, Van Beeck EF. (2005). Pediatrics. 116: e810–817. Sach TH, Barton GR. (2007). Int J Pediatr Otorhinolaryngol. 71: 435–445. Stolk EA, Busschbach JJ, Vogels T. (2000). Qual Life Res. 9: 29–38. Szende A, Oppe M, Devlin N. (2007). EQ-5D Value Sets: Inventory, Comparative Review and User Guide. Springer, Berlin. Szende A, Williams A (ed.). (2004). Measuring SelfReported Population Health: An International Perspective Based on EQ-5D. SpringMed Publishing Ltd, Budapest, Hungary. Van Agt H, Bonsel G. (2005). In: Kind P, Brooks R, Rabin R (ed.) EQ-5D Concepts and Methods: A Developmental History. Springer, Berlin. Wille N, Rabens-Sieberer U, Bonsel G, Burstrom K, Cavrini G, Egmar AC, Greiner W, Gusi N, Herdmann M, Jelsma J, Kind P, Krabbe-Lugner A, Scalone L. (2006). Establishing definitions of the concepts included in CF-EQ-5D: a revision of the definition of EQ-5D concepts for adults. In: Badia X (ed.) Paper Presented at the 23rd Scientific Plenary Meeting of the EuroQOL Group. Health Economics & Outcomes Research IMS Health, Barcelona. Williams A. (2005). In: Kind P, Brooks R, Rabin R (eds.) EQ-5D Concepts and Methods. Springer, Dordrecht, Netherlands. Wolfs CA, Dirksen CD, Kessels A, Willems DC, Verhey FR, Severens JL. (2007). Health Qual Life Outcomes. 5: 33.
99
6 The University of Washington Quality of Life Scale S. N. Rogers . D. Lowe 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
2
Historical Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
3 3.1 3.2 3.3 3.4 3.5
Questionnaire Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 Version 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 Version 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104 Version 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105 Version 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105 Translations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
4 4.1 4.2 4.3 4.4
Scoring and Presenting the UW-QOLv4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105 Presenting Domain Scores . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106 Presenting Global Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110 Presenting the Importance Question . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111 Presentation of Composite Scores . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
5 5.1 5.2 5.3 5.4 5.5 5.6 5.7 5.8
Features of the UW-QOL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112 Feasible and Acceptable . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112 Face and Content Validity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113 Construct Validity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113 Responsiveness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114 Interpretability, Clinical Effect and Sample Size . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114 Stability of UWQOL for Longer-Term Survivors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114 Normative Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115 Composite Scores and Factor Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
6 Comparison of Domains with Other Questionnaires . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118 6.1 Clinically Meaningful Cut-Offs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123 7
Resume of Published Peer Review Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
8
Version 5 of the UW-QOL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
#
Springer Science+Business Media LLC 2010 (USA)
102
6
The University of Washington Quality of Life Scale
Abstract: The need for validated head and neck specific cancer health-related quality of life (HRQOL) questionnaires has long been recognised. One of the first to be published was the University of Washington quality of life scale (UW-QOL) in 1993. The current version 4 covers 12 domains - pain, appearance, activity, recreation, swallowing, chewing, speech, shoulder function, taste, saliva, mood and anxiety. Each question is scaled from 0 (worst) to 100 (best) according to the hierarchy of response. The UW-QOL also contains three global questions, one as a five-point Likert scale asking about HRQOL compared to the month before the cancer, and the other two as a six-point Likert scale, one asking about health-related and the other asking about overall quality of life during ‘the past 7 days’. It also has a question asking about the importance of domains and an option for free-text comment. The lack of any copyright agreements and its simplicity in scoring makes using the UWQOL questionnaire easy to use. It has been translated into numerous languages. The questionnaire has face and content validity and there is considerable evidence to support the construct validity of its domains. Our work indicates the UW-QOL to be sensitive to changes in patient characteristics and to changes over time. The identification of suitable cut-offs in domain scores adds to the interpretability of the UW-QOL in a routine clinic setting to help screen patients with particular problems. The UW-QOL displays stability in patients beyond one year from treatment and shows expected differences in regard to normative values. Factor analyses indicate that two composite scores – physical function and social function – are logical, each being a simple average of 6 domain scores. Further work is necessary to develop a better understanding of these composite scores, which have the potential of being used to assess clinical effect in treatment evaluation studies and to develop sample size calculations. List of Abbreviations: BDI, Beck Depression Inventory; CES-D, Centre for Epidemiology Studies Depression Scale; DAS-24, Derriford Appearance Scale; EORTC, European Organisation for Research and Treatment of Cancer; EQ-5D, EuroQolEQ-5D; FACT, Functional Assessment of Cancer Therapy; FEES, Fibreoptic Endoscopic Evaluation of Swallowing; HADS, Hospital Anxiety and Depression Scale; HRQOL, Health-Related Quality of Life; IMRT, Intensity Modulated Radiotherapy; LORQ, Liverpool Oraql Rehabilitation Questionnaire; MDADI, M.D.Anderson Dysphagia Inventory; NDII, Neck Dissection Impairment Index; SDQ, Shoulder Disability Questionnaire; SF-36, medical outcomes study 36-item Short-Form health survey; SIP, Sickness Impact Profile; SWAL-QOL, Swallowing Quality of Life measure; TOM, Therapy Outcome Measures; UWQOL, University of Washington Quality of Life scale; VAS, Visual Analogue Scale; VHI, Voice Handicap Index; V-RQOL, Voice Related Quality of Life; XeQOLS, Xerostomia- related Quality of Life Questionnaire
1
Introduction
The University of Washington quality of life (UWQOL) scale is one of the more commonly used questionnaires in head and neck cancer. This is probably in part due to its simplicity and suitablity for routine use in a busy clinical setting.This chapter will deal with aspects of historical background, questionnaire development (versions 1, 2, 3 and 4), translations, scoring and presentation (domain, global, importance and composite scores), feasibility and acceptability, face and content validity, construct validity, responsiveness, stability, normative data, composite scores and factor analysis, comparison of domains with other questionnaires, summary of its use as reported in peer review journals and finally a comment about possible
The University of Washington Quality of Life Scale
6
further modifications if there were to be a version 5. Validation of the UW-QOLv4 as a clinical outcomes measure is an ongoing process, as is finding the most suitable way of using it to enhance multidisciplinary team decision making and patient care.
2
Historical Background
The need for validated head and neck specific cancer health-related quality of life (HRQOL) questionnaires was recognised many years ago (Rogers et al., 2008). One of the first to be published was the UW-QOL in 1993 (Hassan and Weymuller, 1993) and it now holds an established place in the reporting of patient derived outcomes following head and neck cancer (Rogers et al., 2008). There is of course no ‘gold standard’ questionnaire and each has it own merits and weaknesses (Ringash and Bezjak, 2001, Rogers et al., 2008). Perhaps one of the most appealing features of the UW-QOL is its simplicity. In the original description, Hassan and Weymuller (1993) stated that the advantages of the head and neck questionnaire are that (1) it is brief and self-administered, (2) it is multifactorial, allowing sufficient detail to identify subtle change, (3) it provides questions specific to head and neck cancer, and (4) it allows no input from the health provider, thus reflecting the QOL as indicated by the patient. There are many barriers to HRQOL collection in a head and neck setting (Rogers et al., 2008) so the UW-QOL has clinical relevance as it easy for both patient and clinician to use. Its application in routine clinical practice can be facilitated by touchscreen computerised technology (Rogers et al., 2008). It is important to include HRQOL as an outcome parameter (Rogers et al., 2008) and also as a means of helping to identify patients doing badly who otherwise can easily go unrecognised (Rogers et al., 2008). The terminology and scoring is an important issue and the UW-QOL uses terms and cut-offs that are easily recognised and hence it gives clinically useful information suitable for healthcare professional, patient and carers alike (Rogers et al., 2008). The conciseness of the UW-QOL means that it can anchor a Unit’s evaluation of itself over time. Because the core questionnaire is brief it is possible to add other questionnaires as required without compromising response rates or causing the patient excessive questionnaire burden and fatigue. Additional validated measures such as for depression, xerostomia, swallowing or shoulder function can be asked concurrently to audit or test a particular issue.
3
Questionnaire Development
There have been four versions of the UW-QOL since it was first published in 1993. The modifications to the questionnaire are shown in > Table 6-1.
3.1
Version 1
The first version of the UW-QOL comprised of 9 domains (pain, appearance, activity, recreation, swallowing, chewing, speech, shoulder, employment). It was compared to two established equality of life evaluation tools, the Karnofsky scale and the Sickness Impact Profile (SIP) (Hassan and Weymuller, 1993). Using the SIP as a gold standard, the UW QOL scale demonstrated an average criterion validity of 0.849, whereas the Karnofsky average criterion validity was 0.826. The UW-QOL questionnaire scored >0.90 on reliability coefficients versus 0.80 for
103
104
6
The University of Washington Quality of Life Scale
. Table 6-1 Summary of development of the UW-QOL Domain
Version 1
Version 2
Version 3
Version 4
Pain
X
X
X
X
Appearance
X
X
X
X
Activity
X
X
X
X
Recreation
X
X
X
X
Swallowing
X
X
X
X
Chewing
X
X
X
X
Speech
X
X
X
X
Shoulder
X
X
X
X
Taste
X
X
Saliva
X
X
Mood
X
Anxiety
X
Employment Global QOL items
X
X X
X
X
Free text
X
X
X
Importance rating
X
X
X
the Karnofsky and 0.87 for the SIP scale. The UW QOL scale faired better than the Karnofsky and the SIP scale in detecting change (responsiveness). It was the face and construct validity that appealed to authors when looking for a head and neck measure at a time when there was a paucity of specific questionnaires available. Questionnaire content was also important as the UW-QOL included a question about shoulder function, missing on other questionnaires but important to a surgical head and neck oncologist. The lack of any copyright agreements and its simplicity in scoring made using the UW-QOL questionnaire quite straightforward. The authors thought that the questionnaire had potential to be a standard way of capturing a patient’s perspective in a busy routine clinical setting and that could be used in national audit.
3.2
Version 2
In version 2 (Deleyiannis et al., 1997) each of the 9 original domains was followed by an importance-rating scale. Also three new single item ‘quality of life’ questions were added (> Table 6-1). These amendments were useful as they brought in the concepts of importance and global measures of quality of life. Importance has been shown to be helpful as there is a wide variation between patients (Rogers et al., 2008). Both pre- and post-treatment there was a general lack of correlation between importance-rating and domain scores. At all time points, patients tended to rate speech, chewing and swallowing as more important than the other UW-QOL domains. The inclusion of importance with domain score gives a way of identifying patients with problems who might benefit from intervention. Prior to the addition of the three general quality of life questions it was shown that a composite score of UW-QOL domains
The University of Washington Quality of Life Scale
6
correlated well with global questions in other questionnaires (Rogers et al., 2008). Another new feature of the UW-QOL was the freetext element. Freetext allows the questionnaire to record other issues of importance to the patient and in so doing provides a better overall questionnaire experience for the patient. Freetext also gives the multidisciplinary team a better insight into the concerns of the patient and can be used to promote a holistic approach. We found that around 61% of patients make a comment at some time, predominantly head and neck (39%) and medical (35%) (Rogers et al., 2008), and about one-third of all questionnaires have some freetext added. The analysis of freetext has shown that the UW-QOL has patient approval, however there are one or two items that could be changed (as is discussed in section vii) such as the addition of too much saliva to the saliva domain.
3.3
Version 3
In version 3 (Weymuller et al., 2000, 2001) two new domains (taste, saliva) were added and the employment domain dropped (> Table 6-1). Rather than asking patients to rank the importance of each individual domain, version 3 asked patients to indicate which three domains had been most important to them in the last seven days. This resulted in a 10-item instrument (UW-QOL-R) with an overall internal consistency score of 0.85.
3.4
Version 4
Health-related quality of life refers to the physical, emotional, and social impact of diseases and their treatments on patients. Version 3 did not include an emotional domain and hence two new domains (anxiety and mood) styled in a similar fashion to the existing questionnaire were added (Rogers et al., 2008). The new domains correlated significantly with the emotional functioning domains from the EORTC C30 and with the pain and appearance domains of UW-QOL. There were also significant correlations between the ‘‘global quality of life’’ item and the two new domains.
3.5
Translations
The UW-QOL has been translated into several different languages (> Table 6-2) (Andrade et al., 2006, Lovell et al., 2005, Omoro et al., 2006, Thone et al., 2003, Vartanian et al., 2004, Vartanian et al., 2006b, Wang et al., 2002). Such papers also compare the UW-QOL with other validated measures, for example the Medical Outcomes Study 36-Item Short-Form Health Survey (SF-36) and a Hospital Anxiety and Depression Scale (HADS) (Vartanian et al., 2006b).
4
Scoring and Presenting the UW-QOLv4
The scoring of the UW-QOLv4 is based on the earlier scoring of the questionnaire. (Deleyiannis and Weymuller, 1996, Deleyiannis et al., 1997) and shown in > Table 6-3. The UW-QOL has domains and general questions based upon discrete ordinal responses. Scoring is scaled so that a score of 0 represents the worst possible response, and a score of 100 represents the best possible response. Scoring is scaled in equal stages from 0 to 100 to reflect
105
106
6
The University of Washington Quality of Life Scale
. Table 6-2 Translations of the UW-QOL (as currently known to the authors) America Asia and Africa Chinese (both simplified and traditional) Dutch English Europe French German Greek Hindi Italian Japanese Malay Marathi Norwegian Portuguese Spanish Swahili Turkish www.headandneckcancer.co.uk
the number of possible responses. Thus the pain domain has five possible responses which are scored as 0, 25, 50, 75 and 100. The global question asking about overall QOL has six possible responses, which are scored as 0, 20, 40, 60, 80 and 100. The UW-QOL also has a question asking about which three domain issues were the most important during the past seven days. A column for each domain should be created in the dataset with each column being scored either as ‘1’ if that domain is chosen as important, otherwise scored as ‘0’. We illustrate next how the UW-QOL data can be presented in a compact form suitable for peer review journals. The actual data used here comes from our use of the UW-QOL questionnaire since 1995, version 4 since 2000, by patients with oral/oro-pharyngeal SCC cancer whose primary treatment was by surgery with or without adjuvant radiotherapy. These data from 550 patients were selected so as to give the QOL response closest to 12 months after surgery, median 16 months, inter-quartile range 12–25 months.
4.1
Presenting Domain Scores
For each domain the > Table 6-4 gives the number of patients with each score, the mean scores, Standard Error (SE) of mean scores, and the percentage of patients selecting the best possible response (100). Though the data are quite clearly skewed towards the higher and better scores no notable ‘floor’ or ‘ceiling’ effects are observed (> Table 6-4). Note that the
The University of Washington Quality of Life Scale
6
. Table 6-3 Scoring of the UW-QOLv4 University of Washington Quality of Life Questionnaire (UW-QOL v4) This questionnaire asks about your health and quality of life over the past seven days. Please answer all of the questions by ticking one box for each question. 1. Pain. (Tick one box: ☐) ☐ I have no pain.
(100)
☐ There is mild pain not needing medication.
(75)
☐ I have moderate pain - requires regular medication (e.g. paracetamol).
(50)
☐ I have severe pain controlled only by prescription medicine (e.g. morphine).
(25)
☐ I have severe pain, not controlled by medication.
(0)
2. Appearance. (Tick one box: ☐) ☐ There is no change in my appearance. ☐ The change in my appearance is minor.
(100) (75)
☐ My appearance bothers me but I remain active.
(50)
☐ I feel significantly disfigured and limit my activities due to my appearance.
(25)
☐ I cannot be with people due to my appearance.
(0)
3. Activity. (Tick one box: ☐) ☐ I am as active as I have ever been.
(100)
☐ There are times when I can’t keep up my old pace, but not often.
(75)
☐ I am often tired and have slowed down my activities although I still get out
(50)
☐ I don’t go out because I don’t have the strength.
(25)
☐ I am usually in bed or chair and don’t leave home.
(0)
4. Recreation. (Tick one box: ☐) ☐ There are no limitations to recreation at home or away from home.
(100)
☐ There are a few things I can’t do but I still get out and enjoy life.
(75)
☐ There are many times when I wish I could get out more, but I’m not up to it.
(50)
☐ There are severe limitations to what I can do, mostly I stay at home and watch TV
(25)
☐ I can’t do anything enjoyable.
(0)
5. Swallowing. (Tick one box: ☐) ☐ I can swallow as well as ever.
(100)
☐ I cannot swallow certain solid foods.
(70)
☐ I can only swallow liquid food.
(30)
☐ I cannot swallow because it "goes down the wrong way" and chokes me.
(0)
6. Chewing. (Tick one box: ☐) ☐ I can chew as well as ever. ☐ I can eat soft solids but cannot chew some foods. ☐ I cannot even chew soft solids.
(100) (50) (0)
107
108
6
The University of Washington Quality of Life Scale
. Table 6-3 (continued) University of Washington Quality of Life Questionnaire (UW-QOL v4) 7. Speech. (Tick one box: ☐) ☐ My speech is the same as always.
(100)
☐ I have difficulty saying some words but I can be understood over the phone.
(70)
☐ Only my family and friends can understand me.
(30)
☐ I cannot be understood.
(0)
8. Shoulder. (Tick one box: ☐) ☐ I have no problem with my shoulder.
(100)
☐ My shoulder is stiff but it has not affected my activity or strength.
(70)
☐ Pain or weakness in my shoulder has caused me to change my work/hobbies.
(30)
☐ I cannot work or do my hobbies due to problems with my shoulder.
(0)
9. Taste. (Tick one box: ☐) ☐ I can taste food normally.
(100)
☐ I can taste most foods normally.
(70)
☐ I can taste some foods.
(30)
☐ I cannot taste any foods.
(0)
10. Saliva. (Tick one box: ☐) ☐ My saliva is of normal consistency.
(100)
☐ I have less saliva than normal, but it is enough.
(70)
☐ I have too little saliva.
(30)
☐ I have no saliva.
(0)
11. Mood. (Tick one box: ☐) ☐ My mood is excellent and unaffected by my cancer. ☐ My mood is generally good and only occasionally affected by my cancer.
(100) (75)
☐ I am neither in a good mood nor depressed about my cancer.
(50)
☐ I am somewhat depressed about my cancer.
(25)
☐ I am extremely depressed about my cancer.
(0)
12. Anxiety. (Tick one box: ☐) ☐ I am not anxious about my cancer.
(100)
☐ I am a little anxious about my cancer.
(70)
☐ I am anxious about my cancer.
(30)
☐ I am very anxious about my cancer. Which issues have been the most important to you during the past 7 days? Tick ☐ up to 3 boxes. ☐ Pain ☐ Swallowing ☐ Taste ☐ Appearance ☐ Chewing ☐ Saliva
(0)
The University of Washington Quality of Life Scale
. Table 6-3 (continued)
6
University of Washington Quality of Life Questionnaire (UW-QOL v4) ☐ Activity ☐ Speech ☐ Mood ☐ Recreation ☐ Shoulder ☐ Anxiety General Questions Compared to the month before you developed cancer, how would you rate your health-related quality of life? (Tick one box: ☐) ☐ Much better
(100)
☐ Somewhat better
(75)
☐ About the same
(50)
☐ Somewhat worse
(25)
☐ Much worse
(0)
In general, would you say your health-related quality of life during the past 7 days has been: (Tick one box: ☐) ☐ Outstanding
(100)
☐ Very good
(80)
☐ Good
(60)
☐ Fair
(40)
☐ Poor
(20)
☐ Very poor
(0)
Overall quality of life includes not only physical and mental health, but also many other factors, such as family, friends, spirituality, or personal leisure activities that are important to your enjoyment of life. Considering everything in your life that contributes to your personal well-being, rate your overall quality of life during the past 7 days. (Tick one box: ☐) ☐ Outstanding
(100)
☐ Very good
(80)
☐ Good
(60)
☐ Fair
(40)
☐ Poor
(20)
☐ Very poor
(0)
Please describe any other issues (medical or nonmedical) that are important to your quality of life and have not been adequately addressed by our questions (you may attach additional sheets if needed)
standard deviation measures the scatter of raw data scores symmetrically about a mean and is less useful with ordered categorical data with few categories. Standard error measures the precision of the mean, and mean +/2 SE is the approximate 95% confidence interval for the mean. Having few categories renders the median to be an insensitive measure and we therefore do not recommend it to summarise UW-QOL domain scores.
109
110
6 4.2
The University of Washington Quality of Life Scale
Presenting Global Questions
The UW-QOL asks three global questions, one about how patients feel relative to before they developed their cancer, one about their health-related QOL and one about their overall QOL. These can also be scaled from 0 to 100 to enable ease of presentation of all key results using the same scale. No notable ‘floor’ or ‘ceiling’ effects can be observed (> Table 6-5). . Table 6-4 Presentation of UW-QOL domain results UW-QoL Scores UW-QOL
N
0
25
Pain
545
6
31
Appearance
545
3
31
Activity
545
16
25 47
30
50
70
75
100 Mean SE % Best Scores
107
131
270
79
105
260
146
74
1 27
185
154
165
70
1 30
210
164
72
1 30
267
77
1 49
191
60
1 35
116
1 50
Recreation
547
10
Swallowing
544
25
Chewing
548
81
Speech
538
3
41
290
204
78
1 38
Shoulder
531
32
55
103
341
81
1 64
Taste*
365
20
69
96
180
73
2 49
Saliva*
360
24
182
73
2 51
Mood*
357
7
157
76
1 44
Anxiety*
354
18
148
74
2 42
58
194 276
66 45
88 28
120
46
142
*These were not in the earliest versions of the UW-QOL but were added later, hence fewer patients
. Table 6-5 Presentation of UW-QOL global results UW-QOL Scores UW-QOL
N
0 20 25 40 50
A. Health-related QOL 343 25 compared to month before had cancer
58
% Best 60 75 80 100 Mean SE Scores*
139
55
66
56
2
76
B. Health-related QOL during the past 7 days
306
8 17
63
115
87 16
60
1
71
C. Overall QOL during the past 7 days
306
2 14
64
105
105 16
63
1
74
*BEST SCORES: A: % scoring 50, 75 or 100; B & C: % scoring 60, 80 or 100 KEY to ratings: A: (0) Much worse (25) Somewhat worse (50) About the same (75) Somewhat better (100) Much better B: (0) V Poor (20) Poor (40) Fair (60) Good (80) V Good (100) Outstanding C: (0) V Poor (20) Poor (40) Fair (60) Good (80) V Good (100) Outstanding
The University of Washington Quality of Life Scale
4.3
6
Presenting the Importance Question
The UW-QOL asks about which three domain issues were the most important during the past seven days. Patients can choose up to three domains. Very occasionally patients may choose more than three and when this occurs we suggest you score all those they have chosen as ‘1’ and all those not chosen as ‘0’. Results can be presented as % of patients choosing each domain. The domains can also be ranked in order. The five main concerns chosen at about 1–2 years after surgery were saliva, swallowing, chewing, speech and appearance. These patients chose a mean of 2.3 domains (> Table 6-6).
4.4
Presentation of Composite Scores
An overall composite score (a simple average of UW-QOL domain scores) has not been recommended for use since version 4 because as more domains have been added it became less interpretable. However, more recent work applying factor analysis to the QOL data (see later section), has suggested two composite scores, one for ‘physical function’ and another for ‘social–emotional function’. The physical function score is computed as the simple average of 6 domain scores – those of chewing, swallowing, speech, taste, saliva and appearance. The social–emotional function is also computed as the simple average of 6 domain scores – those of anxiety, mood, pain, activity, recreation and shoulder function. Missing data for the UW-QOL is rare but to accommodate this it is suggested that the composite scores only be computed if there are at least four component domain scores available. Further details of this work will be given later. This section explains how to score and present the results. These composite scores can be regarded as numerical for the purpose of presentation. The overall median (Inter-Quartile Range) physical function score for patients at 1–2 years was 74
. Table 6-6 Presentation of UW-QOL importance UW-QOL Pain
N of patients
N of patients choosing the domain
% of patients choosing the domain
Rank order
358
46
13
10
Appearance
358
76
21
5
Activity
358
67
19
6
Recreation
358
32
9
12
Swallowing
358
94
26
2
Chewing
358
89
25
3
Speech
357
87
24
4
Shoulder
358
37
10
11
Taste
358
56
16
9
Saliva
358
102
28
1
Mood
358
66
18
7
Anxiety
358
62
17
8
111
112
6
The University of Washington Quality of Life Scale
. Figure 6-1 Box-plot representation of the UW-QOL physical function composite score
(59 to 91) whilst that for social function was 80 (61 to 94). No notable ‘floor’ or ‘ceiling’ effects can be observed. A box-plot graphical representation is also appropriate, as illustrated for physical function below (> Figure 6-1). The patients are grouped according to key clinical characteristics. Note that each box incorporates 50% of the data (i.e., that between the 25th and the 75th percentiles) and is therefore the graphical equivalent of the Inter-Quartile Range. One quarter of scores will lie at the bottom of or below the box and one quarter will lie at the top of or above the box. The thick black line within each box represents the median score. Individual patient ‘outlier’ scores are marked separately with a ‘circle’. The best set of physical function scores were for those patients with less advanced oral cancer tumours and not requiring free-flap surgery nor adjuvant radiotherapy (> Figure 6-1).
5
Features of the UW-QOL
5.1
Feasible and Acceptable
Instruments for routine use in clinical care need to be brief and simple to use. Too brief and they lack content and precision; too complex and they reduce response rates and require programming skills to process. The ideal is for a well-defined questionnaire that can be applied
The University of Washington Quality of Life Scale
6
and interpreted easily and for which there is no great burden of labour. Our experience of the UW-QOL is that it is a practical, feasible and acceptable questionnaire, suitable for routine use as it takes at most ten minutes for patients to complete, and the data processing is relatively straightforward. Postal response rates are typically two-thirds to three-quarters of those targeted and missing data rates are very low (1–3% amongst domains). For each domain patients are required to choose one from only a few discrete hierarchical options and though the distribution of responses is typically skewed there are no notable ‘floor’ or ‘ceiling’ effects (seen if over 95% of responses go to one option only).
5.2
Face and Content Validity
Many of our published papers provide evidence in supporting the validity of the domains and global questions of the UW-QOL. A valid instrument will measure what it is supposed to measure and will not measure what it is not supposed to measure. It is our belief that the UW-QOL version 4 has good ‘face’ and ‘content’ validity. ‘Face validity’ is a loose basic concept and indicates that the questions have face value to the profession and to the patient. Content validity is slightly more precise and indicates that the questions adequately cover the breadth or content of what is meant to be included. It can apply to the breadth amongst domains and of the detail within a domain. Our belief arises from verbal exchanges between patient and consultant, from focus group work, from administering patients to respond via touch-screen and from hundreds of written responses to the free-text question on the UWQOL that asks patients to describe any other important issues not covered by the main questions.
5.3
Construct Validity
Construct validity is particularly relevant because there are no gold standards against which to compare. Therefore we became involved in various studies collecting concurrent evidence from other more detailed and established questionnaires to support the inference that the UW-QOL has meaning. We have been able to demonstrate positive correlations where expected between UW-QOL domains conceptually linked with the concurrent measures and have demonstrated lesser correlations involving the other UW-QOL domains. Where concurrent measures have established definitions of ‘case-ness’ (Rogers et al., 2008) we have assessed sensitivity and the positive predictive value of the UW-QOL. In the absence of caseness in concurrent measures we have assessed the extent to which the UW-QOL domains identify the more severely affected patients. We have also stratified our analysis of UW-QOL domain and concurrent data by clinical characteristics of patients to see if the same clinical differences exist for both measures. On the whole we have been surprised at the ability of the UW-QOL domains, given the limited number of distinctions made within each scale, to reflect the key essence of the more detailed questionnaires. We accept that the UW-QOL has less precision than the alternative battery of detailed questionnaires but believe that as it seems able to flag a high proportion of problem cases, and is simple and quick to administer, it is well suited as a quick screening tool in routine clinics, as well as having potential for use in clinical research.
113
114
6 5.4
The University of Washington Quality of Life Scale
Responsiveness
To use a QOL scale as an outcome variable to evaluate treatment efficacy requires evidence that it can react or respond to real changes in a patient’s condition. Our work on the changes in the UW-QOL after treatment, and between different clinical groups of patients, adds to the evidence supporting the responsiveness or sensitivity of the UW-QOL to changes in patient characteristics and to changes over time.
5.5
Interpretability, Clinical Effect and Sample Size
Researchers have quite a good handle on what constitutes a meaningful change in biochemical measures and other well-used scales but lack experience in using QOL scales in everyday practice. Interpretability is to do with clinical meaning and if we want to use the UW-QOL scale as an outcome measure we need to know what constitutes a small but relevant clinical change over time. One possible way to interpret and compare change is through what are known as effect sizes (Kazis et al., 1989) with an effect size in this context being the mean change in score between two time points divided by the standard deviation at the first time point, an effect size of 0.20 being ‘small’, 0.50 ‘moderate’ and 0.80 ‘large’. However the problem with UW-QOL domains and global scales is that they are not numerical, have at most six discrete options and a skewed response. On the other hand the new composite ‘Physical Function’ and ‘Social–emotional Function’ scores are numerical and not particularly skewed and these offer the prospect of being used to assess clinical effect in treatment evaluation studies and being used to develop sample size calculations. More validation work is necessary to establish these composite scores. An alternative to effect size is to categorise domain scores into binary measures from which the % with the best score of 100, or the % below a certain cut-off score can be derived. We have worked with the UW-QOL to identify suitable cut-offs in domain scores to help find patients with more serious problems (Rogers et al., 2008) but there is a lack of evidence as to what might constitute a meaningful change in these measures. Having suitable cut-off values does however add to the interpretability of the UW-QOL in a routine clinic setting in helping to screen patients with particular problems.
5.6
Stability of UWQOL for Longer-Term Survivors
Our longitudinal work indicates that changes in patient condition are most often seen within the first few months after treatment and our cross-sectional work suggests stability after 1 year. However, we have not yet investigated within-patient change using the UW-QOL beyond one year. The analyses that follow are akin to ‘test-retest’ reliability and are unique to this chapter and they describe the within-patient agreement between one annual survey and the next. The data comes from version 4 of the UW-QOL questionnaire in annual surveys since 2000, from the cohort of MFU patients treated since 1992. There were 79 patients responding in both 2000 and 2001, 63 in 2001 & 2002, 115 in 2002 & 2003, 150 in 2003 & 2004, 169 in 2004 & 2005, 173 in 2005 & 2006 and 202 in 2006 & 2007. The median time from surgery increased from 24 months in 2000 to 53 months in 2006. In all 951 pairs of questionnaires were analysed for agreement, 89% (851) being returned by patients at least 12 months after surgery. Agreement was measured by the kappa coefficient weighted by applying standard
The University of Washington Quality of Life Scale
6
weights according to the number of categories in error. Kappa values (k) of 0.41–0.60 indicate ‘moderate’ agreement, 0.61–0.80 ‘good’ agreement and over 0.80 ‘very good’ agreement. In summary, for survivors beyond one year (median 50 months) the 12 UW-QOL domains and 2 global QOL items show moderate to good agreement from one annual survey to the next (> Table 6-7). Domain importance was considerably less stable. Strongest agreement was for swallowing (k = 0.73), chewing (k = 0.70) and saliva (k = 0.70) scores. The overall kappa statistics are confirmed by narrow ranges from each separate set of paired surveys (2000 & 2001, 2001 & 2002 etc). There was moderate agreement for the global questions about healthrelated (k = 0.50) and overall (k = 0.51) QOL. Less agreement (range for overall k: 0.26–0.49) was seen for stating a domain issue as important and for the global question comparing to before cancer had developed. There was also less agreement, as expected, across the UW-QOL for patients first responding before 12 months (median 7 months). There was no evidence found (from McNemars or McNemar-Bowker’s tests as appropriate at p < 0.01) for any score to show a systematic improvement or deterioration between surveys.
5.7
Normative Data
Age and gender reference data for the UW0QOLv4 have been collected from 372 patients attending general dental practices (Rogers et al., 2008). Oral and oropharyngeal patients compared to the normative data showed that at baseline the key differences were anxiety, pain, swallowing, chewing, and mood. At one year there were big differences in all domains with deterioration in the oral cancer group. The difference was least notable in pain, shoulder, mood and anxiety. Also the UW-QOL was compared with other questionnaires with established reference data such as the Medical Outcomes Short Form 36 (Rogers et al., 2008) and the EuroQol EQ-5D (Rogers et al., 2008). Compared to the UK population survey of 1993 using the EQ-5D, oral and oropharyngeal patients under 60 years of age fared significantly worse than expected for their age. The mean (SE) VAS values of 74 (1) were at levels expected for the most elderly (80 + years) of UK residents.
5.8
Composite Scores and Factor Analysis
The multi-dimensionality of the UW-QOL means that to obtain an overall feel of the UWQOL profile we have to trade off results from one domain against another. Some researchers would prefer a single global index, though the difficulty with this is the loss of sensitivity to subtle and perhaps important changes. There is no single clinical, physiological or biochemical measure of health and no logical reason for imagining that any single value could possibly summarise health-related QOL. Nevertheless in our early publications we did report a composite score that was a simple average of domain scores. The positive correlation amongst domain scores helped to justify the use of this composite score and to help establish it we published a paper (Rogers et al., 2008) that demonstrated that it was able to discriminate in a very noticeable way across all EORTC and SF-36 domains except for emotional function (EORTC) and mental health (SF-36). As the UW-QOL has developed to include more domains, including mood and anxiety, we moved away from reporting an overall score believing it to be less interpretable. Unique to this chapter we have revisited our pool of 2728 UW-QOL responses since 1995 to use factor analysis to investigate whether composite scores may be possible. These data
115
116
6
The University of Washington Quality of Life Scale
. Table 6-7 Within-patient agreement in UW-QOL responses between one annual survey and the next
Patients responding at least 12 months after surgery
Patients first responding Figure 6-2) and social (> Figure 6-3) function scores over time, from
. Figure 6-2 Variation in UW-QOL physical function composite scores over time
117
118
6
The University of Washington Quality of Life Scale
. Figure 6-3 Variation in UW-QOL social–emotional function composite scores over time
before operation up to seven years after operation, the time frames being chosen to minimise the possibility of multiple responses per patient within each time frame. Each bar includes a different patient mix. The data clearly show the effect of treatment on physical function, with similar levels of deficit reported from about one year. The effect of treatment on social– emotional function is less pronounced and levels of function beyond 12 months are comparable to before the operation. These observations do of course mask the more subtle changes that can be observed from a full profile of results and our conclusion is that these physical and social–emotional function scores have a part to play in the presentation of results of the UW-QOL but in addition to and not instead of the full profile of domain results. Clearly we need to do further work to develop a better understanding of these composite scores and we expect to publish further clinical validation work in the near future.
6
Comparison of Domains with Other Questionnaires
Initially the UW-QOL was compared with the Medical Outcomes Short Form 36. (Rogers et al., 2008). By using the SF36 it was possible to compare the impact of surgery for oral cancer on quality of life to normative reference data and hence put the disease and the affect of treatment into a broader context. The same year Rogers et al. published a paper comparing the
The University of Washington Quality of Life Scale
6
UW-QOL to a well-validated European general cancer and head and neck specific cancer questionnaire, EORTC QOQ-C33 and EORTC Head and Neck 35. Also (D’Antonio et al., 1996) compared the UW-QOL with the Functional Assessment of Cancer Therapy (FACT). More recently as head and neck cancer function specific questionnaires have emerged the individual domains of the UW-QOLv4 have been compared with these (> Table 6-8). Correlations have been strong. > Figures 6-4–6-7 illustrate the strong correlations seen between saliva and xerostomia scale (XeQOLS), shoulder function and Neck Dissection Index (NDII), appearance and Derriford scale (DAS-24), and between overall QOL and the EQ-5D visual analogue scale. Activity and recreation. The physical domain is common to many HRQOL questionnaires. The UW-QOL activity and recreation domains have been compared to similar domains in several other questionnaires such as the SF-36 (Rogers et al., 2008) SF-12 (Vilaseca et al., 2006), EQ-5D (Rogers et al., 2008) On each occasion a strong relationship existed between UW-QOL and these other measures in the appropriate domains. Rogers et al showed that there are strong correlations between the five dimensional classification of the EQ-5D (mobility, self-care, usual activities, pain/discomfort and anxiety/depression) and appropriate domains in the UW-QOLv4 (pain, activity, recreation, anxiety and mood). Appearance. The appearance domain of the UW-QOLv4 has been compared with the Derriford Appearance Scale (DAS24) (Rogers et al., 2008). There was strong correlation between the UWQOL appearance domain and the DAS24 score (> Figure 6-6). A score of Figure 6-7). As with the EQ-5D VAS patient age was not correlated with either the health-related UW-QOL item or with the overall UW-QOL item. The strongest correlations between clinical demographic factors and both global items were regarding tumour size, type of surgery and radiotherapy, again as found with the EQ-5D VAS. Pain. Pain is a frequent component to HRQOL evaluation. The UW-QOL pain domain has been shown to have strong correlation with the pain domain in other measures such as SF-36 (Rogers et al., 2008), SF-12 (Vilaseca et al., 2006), EQ-5D (Rogers et al., 2008), LORQ (Rogers et al., 2008), EORTC C30 and H&N35 (Rogers et al., 2008). Saliva. The UW-QOL saliva domain has been compared with the Xerostomia- Related Quality of Life Questionnaire (XeQOLS). This work, yet to be published, showed a strong correlation between the UWQOL saliva domain and the XeQOLS (> Figure 6-4). The strong association was present across all of the four attributes of the XeQOLS namely physical, personal/psychological, pain/discomfort, and social function. Shoulder. There have been several studies specifically reporting shoulder dysfunction using the UW-QOL both from a clinician-rated objective assessment and comparison with other validated questionnaires (Rogers et al., 2008) Although the UW-QOL shoulder domain is limited to one of four responses, when compared with the neck dissection impairment index (NDII), and the shoulder disability questionnaire (SDQ) it performed very well (> Figure 6-5). Speech. Speech is a key outcome following head and neck cancer. Radford and co-workers (Rogers et al., 2008) showed that there was significant correlation between UW-QOL speech and Therapy Outcome Measures (TOM) laryngectomy, voice, phonology and dysarthria subscales. Thomas et al (Rogers et al., 2008) compared the speech domain of the UW-QOL with the Voice Handicap Index (VHI) and Voice related Quality of Life (VRQOL). In addition the scores were compared with objective testing; GRBAS rating, speech intelligibility and dysarthria rating. There were strong correlations and clear demarcations between UW-QOL scores of 30, 70 and 100 on the voice questionnaires. Swallowing. Swallowing is a key outcome parameter. Radford and co-workers (Rogers et al., 2008) showed correlations between UW-QOL swallowing and Therapy Outcome Measures (TOM) dysphagia. It has also been compared with the M. D. Anderson Dysphagia Inventory (MDADI) and the SWALQOL together with an objective evaluation (Fibreoptic Endoscopic Evaluation of Swallowing- FEES) (Rogers et al., 2008) with strong correlations reported between the UW-QOL swallowing domain and the other measures. Percutaneous Endoscopic Gastrostomy feeding (PEG) has an important role in providing nutritional
The University of Washington Quality of Life Scale
6
support in selected patients. Rogers et al., 2008 using a PEG specific questionnaire showed that the UW-QOL swallowing domain strongly related to the continued need for a PEG tube in long-term survivors. Taste. The taste domain of the UW-QOL has been compared favourably with the EORTC H&N taste question (Rogers et al., 2008).
6.1
Clinically Meaningful Cut-Offs
The comparison of the UW-QOL domains with other specific domain questionnaires has allowed us to bring forward the concept that the UW-QOLv4 is an appropriate screening tool in a busy clinical setting. It has been possible to postulate cut-offs that identify clinical distress and could trigger intervention (Rogers et al., 2008). The algorithm is based on the UW-QOL score and whether the patient considers this as important. The > Table 6-9 shows what we consider as our trigger criteria for each domain and also shows considerable variation in the percentage of patients selected at around 2 years from surgery according to key clinical patient characteristics.
7
Resume of Published Peer Review Studies
There is a lot of information concerning quality of life outcomes following head and neck cancer Rogers et al., 2008). The UW-QOL has help to inform clinicians and healthcare professions on the multidisciplinary team regarding patient based outcomes and has helped shape treatment protocols. In the section below papers from the literature that have reported the UW-QOL to help inform outcome following treatment and evaluate treatment will be briefly summarised. This will give an indication of the range of clinical settings the questionnaire has been used in. Access. The UW-QOL has been used to help evaluate the best way of getting to a tumour to facilitate as removal. In a comparison of aesthetic, functional and patient subjective outcomes following lip-split mandibulotomy and mandibular lingual releasing access procedures the lipsplit mandibulotomy group reported significantly better speech, swallowing and chewing (Rogers et al., 2008). Base of tongue. Base of tongue cancer is a site that tends to confers a lot of problems in terms of function and HRQOL following treatment. Using the UW-QOL as an outcome measure surgical resection can offer good functional and overall QOL results for advanced tumours when combined with reconstruction (Winter et al., 2004). Chemoradiation. Chemoradiation (organ preservation) has recently become a strategy of choice for certain H&N tumours. The UW-QOL has been used to report HRQOL following chemoradiation and compare ‘organ preservation’ to surgery (LoTempio et al., 2005, Mowry et al., 2006a, b). Comorbidity. No significant association was found between comorbidity and pretreatment UW-QOL scores, (Gourin et al., 2005) although it has been reported that American Society of Anesthesiologists’ (ASA), which is often recorded as part of preoperative assessment, reflects both survival and UW-QOL scores (Rogers et al., 2008). Coping. The UW-QOL has been used in studies exploring issues such as coping style, fighting spirit, level of social support and satisfaction with that support (Hassanein et al., 2001, Hassanein et al., 2005) and in addition personality (Fang et al., 2004).
123
0
0, 30
0, 30 + IMP
0, 30 + IMP
0, 30 + IMP
0, 25, 50 + IMP
0 or 30
Chewing
Speech
Shoulder
Taste
Saliva
Mood
Anxiety
15
13
7
6
8
4
4
4
17
51
9
13
4
6
17
7
4
8
5
15
11
16
21
64
20
22
32
14
15
10
12
10
3
8
10
5
Early Oral Flap RT % of N ¼ 42
17
52
14
9
9
9
9
11
11
9
20
15
15
11
Late Oral Flap No RT % of N = 46
This is the trigger criteria preferred by the authors For example with pain a scores of 0 alone, or 25 alone, or 50 plus the patient marking this as (IMP) would trigger the algorithm Each patient is represented only the once; the closest to 24 months (but after 15months) was selected for analysis
a
0, 30
Swallowing
5
13
0, 25, 50 + IMP
Recreation
9
2
42
0, 25, 50 + IMP
Activity
Three or more domains triggered
0, 25, 50 + IMP
Appearance
8
One or more domains triggered
0, 25, 50 + IMP
Pain
Trigger criteriaa
Early Oral Flap No RT % of N ¼ 82
42
81
24
20
30
30
15
24
39
35
16
18
29
20
Late Oral Flap RT % of N = 57
35
70
22
11
37
13
11
11
24
35
13
17
13
13
Oropharynx % of N = 46
6
Early Oral No flap No RT % of N ¼ 101
. Table 6-9 Trigger rates for patients at around 2 years from surgery (median 29, inter-quartile range23–45 months), by main clinical characteristics
124 The University of Washington Quality of Life Scale
The University of Washington Quality of Life Scale
6
Deprivation. The issue of deprivation (Rogers et al., 2008) and socioeconomic factors (Vartanian et al., 2006a) have been evaluated in the context of the UW-QOL. Deprivation is associated with poorer scores and additional rehabilitation social support should be considered for this vunerable group. Donor site and free tissue transfer. Parts of the body are used to reconstuct the defect following the removal of the cancer. The radial forearm free flap is the work horse in reconstruction. The outcome of this donor site has been assessed (Smith et al., 2006). The UW-QOL was used to help assess function in a paper comparing two bone flaps, the fibula (leg) and deep circumflex laic (hip) (Rogers et al., 2008). Markkanen-Leppanen et al., 2006 used the UW-QOL to report outcome following surgery for large carcinoma in the oral cavity, oral pharynx, or hypopharynx with free-flap surgery with or without radiotherapy. Dysphagia (difficulty swallowing). Dysphagia is a significant morbidity of head-and-neck cancer treatment, and the severity of dysphagia correlated with a compromised QOL as reflected in the UW-QOL (Nguyen et al., 2005). Papers exploring the association with swallowing specific questionnaires and the swallowing domain of the UW-QOL also reflect the importance of swallowing as an outcome measure. Laryngeal cancer. Loughran et al. (2005) has used the UW-QOL to compare outcomes and voice following endoscopic resection or radiotherapy for early glottic cancer. Campbell et al. (2004) used the UW-QOL in an evaluation of swallowing function and weight change and HRQOL. The UW-QOL has also been used to report HRQOL following total laryngectomy (Kazi et al., 2007, Vilaseca et al., 2006, Zhang et al., 2002), following partical laryngectomy (Vigili et al., 2002) and to compare partial laryngectomy to total laryngectomy (Wang et al., 2002). Length of hospital admission. Length of stay is potentially a useful indicator of HRQOL as measured by the UW-QOL because it is linked by tumour size, however, the relationship is confounded by age, which tends to influence length of stay more than health related quality of life (Rogers et al., 2008). Long-term outcomes. As cure rates for head and neck cancer improve so the HRQOL of survivors is an important consideration. There are at least three papers that have specifically explored this area using the UW-QOL (Campbell et al., 2000, Rogers et al., 2008, Vartanian et al., 2004). Mandibular (Lower jaw) resection. Mandibular resection has an important influence on HRQOL. The UW-QOL has been used to help quantify this from the patient perspective (Petruzzelli et al., 2003, Rogers et al., 2008, Young et al., 2007). Nasopharyngeal (back of the nose). The UW-QOL has been used to assess outcome in nasopharyngeal cancer (Lovell et al., 2005, Talmi et al., 2002). Dry mouth, chewing, and ear problems were of major concern with the majority of patients. Neck dissection (clearance of the drainage lymph glands). Neck dissection is a common surgical procedure in the management of head and neck cancer. The UW-QOL has helped put the function deficit into a patient perspective and perhaps has helped lead to a less radical approach and hence a better HRQOL outcome following this operation (Kuntz and Weymuller 1999, Rogers et al., 2008). Oral and Oropharyngeal. There has been a considerable amount of interest in the outcome of oral and oropharyngeal cancer using the UW-QOL as a predictor of outcome and for comparing surgical techniques (Klozar et al., 2001, Rogers et al., 2008).
125
126
6
The University of Washington Quality of Life Scale
Oral rehabilitation. Oral rehabilitation is an important aspect of maintaining HRQOL. The UW-QOL has been used to explore this area (Rogers et al., 2008). Palliative surgery. Smucler and Mazanek (2004) reported the effect of the combination of laser excision and interstitial hyperthermia in palliative therapy of head and neck tumours in the advanced stage of the disease. Parotidectomy. The quality of life following parotidectomy for malignant and benign disease has been published using the UW-QOL (Nitzan et al., 2004). Photodynamic therapy. The UW-QOL has helped to show that patients with advanced cancer of the head and neck, who have exhausted other treatment options, can achieve significant clinical benefit and improvement in quality of life by using this laser based treatment (D’Cruz et al., 2004). Prosthetic voice restoration. Speech after laryngectomy is a key area of HRQOL outcome and the UW-QOL has helped to inform practice (Gerwin and Culton 2005, Kazi et al., 2006). Radiotherapy. There have been recent innovations aimed at reducing the radiation dose to the parotid glands. The technique is called intensity modulated radiotherapy (IMRT) and has been shown to reduce the side effect of dry mouth (Scrimger et al., 2007). Submandibular salivary gland transfer. Submandibular salivary gland transfer has a potential role of in reducing or preventing dry mouth (xerostomia) in selected head and neck patients receiving chemoradiotherapy (Al-Qahtani et al., 2006, Jha et al., 2000, Jha et al., 2003). Trauma (facial). Sen and co-workers (Rogers et al., 2008) used a modification of the UW-QOL to assess outcome following maxillofacial trauma.
8
Version 5 of the UW-QOL
It seems appropriate to end this chapter by posing the question, is there a need for a version 5 of the UW-QOL? There is a natural reticence to modify a questionnaire that was first published in 1993 with the most recent version (the addition of two extra domains in version 4) published more than five years ago. Changing the questionnaire means that its reporting is difficult to compare. There is no single pressing concern that argues directly for a version 5. Possible changes to existing domains would be to consider building in the fact that some patients report they have too much saliva, and also to add extra precision to the chewing domain by increasing the number of responses from the current three levels. One additional domain that might be added is that of intimacy. What could also be built in is an indication of how much of the domain responses are related to comorbidity and how much to the cancer. This apportioning may be impossible for the patient to judge in which case it might help just to know whether or not the response was entirely due to something else other than the cancer.
Summary Points UW-QOL was first published in 1993 It has undergone three revisions with UW-QOLv4 being published in 2002
The University of Washington Quality of Life Scale
6
It is a simple questionnaire suitable for routine clinical practice It has been used in many different head and neck cohorts It is one of the commonest reported validated head and neck questionnaires It does not have a copyright It has been translated into numerous languages When compared to other specific functional questionnaires the single items correlate It is an ideal screening tool in routine head and neck cancer practice
References Al-Qahtani K, Hier MP, Sultanum K, Black MJ. (2006). Oral Surg Oral Med Oral Path Oral Rad Endo. 101: 753–756. Andrade FP, Antunes JL, Durazzo MD. (2006). Braz Oral Res. 20: 290–296. Campbell BH, Marbella A, Layde PM. (2000). Laryngology. 110: 895–906. Campbell BH, Spinelli K, Marbella AM, Myers KB, Kuhn JC, Layde PM. (2004). Otolaryngology. 130: 1100–1103. D’Antonio LL, Zimmerman GJ, Cella DF, Long SA. (1996). Otolaryngol Head Neck Surg. 122: 482–487. D’Antonio LL, Long SA, Zimmerman GJ, Peterman AH, Petti GH, Chonkich GD. (1998). Laryngology. 108: 806–811. D’Cruz AK, Robinson MH, Biel MA. (2004). Head Neck. 26: 232–240. Deleyiannis FW-B Weymuller EA. (1996). Quality of life in patients with head and neck cancer In: Myers EN, Suen JY (ed.) Cancer of the Head and Neck, 3rd ed. Saunders Company, pp. 904–916. Deleyiannis FW-B, Weymuller EA, Coltrera MD. (1997). Head Neck 19: 466–473. Fang WQ, Chen XY, Guan CH, Chen JF, Wang JB. (2004). Chin J Otolaryngol. 39(4): 227–231. Gerwin JM, Culton GL. (2005). Otolaryngol Head Neck Surg. 133(5): 685–688. Gourin CG, McAfee WJ, Neyman KM, Howington JW, Podolsky RH, Terris DJ. (2005). Laryngology 115(8): 1371–1375. Hassan SJ, Weymuller EA Jr. (1993). Head Neck. 15(6): 485–496. Hassanein KA, Musgrove BT, Bradbury E. (2001). Brit J Oral Maxfax. 39(5): 340–345. Hassanein KA, Musgrove BT, Bradbury E. (2005). J Cranio Maxfax Surg. 33(6): 404–409. Jha N, Seikaly H, McGaw T, Coulter L. (2000). Int J Rad Onc. 46(1): 7–11. Jha N, Seikaly H, Harris J, Williams D, Liu R, McGaw T, Hofmann H, Robinson D, Hanson J, Barnaby P. (2003). Rad Onc 66(3): 283–289.
Kazis LE, Anderson JJ, Meehan RF. (1989). Med Care. 27: S178–189. Kazi R, Kiverniti E, Prasad V, Venkitaraman R, Nutting CM, Clarke P, Rhys-Evans P, Harrington KJ. (2006). Clin Otolaryng. 31: 511–517. Kazi R, De Cordova J, Kanagalingam J, Venkitaraman R, Nutting CM, Clarke P, Rhys-Evans P, Harrington KJ. (2007). J Otolaryngol 69(2): 100–106. Klozar J, Lischkeova B, Betka J. (2001). Eur Arch Otolaryngol. 258(10): 546–551. Kuntz AL, Weymuller EA Jr. (1999). Laryngology. 109(8): 1334–1338. LoTempio MM, Wang KH, Sadeghi A, Delacure MD, Juillard GF, Wang MB. (2005). Otolaryngol Head Neck Surg. 132(6): 948–953. Loughran S, Calder N, MacGregor FB, Carding P, MacKenzie K. (2005). Clin Otolaryngol. 30(1):42–47. Lovell SJ, Wong HB, Loh KS, Ngo RY, Wilson JA. (2005). Head Neck. 27(10): 864–872. Markkanen-Leppanen M, Makitie AA, Haapanen ML, Suominen E, Asko-Seljavaara S. (2006). Head Neck. 28(3): 210–216. Mowry SE, Ho A, Lotempio MM, Sadeghi A, Blackwell KE, Wang MB. (2006a). Laryngology. 116(9): 1589–1593. Mowry SE, LoTempio MM, Sadeghi A, Wang KH, Wang MB. (2006b). Otolaryngol Head Neck Surg. 135(4): 565–570. Nguyen NP, Frank C, Moltz CC, Vos P, Smith HJ, Karlsson U, Dutta S, Midyett A, Barloon J, Sallah S. (2005). Int J Rad Oncol Biol Phys. 61(3): 772–778. Nitzan D, Kronenberg J, Horowitz Z, Wolf M, Bedrin L, Chaushu G, Talmi YP. (2004). Plastic Recon Surg. 114(5): 1060–1067. Omoro SA, Fann JR, Weymuller EA, Macharia IM, Yueh B. (2006). Int J Psychiatr Med. 36(3): 367–381. Petruzzelli GJ, Knight FK, Vandevender D, Clark JI, Emami B. (2003). Otolaryng Head Neck Surg. 129(6): 713–719. Ringash J, Bezjak A. (2001). Head Neck. 23: 201–213. Rogers et al. (2008). www.headandneckcancer.co.uk.
127
128
6
The University of Washington Quality of Life Scale
Scrimger R, Kanji A, Parliament M, Warkentin H, Field C, Jha N, Hanson J. (2007). Am J Clin Oncol 30(3): 271–277. Smith GI, Yeo D, Clark J, Choy ET, Gao K, Oates J, O’Brien CJ. (2006). Br J Oral Maxfax Surg. 44(3): 187–192. Smucler R, Mazanek J. (2004). Lasers Surg Med. 34(1): 12–17. Talmi YP, Horowitz Z, Bedrin L, Wolf M, Chaushu G, Kronenberg J, Pfeffer MR. (2002). Cancer. 94(4): 1012–1017. Thone M, Karengera D, Siciliano S, Reychler H. (2003). Revue de Stomatologie et de Chirurgie MaxilloFaciale. 104(1): 19–24. Vartanian JG, Carvalho AL, Yueh B, Priante AV, de Melo RL, Correia LM, Kohler HF, Toyota J, Kowalski IS, Kowalski LP. (2004). Arch. Otolaryngol Head Neck Surg. 130(10): 1209–1213. Vartanian JG, Carvalho AL, Toyota J, Kowalski IS, Kowalski LP. (2006a). Arch Otolaryngol Head Neck Surg. 132(1): 32–35. Vartanian JG, Carvalho AL, Yueh B, Furia CL, Toyota J, McDowell JA. (2006b). Head Neck Surg. 28(12): 1115–1121.
Vigili MG, Colacci AC, Magrini M, Cerro P, Marzetti A. (2002). Eur Arch Otolaryngol. 259(1): 11–16. Vilaseca I, Chen AY, Backscheider AG. (2006). Head Neck 28(4): 313–320. Wang G, Ji W, Pan Z, Guo X. (2002). Chung-Hua Chung Liu Tsa Chih. Chin J Oncol. 24(1): 53–56. Weymuller EA, Yueh B, Deleyiannis FWB, Kuntz AL, Alsarraf R, Coltrera MD. (2000). Arch Otolaryngol Head Neck Surg. 126: 329–335. Weymuller EA Jr. Alsarraf R, Yueh B, Deleyiannis FW, Coltrera MD. (2001). Arch Otolaryngol Head Neck Surg. 127(5): 489–493. Winter SC, Cassell O, Corbridge RJ, Goodacre T, Cox GJ. (2004). Clin Otolaryngol Allied Sci. 29(3): 274–278. Young CW, Pogrel MA, Schmidt BL. (2007). J Oral Maxfax Surg. 65(4): 706–712. Zhang L, Luan X, Pan X, Xie G, Xu F, Liu D, Lei D. (2002). [Chinese] Chung Hua Erh Pi Yen Hou Ko Tsa Chih - Chin J Otolaryngol. 37(1): 11–14.
7 Comparison of Three Quality of Life Questionnaires in Urinary Incontinence Ja Hyeon Ku . Seung-June Oh 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
2
Types of Urinary Incontinence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
3
Limitations of Objective Tests on Urinary Incontinence . . . . . . . . . . . . . . . . . . . . . . . . . . 131
4 Generic QOL Questionnaires on Urinary Incontinence . . . . . . . . . . . . . . . . . . . . . . . . . . . 132 4.1 Advantages and Disadvantages of Generic QOL Questionnaires . . . . . . . . . . . . . . . . . . . . 132 4.2 Use of Generic QOL Questionnaires in Urinary Incontinence . . . . . . . . . . . . . . . . . . . . . . 132 5 5.1 5.2 5.3
Three Disease-Specific QOL Questionnaires in Urinary Incontinence . . . . . . . . . . . . 133 The BFLUTS Questionnaire . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133 The I-QOL Questionnaire . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135 The KHQ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
6
Comparisons Between the Three QOL Questionnaires in Urinary Incontinence Patients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138 6.1 Correlations Between the Three QOL Questionnaires . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138 6.2 Minimal Clinically Importance Change . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140 6.3 Recommendations for Use . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141 7
Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
#
Springer Science+Business Media LLC 2010 (USA)
130
7
Comparison of Three Quality of Life Questionnaires in Urinary Incontinence
Abstract: Urinary incontinence is not life-threatening, but loss of urinary control can affect the social, psychological, domestic, occupational, physical, and the sexual aspects of patients’ lives. A number of reliable, valid instruments are available for assessing the impact of urinary incontinence on quality of life (QOL). Of these many questionnaires, we selected three, namely, the Bristol Female Lower Urinary Tract Symptoms (BFLUTS) questionnaire, the Incontinence Quality of Life (I-QOL) questionnaire, and the King’s Health Questionnaire (KHQ), and compared the results obtained. These questionnaires have been developed primarily to assess urinary incontinence in women and the impact that urinary incontinence has on aspects of everyday QOL. In the present study, when BFLUTS and I-QOL were compared, the “BFLUTS-IS” and “BFLUTS-QOL” domains were found to correlate inversely with only two and three domains in the I-QOL, respectively. When BFLUTS and the KHQ were compared, the scores of the “BFLUTS-IS” and “BFLUTS-QOL” domains were found to correlate with six domains in the KHQ questionnaire, but these correlations were low to moderate, with correlation coefficients that ranged from 0.412 to 0.649. When I-QOL and the KHQ were compared, “Role limitations” and “Emotional problems” domains in the KHQ were found to correlate with all domains in the I-QOL. Moreover, significant negative correlations were noted between the “Severity measures” domain in the KHQ and all domains in the I-QOL; however, correlations were low to moderate (range: –0.384 to –0.650). Other domains in the three questionnaires showed no correlation with each other. Since a plethora of measurement instruments are available that vary in scope and content, subjective QOL results of urinary incontinence using specific-condition QOL questionnaires may differ. Therefore, before deciding on an instrument, the content on the instrument’s items should be thoroughly reviewed to ensure that a particular aspect of QOL does not need additional assessment. List of Abbreviations: BFLUTS, Bristol female lower urinary tract symptoms; BFLUTS-FS, BFLUTS filling symptoms; BFLUTS-IS, BFLUTS incontinence symptoms; BFLUTS-QOL, BFLUTS quality of life; BFLUTS-sex, BFLUTS sexual function; BFLUTS-VS, BFLUTS voiding symptoms; I-QOL, incontinence quality of life; KHQ, King’s health questionnaire; QOL, quality of life; SF-36, medical outcomes study short form-36
1
Introduction
Urinary incontinence is defined by the International Continence Society as “a complaint of any involuntary leakage of urine” (Abrams et al., 2002), and is a common problem that affects around 20–30% of the adult population (Sandvik et al., 1993; Thomas et al., 1980). Although urinary incontinence is not life-threatening, loss of urinary control can affect the social, psychological, domestic, occupational, physical, and the sexual aspects of patients’ lives (Thomas et al., 1980). Assessments of incontinence severity have been the mainstay of investigations on the burden imposed by this condition, but it is now recognized that psychosocial adjustment to illness is as important as the status of the physical disease. Therefore, it is essential that health care evaluations focus on this concern by incorporating some measure of abstract subjective feelings or quality of life (QOL). However, it is difficult to appreciate the extent to which troublesome incontinence symptoms disrupt daily life. Generic measures of health status have proven useful for comparing general and specific populations, estimating the relative burdens of different diseases, differentiating the health benefits produced by a wide range of different treatments, and screening individual patients
Comparison of Three Quality of Life Questionnaires in Urinary Incontinence
7
(Manocchia et al., 1998). However, although generic QOL instruments can be used for such assessments, they may not be sensitive enough to detect the characteristics of urinary incontinence. Thus, several disease-specific QOL instruments have been developed to provide detailed information on how urinary incontinence affects patients’ lives. The Bristol Female Lower Urinary Tract Symptoms (BFLUTS) questionnaire has good psychometric > validity and reliability (Jackson et al., 1996). Using data from a large, randomized clinical trial that assessed treatments for women with urinary incontinence, Wagner et al. (1996) demonstrated that the Incontinence Quality of Life (I-QOL) questionnaire has good > internal consistency, > reproducibility, and validity. Moreover, the King’s Health Questionnaire (KHQ) is a valid and reliable instrument for assessing QOL in women with voiding symptoms, including urinary incontinence (Kelleher et al., 1997). Nevertheless, no questionnaire is generally accepted for the assessment of urinary incontinence. Based on the results of a previous study (Oh and Ku, 2007), we focused on the impact of urinary incontinence on QOL using three disease-specific QOL instruments and compared the results obtained.
2
Types of Urinary Incontinence
Urinary incontinence may be classified as stress urinary incontinence (involuntary leakage during effort or exertion, or during sneezing or coughing), urge urinary incontinence (involuntary leakage accompanied by or immediately preceded by urgency), or mixed incontinence (a combination of stress and urge urinary incontinence) (Abrams et al., 2002). Since the underlying pathophysiologies of stress and urge urinary incontinence differ (i.e., stress urinary incontinence is attributable to urethral hypermobility or sphincter weakness and urge urinary incontinence results from detrusor overactivity) the symptomatic presentations of stress and urge urinary incontinence differ. Thus, incontinence types may affect QOL differentially, i.e., urge urinary incontinence has been reported to have a greater impact on QOL than stress urinary incontinence (Robinson et al., 1998; Wyman et al., 1987). Nevertheless, regardless of type, incontinence has been shown to have a detrimental impact on patient healthrelated QOL.
3
Limitations of Objective Tests on Urinary Incontinence
The principle aim of objective tests is to reproduce a patient’s symptoms and to provide a pathophysiological explanation. Several investigators have demonstrated that low maximal urethral closure pressure and low Valsalva leak point pressure values are associated with a higher grade of incontinence severity (Bump et al., 1997; Nitti and Combs, 1996). However, objective diagnostic tests, which are usually pad and/or urodynamic tests, may not take into account patients’ perceptions of their problems. Moreover, several authors have found that symptom scores and QOL measures are not correlated with urodynamic measures that they inadequately predict urodynamic outcome (FitzGerald and Brubaker, 2002; Swift and Ostergard, 1995). Nager et al. (2001) found insignificant correlations between urodynamic parameters and pad loss or QOL measures, and a poor correlation was found between the subjective degrees of the bothersomenesses of symptoms and objective measures of degree of urinary incontinence (Ryhammer et al., 1995; Wyman et al., 1987). Moreover, Oh and
131
132
7
Comparison of Three Quality of Life Questionnaires in Urinary Incontinence
Ku (2007) found a poor correlation between QOL questionnaire findings and objective test results. Several studies have suggested that the impact of urinary incontinence is not solely a function of its severity, but that it also depends on abilities of individuals to cope (Lenderking et al., 1996). Moreover, an individual’s perception of continence status and the reporting of leakage episodes have been reported to be modulated by differences in personality characteristics (Frazer et al., 1989). These findings support the notion that since objective clinical measures do not reflect the patient’s view, QOL measures should be included in clinical practice.
4
Generic QOL Questionnaires on Urinary Incontinence
Two broad categories of QOL questionnaires exist. Generic QOL questionnaires have been developed for use in a wide range of clinical populations, whereas disease-specific QOL questionnaires include items directly related to a specific medical condition.
4.1
Advantages and Disadvantages of Generic QOL Questionnaires
Generic measures of health-related QOL are recommended for assessing the general quality of healthcare, and are used to monitor health in large populations, and this form of assessment may be preferred in research covering the full adult age spectrum. However, although generic outcome instruments are designed for use in any population, their validity and reliability vis-a`-vis a particular disease should be verified to ensure their appropriateness. Furthermore, the abilities of generic instruments to assess QOL in impairment-specific populations are likely to depend on group characteristics, and these instruments may be more susceptible to factors other than disease severity and reflect wider aspects of daily life.
4.2
Use of Generic QOL Questionnaires in Urinary Incontinence
To date, several studies have used generic health questionnaires to evaluate the impact of urinary incontinence on health-related QOL. The Sickness Impact Profile developed by Hunskaar and Vinsnes (1991) revealed that urge symptoms are associated with greater QOL impairment than the symptoms of stress urinary incontinence. Grimby et al. (1993) used the Nottingham Health Profile Questionnaire to assess QOL and found that women suffering from urge or mixed urinary incontinence reported emotional disturbances more so than continent women. Simeonova et al. (1999) assessed QOL using a visual analogue scale and found that women with urge or mixed urinary incontinence reported a lower QOL than those with stress urinary incontinence. In fact, the majority of studies carried out over the past decade indicate that urge urinary incontinence has a greater impact on health-related QOL than stress urinary incontinence (Grimby et al., 1993; Hunskaar and Vinsnes, 1991; Lenderking et al., 1996; Sandvik et al., 1993; Simeonova et al., 1999). However, Ho-Yin et al. (2003) found that a generic questionnaire (the Medical Outcomes Study Short Form-36 [SF-36]) was unable to detect a significant difference between the QOL measurements of women suffering from genuine stress urinary incontinence and detrusor instability. The effect of stress urinary incontinence in women on QOL is a controversial issue. Women with stress urinary incontinence have reported an inferior QOL than continent
Comparison of Three Quality of Life Questionnaires in Urinary Incontinence
7
women in some studies (Simeonove et al., 1999), whereas other studies have concluded that stress urinary incontinence does not significantly affect QOL (Grimby et al., 1993). Oh and Ku (2006) were unable to confirm the hypothesis that women suffering from stress urinary incontinence have an inferior QOL using the generic SF-36 questionnaire. However, the poor content validities of generic health scales for urinary incontinence would make them less sensitive to disease-specific changes, which makes them unsuitable for assessing treatment efficacy.
5
Three Disease-Specific QOL Questionnaires in Urinary Incontinence
A large number of studies have been undertaken with a view toward developing a valid diseasespecific QOL instrument for use in clinical trials. However, no generally accepted symptom questionnaire is currently available for the assessment of urinary incontinence. Furthermore, questionnaires vary widely, depending on their intended purposes and target populations. Nevertheless, tools are available for differentiating stress and urge urinary incontinence, for quantifying urinary symptom severities, for assessing urinary symptom severities and impacts of symptoms, and for measuring treatment outcomes. The KHQ, I-QOL and BFLUTS questionnaires were developed primarily to assess urinary incontinence and its impacts on aspects of QOL. These questionnaires have been determined to be reliable and valid for use in incontinent women, and the International Consultation on Incontinence recommended that these three questionnaires are suitable for use in urinary incontinence (Donovan et al., 2002). However, these questionnaires have their own particularities, and thus, instruments are invariably chosen based on the population under consideration.
5.1
The BFLUTS Questionnaire
The BFLUTS questionnaire has eight items related to urinary incontinence, four of which were designed specifically to quantify urinary leakage (Jackson et al., 1996) (> Table 7-1). Twelve items address other symptoms; four are associated with the storage phase and eight with the voiding phase. Nine items address other aspects of QOL and four items address sexual function. The format of BFLUTS is similar to that of the International Continence Society male questionnaire developed for the International Continence Society Benign Prostatic Hyperplasia study (Donovan et al., 1996). Although it is biased towards assessing urinary incontinence, the BFLUTS instrument was intended to cover all symptoms pertaining to female lower urinary tract dysfunction (Jackson et al., 1996). Thus, unlike the majority of other questionnaires, BFLUTS also includes questions on a range of troublesome symptoms that are not commonly included in other questionnaires. Recently, Brookes et al. (2004) described the development and validation of a scored form based on the BFLUTS questionnaire. Three domains were identified to assess symptoms: incontinence (five items, BFLUTS-IS); voiding (three items, BFLUTS-VS); and filling (four items, BFLUTS-FS); with additional subscales for sexual function (two items, BFLUTS-sex) and QOL (five items, BFLUTS-QOL). All scales have simple additive scores.
133
134
7
Comparison of Three Quality of Life Questionnaires in Urinary Incontinence
. Table 7-1 Items of urinary symptoms, sexual matters, and quality of life in the BFLUTS questionnaire and its scored form BFLUTS items
Scored form
Urinary questions “During the day, how many times do you urinate on average?
BFLUTS-FS
During the night, how many times do you get up to urinate, on average?
BFLUTS-FS
Do you have to rush to the toilet to urinate?
BFLUTS-FS
Does urine leak before you can get to the toilet?
BFLUTS-IS
Do you have pain in your bladder?
BFLUTS-FS
How often do leak urine?
BFLUTS-IS
Does urine leak when you are physically active, exert yourself, cough or sneeze? BFLUTS-IS Do you ever leak urine for no obvious reason and without feeling that you want BFLUTS-IS to go? How much urinary leakage occurs? Is there a delay before you can start to urinate?
BFLUTS-VS
Do you have to strain to urinate?
BFLUTS-VS
Do you stop and start more than once while you urinate?
BFLUTS-VS
Do you leak urine when you are asleep?
BFLUTS-IS
Would you say that the strength of your urinary stream is. . . Have you ever blocked up completely so that you could not urinate at all and had to have a catheter to drain the bladder? Do you have a burning feeling when you urinate? How often do you feel that your bladder has not emptied completely after you have urinated? Can you stop the flow of urine if you try while you are urinating? How often do you pass urine during the day? Sexual questions Do you have pain or discomfort because of a dry vagina? To what extent do you feel that your sex life has been spoilt by your urinary symptoms?
BFLUTS-sex
Do you have pain when you have sexual intercourse? Do you leak urine when you have sexual intercourse?
BFLUTS-sex
Quality of life Do you have to change your underclothes or wear protection because of your leakage? How many times a day do you change the above items because of leakage? Do you need to change your outer clothing during the day because of urine leakage?
BFLUTS-QOL
Do you cut down on the amount of fluid you drink so that your urinary symptoms BFLUTS-QOL improve, and you can do the tings that you want to do? To what extent have your urinary symptoms affected your ability to perform daily BFLUTS-QOL tasks (e.g., cleaning, DIY, lifting objects)?
Comparison of Three Quality of Life Questionnaires in Urinary Incontinence
. Table 7-1 (continued) BFLUTS items Do you avoid places and situations where you know a toilet is not nearby (e.g., shopping, traveling, theater, church)?
7
Scored form BFLUTS-QOL
Do your urinary symptoms interfere with physical activity (e.g., walking, dancing, swimming)? Overall, how much do your urinary symptoms interfere with your social life (going out, meeting friends and so on)? Overall, how much do your urinary symptoms interfere with your life?
BFLUTS-QOL
How long have you had urinary symptoms that bother you? If you had to spend the rest of your life with your urinary symptoms as they are now, how would you feel? The BFLUTS questionnaire contains items relating to urinary incontinence (eight items), symptoms associated with the storage phase (four items) and the voiding phase (eight items), sexual function (four items), and quality of life (nine items). The scored form of the BFLUTS questionnaire contains five domains; incontinence (five items, BFLUTS-IS), voiding (three items, BFLUTS-VS), filling (four items, BFLUTS-FS), sexual function (two items, BFLUTS-sex), and QOL (five items, BFLUTS-QOL). BFLUTS Bristol female lower urinary tract symptoms; BFLUTS-IS BFLUTS incontinence symptoms; BFLUTS-VS BFLUTS voiding symptoms; BFLUTS-FS BFLUTS filling symptoms; BFLUTS-sex BFLUTS sexual function; BFLUTS-QOL BFLUTS quality of life. Source: Jackson et al. (1996), printed with permission; Brookes et al. (2004), printed with permission
5.2
The I-QOL Questionnaire
This disease-specific, 22-item I-QOL instrument includes questions that evaluate the distress and impact of urinary incontinence in three domains: avoidance and limiting behavior (eight items), social embarrassment (five items) and psychosocial impact (nine items) (> Table 7-2). For example, the I-QOL includes questions such as “I worry about coughing and sneezing” (Item 2 in avoidance and limiting behavior), “I worry about being embarrassed or humiliated because of my incontinence” (Item 14 in social embarrassment) and “My incontinence makes me feel likr I’m not a healthy person” (Item 15 in psychosocial impact). These item scores are summed and then transformed to a 0–100 scale, where a higher score represents a better QOL (Patrick et al., 1999; Wagner et al., 1996). The I-QOL questionnaire has been used to evaluate the effect of different therapies on women with urinary incontinence (Almeida et al., 2004; Bump et al., 2003; Vandoninck et al., 2003). The I-QOL questionnaire is a highly targeted condition-specific questionnaire that assesses the impact and distress of specific incontinence symptoms. The potential advantage of the I-QOL over previous measures is its applicability to patients over a range of ages and with varying types and severities of urinary incontinence. The I-QOL is capable of discriminating between different levels of perceived severity, the frequencies of incontinent episodes, and stress test pad weights. Because this instrument was developed on the basis that its items should be meaningful to those with urinary incontinence, I-QOL scores are not significantly affected by demographic variables.
5.3
The KHQ
The KHQ is a 21-item, condition-specific instrument designed to assess QOL, and has been shown to be a valid and reliable psychometric questionnaire (Kelleher et al., 1997)
135
136
7
Comparison of Three Quality of Life Questionnaires in Urinary Incontinence
. Table 7-2 Items and domains of I-QOL questionnaire Items
Domains
I worry about not being able to get to the toilet on time
Avoidance and limiting behavior
I worry about coughing and sneezing
Avoidance and limiting behavior
I have to be careful about standing up after sitting down
Avoidance and limiting behavior
I worry where the toilets are in new places
Avoidance and limiting behavior
I feel depressed
Psychosocial impacts
I don’t feel free to leave my home for long periods of time
Psychosocial impacts
I feel frustrated because my UI prevents me from doing what I want Psychosocial impacts I worry about others smelling urine on me
Social embarrassment
Incontinence is always on my mind
Psychosocial impacts
It’s important for me to make frequent trips to the toilet
Avoidance and limiting behavior
Because of my incontinence, it is important to plan every detail in advance
Avoidance and limiting behavior
I worry about my incontinence getting worse as I grow older
Social embarrassment
I have a hard time getting a good night’s sleep
Avoidance and limiting behavior
I worry about being embarrassed or humiliated be cause of my incontinence
Social embarrassment
My incontinence makes me feel like I’m not a healthy person
Psychosocial impacts
My UI makes me feel helpless
Psychosocial impacts
I get less enjoyment out of life because of my UI
Psychosocial impacts
I worry about wetting myself
Social embarrassment
I feel like I have no control over my bladder
Social embarrassment
I have to watch what I drink
Avoidance and limiting behavior
My UI limits my choice of clothing
Psychosocial impacts
I worry about having sex
Psychosocial impacts
The I-QOL instrument contains 22 questions that evaluate both the distress and impact of urinary incontinence, i.e., avoidance and limiting behavior (eight items), social embarrassment (five items), and psychosocial impact (nine items). I-QOL incontinence quality of life; UI urinary incontinence. Source: Patrick et al. (1999), printed with permission
(> Table 7-3), and in women with voiding symptoms, including stress urinary incontinence and an overactive bladder (Bidmead et al., 2001; Kelleher et al., 1997; Uemura and Homma, 2004). The KHQ contains seven multi-domains, i.e., role limitations, physical limitations, social limitations, personal relationships, emotional problems, sleep/energy disturbances, and
Comparison of Three Quality of Life Questionnaires in Urinary Incontinence
7
. Table 7-3 Items and domains of KHQ Items How would you describe your health at the present?
Domains General health perception
How much do you think your bladder problem affects your life?
Impact on life
Does your bladder problem affect your household tasks? (cleaning, shopping etc)
Role limitations
Does your bladder problem affect your job, or your normal daily activities outside the home?
Role limitations
Does your bladder problem affect your physical activities (e.g., going for a walk, running, sport, gym etc)?
Physical limitations
Does your bladder problem affect your ability to travel?
Physical limitations
Does your bladder problem limit your social life?
Social limitations
Does your bladder problem limit your ability to see and visit friends?
Social limitations
Does your bladder problem affect your family life?
Social limitations
Does your bladder problem affect your relationship with your partner?
Personal relationships
Does your bladder problem affect your sex life?
Personal relationships
Does your bladder problem make you feel depressed?
Emotional problems
Does your bladder problem make you feel anxious or nervous?
Emotional problems
Does your bladder problem make you feel bad about yourself?
Emotional problems
Does your bladder problem affect your sleep?
Sleep/energy disturbances
Does your bladder problem make you feel worn out and tired ?
Sleep/energy disturbances
Wear pads to keep dry?
Severity measures
Be careful how much fluid you drink?
Severity measures
Change your underclothes because they get wet?
Severity measures
Worry in case you smell?
Severity measures
Get embarrassed because of your bladder problem?
Severity measures
The KHQ has 21 items that assess quality of life and contains seven domains, i.e., role limitations, physical limitations, social limitations, personal relationships, emotional problems, sleep/energy disturbances, severity measures, general health perceptions, and impact on life. KHQ: King’s Health Questionnaire. Source: Kelleher et al. (1997), printed with permission
severity measures, and two single-item domains: general health perceptions and impact on life. The scoring algorithm used has been described (Kelleher et al., 1997). Possible KHQ scores range from zero (best health perception) to 100 (worst health perception). An abridged version of the KHQ has also been developed and validated in Japan (Homma and Uemura, 2004). Instead of the 16 items grouped into eight domains in the KHQ, this edited version consists of six items grouped into two domains (daily life limitations and mental health).
137
138
7
Comparison of Three Quality of Life Questionnaires in Urinary Incontinence
6
Comparisons Between the Three QOL Questionnaires in Urinary Incontinence Patients
6.1
Correlations Between the Three QOL Questionnaires
Oh and Ku (2007) assessed 28 patients with stress urinary incontinence. When BFLUTS and I-QOL were compared, the “BFLUTS-IS” domain was found to be inversely correlated with the following domains in the I-QOL; the “Psychological impact” domain (r ¼ 0.441), the “Social embarrassment” domain (r ¼ 0.456) and “Total sum” (r ¼ 0.488). Significant negative correlations were also noted between the “BFLUTSQOL” domain and IQOL “Avoidance behaviors” (r ¼ 0.649), “Psychological impact” (r ¼ 0.438), and “Social embarrassment” domains (r ¼ 0.600) and IQOL “Total sum” (r ¼ 0.637). The other three domains of BFLUTS were not found to correlate with any domain in the IQOL (> Table 7-4). When BFLUTS and KHQ were compared, the “BFLUTS-FS,” “BFLUTS-VS,” and “BFLUTS-sex” domains were not found to be correlated with most KHQ domains. However, the scores of the “BFLUTS-IS” and “BFLUTS-QOL” domains were correlated with those of six domains in the KHQ questionnaire, but correlations were low to moderate with correlation coefficients that ranged from 0.412 to 0.649. Highest correlation was found between the “BFLUTS-QOL” domain and the KHQ “Emotional problems” domain (r ¼ 0.649) (> Table 7-5). When the I-QOL and KHQ were compared, the “Role limitations” and “Emotional problems” domains in the KHQ were found to be correlated with all domains in the I-QOL. Statistically significant negative correlations were also noted between the “Severity measures” domain in the KHQ and all domains in the I-QOL. However, correlations were low to moderate (range: –0.384 to –0.650). Highest correlation was found between the “Social . Table 7-4 Correlation coefficients between the BFLUTS and I-QOL BFLUTS Filling symptoms
Voiding symptoms
Incontinence symptoms
Sexual function
Quality of life
Avoidance behaviors
0.359
0.024
0.318
0.045
0.649**
Psychological impacts
0.131
0.007
0.441*
0.145
0.438*
Social embarrassment
0.196
0.121
0.456*
0.005
0.600**
Total
0.293
0.014
0.488**
0.141
0.637**
I-QOL
“BFLUTS incontinence symptoms” domain was found to correlate inversely with the I-QOL “Psychological impact” domain (r ¼ 0.441), “Social embarrassment” domain (r ¼ 0.456) and “Total sum” (r ¼ 0.488), and the “BFLUTS quality of life” domain correlated with inversely with I-QOL “Avoidance behaviors” domain (r ¼ 0.649), “Psychological impact” domain (r ¼ 0.438), “Social embarrassment” domain (r ¼ 0.600) and “Total sum” (r ¼ 0.637). The other three domains in BFLUTS did not correlate with any domain in the I-QOL. BFLUTS Bristol female lower urinary tract symptoms; I-QOL incontinence quality of life. Source: Oh and Ku (2007), printed with permission * p < 0.05, **p < 0.01
0.544
0.230
0.615
0.235
0.007 0.540
0.358 b
0.562b
0.266
0.530b
0.277
0.512b
Social limitations
0.082
Physical limitations
0.247
0.649
0.071 b
0.505b
0.172 0.495b
0.058
0.306
Emotional problems
0.031
0.098
Personal relationships
KHQ
0.558
0.361
0.229
0.323
0.365
b
Sleep/energy disturbances
0.539b
0.178
0.599b
0.012
0.082
Severity measures
“BFLUTS filling symptoms,” “BFLUTS voiding symptoms,” and “BFLUTS sexual function” domains did not correlate with most domains in the KHQ questionnaire. Scores for “BFLUTS incontinence symptoms” and “BFLUTS quality of life” domains were found to correlate with six domains in the KHQ questionnaire. BFLUTS Bristol Female Lower Urinary Tract Symptoms; KHQ: King’s Health Questionnaire. Source: Oh and Ku (2007), printed with permission a p < 0.05, bp < 0.01
Quality of life
0.048
0.200
Sexual function
0.309
b
0.074
0.320
0.412a
Incontinence symptoms
a
0.599b
0.198
0.231
Voiding symptoms
0.038
0.219
0.400a
Filling symptoms
Role limitations
Impact on life
General health
BFLUTS
. Table 7-5 Correlation coefficients between the BFLUTS and KHQ
Comparison of Three Quality of Life Questionnaires in Urinary Incontinence
7 139
140
7
Comparison of Three Quality of Life Questionnaires in Urinary Incontinence
. Table 7-6 Correlation coefficients between the I-QOL and KHQ I-QOL Avoidance behaviors
KHQ
Psychological impacts
Social embarrassment
General health
0.304
0.063
0.147
Impact on life
0.209
0.471*
0.342
Role limitations
0.384
0.554
**
Physical limitations
0.341
0.491
**
Social limitations
0.260
0.388
*
0.072
0.029
Personal relationships
*
Total 0.196 0.319
0.612
**
0.532**
0.447
*
0.461*
0.417
*
0.397*
0.074
0.019
Emotional problems
0.495**
0.611**
0.650**
0.639**
Sleep/energy disturbances
0.457*
0.344
0.341
0.464*
Severity measures
0.499**
0.455*
0.570**
0.545**
“Role limitations,” “Emotional problems” and “Severity measures” domains in the KHQ correlated with all domains in the I-QOL, but “General health” and “Personal relationships” domains in the KHQ were not significantly correlates with any domain in the I-QOL. I-QOL incontinence quality of life; KHQ King’s Health Questionnaire. Source: Oh and Ku (2007), printed with permission * p < 0.05, **p < 0.01
embarrassment” domain in the I-QOL and the “Emotional problems” domain in the KHQ (r ¼ 0.650). “General health” and “Personal relationships” domains in the KHQ were not found to be significantly correlated with any domain in the I-QOL (> Table 7-6).
6.2
Minimal Clinically Importance Change
In a study that used data from two clinical trials on tolterodine in overactive bladder, Kelleher et al. (2004) used > anchor-based approach and > distribution-based approach to calculate minimally important differences for the KHQ. They found that a KHQ change of 5 points indicated a clinically important difference in health-related QOL. In a randomized, doubleblind, placebo-controlled study on an oxybutynin transdermal patch, a reduction of three times per week was found to be the minimum important change in incontinence frequency among Japanese patients with an overactive bladder (Homma and Koyama, 2006). However, the KHQ requires further testing before it can be used in men with incontinence symptoms other than an overactive bladder (Symonds, 2003). Recently, in a study using data from two randomized, placebo-controlled duloxetine studies, Yalcin et al. (2006) proposed 2.5 points as a reasonable guide for an I-QOL > between-treatment minimal clinically importance change and 6.3 points as a > withintreatment minimal clinically importance change. However, the I-QOL needs further testing before use in incontinent men (Symonds, 2003). However, further research is needed to establish clinically relevant and interpretable cutoff points for BFLUTS scores (Brookes et al., 2004). Hence, the BFLUTS requires appropriate whole instrument testing, including testing of the health-related QOL and sexual functioning items (Symonds, 2003).
Comparison of Three Quality of Life Questionnaires in Urinary Incontinence
6.3
7
Recommendations for Use
Based on assessments of validity, reliability, responsiveness, utility, and frequency of use, Ross et al. (2006) recommended that some questionnaires, including the I-QOL, be viewed as first choice QOL measures in trials on incontinence treatments. Naughton et al. (2004) identified 1,300 published articles relating to symptoms of urinary incontinence, effects on QOL, and outcomes assessments of incontinence treatments, evaluated the instruments used using priori criteria and graded for quality. To be highly recommended (grade A), an instrument was required to have supportive published data that provided evidence of its reliability, validity, and responsiveness to change. For a grade B recommendation, an instrument was required to have published data that provided evidence of its reliability and validity and of its relevance to individuals with urinary incontinence. Questionnaires must have reached at least grade B to be recommended for use. For assessing symptoms of urinary incontinence, the KHQ was highly recommended (grade A), BFLUTS was also recommended (grade B), but the I-QOL was not. For assessing bothersomeness in persons with incontinence, only BFLUTS was recommended (grade B), and for assessing the effect of incontinence on QOL, the I-QOL and the KHQ were highly recommended (grade A), whereas BFLUTS was not recommended. Recently, Reid et al. (2007) undertook a psychometric validation of three questionnaires, including BFLUTS and KHQ, to assess surgical outcomes of stress urinary incontinence. They found that the all three questionnaires had limitations when used as outcome measures. These findings suggest that when instruments are used in different populations their psychometric properties may change.
7
Conclusions
Health-related QOL is partly a reflection of an individual’s ability to cope and adapt to a new life situation. The effect of physical disability or illness cannot be understood without taking into consideration the specific areas of functioning affected by an individual’s physical condition and those aspects of QOL that are of particular importance to an individual. Subjective QOL results on urinary incontinence using specific-condition QOL questionnaires may differ because of the many measurement instruments available that vary in scope and content. However, this does not imply that several disease-specific QOL instruments should be used simultaneously to evaluate patients with urinary incontinence. Rather, it means that before deciding on an instrument, the content on the instrument’s items should be thoroughly reviewed to ensure that a particular aspect of QOL does not need additional assessment.
Summary Points Urinary incontinence is a common problem and may affect patient QOL. Urge urinary incontinence has a greater impact on QOL than stress urinary incontinence. Since objective clinical measures do not reflect the patient’s viewpoint, QOL measures should be included in clinical practice.
Because generic health scales have poor content validity for urinary incontinence, they may be unsuitable for assessing treatment efficacy.
141
142
7
Comparison of Three Quality of Life Questionnaires in Urinary Incontinence
BFLUTS includes a number of other troublesome symptoms that are not commonly included in other questionnaires, in addition to a range of incontinence symptoms. However, BFLUTS requires appropriate whole instrument testing, including of its health-related QOL and sexual functioning items. The potential advantage of the I-QOL over previous measures is its applicability to patients over a range of ages and with varying types and severities of urinary incontinence. The I-QOL is highly recommended for assessing the effect of incontinence on QOL, but it needs further testing before use in incontinent men. KHQ is highly recommended for assessing symptoms of urinary incontinence and their effects on QOL. However, the KHQ requires further testing in men with incontinence symptoms other than an overactive bladder. Subjective QOL results in women with urinary incontinence using specific-condition QOL questionnaires can differ because the many measurement instruments available in scope and content.
References Abrams P, Cardozo L, Fall M, Griffiths D, Rosier P, Ulmsten U, van Kerrebroeck P, Victor A, Wein A. (2002). Neurourol Urodyn. 21: 167–178. Almeida FG, Bruschini H, Srougi M. (2004). J Urol. 171: 1571–1574. Bidmead J, Cardozo LD, McLellan A, Khullar V, Kelleher CJ. (2001). BJOG. 108: 408–413. Brookes ST, Donovan JL, Wright M, Jackson S, Abrams P. (2004). Am J Obstet Gynecol 191: 73–82. Bump RC, Coates KW, Cundiff GW, Harris RL, Weidner AC. (1997). Am J Obstet Gynecol. 177: 303–310. Bump RC, Norton PA, Zinner NR, Yalcin I. (2003). Duloxetine Urinary Incontinence Study Group Obstet Gynecol. 102: 76–83. Donovan JL, Abrams P, Peters TJ, Kay HE, Reynard J, Chapple C, De La Rosette JJ, Kondo A. (1996). Br J Urol. 77: 554–562. Donovan JL, Badia X, Corcos J, Gotoh M, Kelleher C, Naughton M. (2002). In: Abrams P, Cardozo L, Khoury S, Wein A (eds.) Incontinence. Health Publication Ltd, Plymouth, UK, pp. 267–316. FitzGerald MP, Brubaker L. (2002). Neurourol Urodyn. 21: 30–35. Frazer MI, Haylen BT, Sutherst JR. (1989). Br J Urol. 63: 14–15. Grimby A, Milsom I, Molander U, Wiklund I, Ekelund P. (1993). Age Ageing 22: 82–89. Homma Y, Koyama N. (2006). Neurourol Urodyn. 25: 228–235. Homma Y, Uemura S. (2004). BJU Int. 93: 1009–1013. Ho-Yin PL, Man-Wah P, Shing-Kai Yip. (2003). Acta Obstet Gynecol Scand. 82: 275–279.
Hunskaar S, Vinsnes A. (1991). J Am Geriatr Soc. 39: 378–382. Jackson S, Donovan J, Brookes S, Eckford S, Swithinbank L, Abrams P. (1996). Br J Urol. 77: 805–812. Kelleher CJ, Cardozo LD, Khullar V, Salvatore S. (1997). Br J Obstet Gynaecol. 104: 1374–1379. Kelleher CJ, Pleil AM, Reese PR, Burgess SM, Brodish PH. (2004). Br J Obstet Gynaecol. 111: 605–612. Lenderking WR, Nackley JF, Anderson RB, Testa MA. (1996). Pharmacoeconomics. 9: 11–23. Manocchia M, Bayliss MS, Connor J. (1998). SF-36 Health Survey Annotated Bibliography: Second Edition (1988–1996). The Health Assessment Lab, New England Medical Center, Boston, MA. Nager CW, Schulz JA, Stanton SL, Monga A. (2001). Int Urogynecol J. 12: 395–400. Naughton MJ, Donovan J, Badia X, Corcos J, Gotoh M, Kelleher C, Lukacs B, Shaw C. (2004). Gastroenterology. 126(Suppl 1): S114–S123. Nitti VW, Combs AJ. (1996). J Urol. 155: 281–285. Oh SJ, Ku JH. (2006). Qual Life Res. 15: 493–501. Oh SJ, Ku JH. (2007). Scand J Urol Nephrol. 41: 66–71. Patrick DL, Martin ML, Bushnell DM, Yalcin I, Wagner TH, Buesching DP. (1999). Urology. 53: 71–76. Reid FM, Smith ARB, Dunn G. (2007). Neurourol Urodyn. 26L: 123–128. Robinson D, Pearce KF, Preisser JS, Dugan E, Suggs PK, Cohen SJ. (1998). Obstet Gynecol. 91: 224–228. Ross S, Soroka D, Karanhalios A, Glazener CMA, Hay-Smith EJ, Drutz HP. (2006). Int Urogynecol J. 17: 272–285.
Comparison of Three Quality of Life Questionnaires in Urinary Incontinence Ryhammer AM, Djurhuus JC, Laurberg S, Hermann AP. (1995). Neurourol Urodyn. 14: 456–457. Sandvik H, Kveine E, Hunskaar S. (1993). Scand J Caring Sci. 7: 53–56. Simeonova Z, Milsom I, Kullendorff AM, Molander U, Bengtsson C. (1999). Acta Obstet Gynecol Scand. 78: 546–551. Swift SE, Ostergard DR. (1995). Obstet Gynecol. 85: 704–708. Symonds T. (2003). Eur Urol. 43: 219–225. Thomas TM, Plymat KR, Blannin J, Meade TW. (1980). Br Med J. 281: 1243–1245.
7
Uemura S, Homma Y. (2004). Neurourol. Urodyn. 23: 94–100. Vandoninck V, Van Balken MR, Finazzi Agro E, Petta F, Caltagirone C, Heesakkers JP, Kiemenev LA, Debruyne FM, Bemelmans BL. (2003). Neurourol Urodyn. 22: 17–23. Wagner TH, Patrick DL, Bavendam TG, Martin ML, Buesching DP. (1996). Urology. 47: 67–71. Wyman JF, Harkins SW, Choi SC, Taylor JR, Fantl JA. (1987). Obstet Gynecol. 70: 378–381. Yalcin I, Patrick DL, Summers K, Kinchen K, Bump RC. (2006). Urology. 67: 1304–1308.
143
8 Overview of Instruments Used to Assess Quality of Life in Dentistry C. McGrath . S. N. Rogers 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146 2 Use of Generic Health-Related Quality of Life Measures in Dentistry . . . . . . . . . . . . . 147 3 Oral Health-Related Quality of Life Measures in Dentistry . . . . . . . . . . . . . . . . . . . . . . . . 149 4 Condition Specific Oral Health-Related Quality of Life Measures in Dentistry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152 5 Assessing the Impact of Oral Health on the Quality of Life of Children . . . . . . . . . . . 154 6 Uses and Future Directions of Health-Related Quality of Life Measures in Dentistry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
#
Springer Science+Business Media LLC 2010 (USA)
146
8
Overview of Instruments Used to Assess Quality of Life in Dentistry
Abstract: In recent times there has been an explosion of interest in assessing the impact of > oral health status on quality of life (QOL) and in assessing the impact of oral health care services and initiatives to QOL within dentistry. This has following developments within other fields of medicine as well as the understanding that oral health is not equated with the absence of oral disease. Clinical oral health status measures have provided little insight into the physical, social and psychological consequences of oral health are of little use in assessing the impact of oral health on QOL. Attempts have been made to use generic health-related quality of life (HQOL) measures which have been used in other fields within medicine to compare the impact on different oral health states on QOL. However, for the most part generic HQOL have not been useful in assessing subtle difference in oral health states or changes to oral health following oral health care interventions except in relation to more severe oral diseases, such as oral cancer. This has led to the development of what are termed generic oral health related quality of life (OHQOL) measures to assess the impact of the range of oral diseases and conditions to QOL. A plethora of OHQOL measures exist and they have generally been more useful in describing the impact of different oral health states to QOL. In assessing the impact of specific oral health states condition specific HQOL have been developed and these have proved useful in describing the impact of specific condition and treatment initiatives on QOL. In many situations the preference of condition specific measure relative to generic HQOL and OHQOL has not been established. More recently advances have been in assessing the impact of oral health on QOL of children which has proved challenging. For the most part attention has focused on developing generic child OHQOL measures. Despite the availability of QOL measure within dentistry there has been little use in clinical practice and limited determination of the effects of different oral health care intervention to QOL. Use and future directions of research of QOL measures within dentistry are provided. List of Abbreviations: CHQ, child health questionnaire; COIDP, child oral impact on daily performance measure; COIHP, child oral health impact profile; CPQ, child perception questionnaire; DIDL, dental impact on daily living; DIP, dental impact profile; ECOHIS, early childhood oral health impact scale; EORTC: HN, European Organization for Research and Treatment of Cancer head and neck questionnaire; EuroQOL, European quality of life measure; FIS, family impact scale; GOHAI, geriatric oral health assessment index; HQOL, healthrelated quality of life; LORQ, Liverpool oral rehabilitation questionnaire; LSI, Liverpool sicca index; MCOHQOL, Michigan child oral health-related quality of life scale; MPDS, Manchester orofacial pain disability; OHIP, oral health impact profile; OHQOL, oral health-related quality of life; OH-QOL UK, United Kingdom oral health-related quality of life measure; OIDP, oral impacts on daily performances; OQLQ, orthognathic quality of life questionnaire; PedsQL, paediatric quality of life inventory; PPQ, parental perception questionnaire; QOL, quality of life; SF12, (health survey) short form 12; SF36, (health survey) short form 36; SIDS, social impacts of dental diseases; SIP, sickness impact profile; SOHSI, subjective oral health status indicator; UWQOL, University of Washington quality of life questionnaire; XeQOLS, xerostomia- related quality of life questionnaire; XI, xerostomia inventory
1
Introduction
This value and use of socio-dental indicator to assess the impact of oral health status on quality of life (QOL) and to assess changes to QOL following oral health care interventions
Overview of Instruments Used to Assess Quality of Life in Dentistry
8
has been advocated since the 1970s (McGrath and Newsome, 2007). This followed the considerable development and widespread use of health status measures (Health-related Quality of Life (HQOL) measures) in medicine on the basis that the extension of people’s life spans and the enhancement of their QOL are two central goals of health care systems. It has long been recognized that clinical “objective” oral health status measures, fail to measure the consequences of oral diseases and conditions and thus provide little insight into the likely physical, social and psychological effects of oral health. The prevailing concept of oral health has evolved from a narrowly reductive perspective, where oral health was equated with the absence of oral disease to a multi dimensional concept and thus the need to incorporate HQOL measures in the assessment of oral health (Locker, 1998). Moreover, there is a growing consensus that oral health states (aside from oral cancer) are rarely if ever life threatening and thus the effects of oral ill health and the consequences of oral health care are largely related to QOL. This has resulted in paradigm shift within dentistry to patient/ client centered assessments and a rapid expansion in the use of HQOL measures to assess oral health needs, describe treatment consequences and in evaluating the outcomes from oral health care interventions (Buck and Newton, 2001).
2
Use of Generic Health-Related Quality of Life Measures in Dentistry
The use of generic HQOL measures in assessing the impact of oral health status on life quality has obvious advantages in that many of these measures have already undergone rigorous psychometric analysis (Reisine, 1996). Moreover, their use in dentistry has the potential to be a useful mechanism for comparing the impact of oral health with other health conditions, and in that way provide a useful tool to describe their relative importance to people’s lives, and also possibly play a role in prioritizing oral health care within health care systems. However, others argue that generic health related quality of life instruments have very limited use in the assessment of the impact of oral health on life quality because they are insufficiently sensitive to measure the more subtle psychosocial impacts of oral problems. Moreover, in one community-base study, oral health represented a dimension of health separate from other health measures and should arguably be seen as a separate construct (Dolan et al., 1991). In contrast, others have described oral health as part of the broader condition of general health and thus the role of generic health-related quality of life measures in dentistry. Commonly employed generic health related quality of life instruments used in dentistry are listed in > Table 8-1. One generic HQOL measure the Sickness Impact Profile (SIP) has received particular attention for being of use in assessing the impact of oral health on life quality (Bergner et al., 1976).
. Table 8-1 Examples of generic health-related quality of life measures used in dentistry References
Name of measure
Bergner et al. (1976)
Sickness Impact Profile (SIP)
Ware and Sherbourne (1992)
Medical Outcome Survey short form (SF36)
EuroQOL Group (1990)
European Quality of Life measure (EuroQOL)
147
148
8
Overview of Instruments Used to Assess Quality of Life in Dentistry
The psychometric properties of the SIP in a wide variety of settings are well documented. In one study population chosen to validate SIP use in relation to oral health, a small convenient non-random group of 152 patients from private dental practices, comprising of 48 > tempromandibular joint dysfunction (TMD) patients, 33 patients with > periodontal disease, 23 denture patients and 48 recall patients (> recall dental patients) were chosen (Reisine et al., 1989). The construct validity of the measure was assessed through relating the domain scores of SIP to the four patient groups. The hypothesis was that recall patients would have the lowest impact scores. The findings indicated that TMD patients experienced a high degree of impact, in particular the domains of well-being, social functioning and symptoms were affected, whereas, recall patients experienced low impact, thus supporting its construct validity. That aside, the principal investigators raised concerns about its sensitivity in relation to less extreme oral health problems and in relation to clinical oral health status such as > dental caries status (number of decayed, missing or filled teeth). Likewise, the sensitivity of the instrument in relation to changes with dental treatment and disease has been questioned. Another generic HQOL of life measure also reputed for its psychometric properties, the Medical Outcome Survey short form SF36 (Ware and Sherbourne, 1992), was utilized in a study to compare its performance with an oral health specific quality of life measure among a group of adults seeking prosthetic care at a UK dental hospital (Allen et al., 1999). It was concluded that a generic oral health related quality of life (OHQOL) measure demonstrated better discriminate and construct validity properties compared with the generic HQOL measure. Furthermore, the SF36 has been reported to be of limited use in determining changes to quality of life following oral implant therapy (Allen and McMillan, 2003). In studies comparing different measurement approaches to assessing the impact dentofacial deformity on QOL it was concluded that the SF36 was unable to distinguish difference in HQOL between those with and without dentofacial deformities and thus it has limited use in dentofacial deformities research or practice (Cunningham et al., 2002). A study evaluating the performance of SF12 in the oral surgery setting (in response to > dento-alveolar surgery) reported that whilst SF-12 may have use in describing the consequences of dento-alveolar surgery it had little value in assessing the need for dento-alveolar surgery or in assessing the outcome from dento-alveolar surgery (McGrath et al., 2003). The EuroQOL was developed by an international research group set up in 1987 to develop a standardized, non-specific instrument for describing and valuing HQOL (EuroQOL, 1991). The EuroQOL EQ-5D provides both a compact descriptive profile and a single index value that can be used in the clinical and economic evaluation of health care. It consist of in twoparts, part 1 is a self-reported description using a five dimensional classification of mobility, self-care, usual activities, pain/discomfort and anxiety/depression. Each dimension has three statements organized hierarchically according to severity and the patient is asked to tick the statement that best describes how they feel “today.” A total of 243 possible health states are defined in this way and each heath state may be converted to a “utility health index” score using tables of values in the EQ-5D user guide and for which the maximum score of 1 indicates the best health state. Part 2 is a self-rated valuation using a visual analogue scale thermometer in which patients’ rate how good or bad their own health is “today.” The best state they could imagine would score 100 and the worst stage imaginable would score 0. The use of EQ-5D among oral and oral pharyngeal cancer suffers has proved useful in understanding how these cancers impact on life quality (Rogers et al., 1996). In addition, EQ-5D proved useful in distinguishing different periodontal health states and thus has potential for describing the burden of different oral disease on QOL (Brennan et al., 2007). Moreover, the EuroQOL is
Overview of Instruments Used to Assess Quality of Life in Dentistry
8
reported to have comparable discriminant validity as a generic OHQOL measure (Brennan and Spencer, 2005). Various other generic HQOL status measures including assessments of pain, anxiety and depression, life satisfaction, adjustment and coping, coherence and social functioning have been used in dental research (McGrath and Newsome, 2007). However, it is beyond the scope of this chapter to describe these in specific detail since the research focus has been on aspect of QOL per se rather than a comprehensive assessment of the impact of oral diseases and conditions to life quality. Thus far the value and use of generic HQOL measures in dentistry is unclear as there are concerns about their ability to distinguish between different oral states and to measure subtle changes in oral health. Moreover, in many cases they perform less well than oral health specific measure at assessing the impact of oral health on quality of life.
3
Oral Health-Related Quality of Life Measures in Dentistry
Since the 1980s a plethora of generic oral health-related quality of life (OHQOL) measures have been developed to asses the impact of oral health status on life quality (Allen, 2003), > Table 8-2. Increasingly these measures have been employed in oral health research as they are valuable tools in describing the physical, social and psychological consequences of oral health states. These measures differ in the their underlying theoretical frameworks (some having none); the dimensions of oral health which they assess (most assesses only the burden of oral disease and deformities); the degree to which they capture the physical, social and psychological aspects of oral health (some focus primarily on symptoms and the physical domain); the number and type of items they contain, as well as differences in scoring methods: difference in response categories and methods of scoring, and some use “weighting” system to provide different weighting for different oral heath experiences (McGrath and Newsome, 2007). It is outside the scope of this chapter to describe all the measures in detail except and so attention will be placed on oral health-related quality of life measures commonly use in dental research and practice and for whom their psychometric properties have been more comprehensively assessed.
. Table 8-2 Examples of currently available generic oral health-related quality of life measures References
Name of measure
Cushing et al. (1986)
Social Impacts of Dental Diseases (SIDS)
Atchison and Dolan (1990)
Geriatric Oral Health Assessment Index (GOHAI)
Strauss and Hunt (1993)
Dental Impact Profile (DIP)
Slade and Spencer (1994)
Oral Health Impact Profile (OHIP)
Locker and Miller (1994)
Subjective Oral Health Status Indicators (SOHSI)
Leao and Sheiham (1996)
Dental Impact on Daily Living (DIDL)
Adulyanon and Sheiham (1997) Oral Impacts on Daily Performances (OIDP) McGrath and Bedi (2001)
United Kingdom Oral health-related Quality of Life mesure (OH-QOL UK)
149
150
8
Overview of Instruments Used to Assess Quality of Life in Dentistry
Perhaps the most sophisticated and most widely used measure has been the Oral Health Impact Profile: OHIP (Slade and Spencer, 1994). The underlying theoretical framework for OHIP is based on Locker’s adaptation of the World Health Organization’s classification of Impairments, Disabilities and Handicaps for measuring oral health (Locker, 1998). This places functional disorders relating to oral health and their social consequences in a hierarchy of outcomes and thus provides a clearer insight into the multidimensional nature of oral health. OHIP is a 49 item measure, with statements divided into seven theoretical domains, namely functional limitation, pain, psychological discomfort, physical disability, psychological disability, social disability and, handicap. A Likert response format (0 = never, 1 = hardly ever, 2 = occasionally, 3 = fairly often, 4 = very often) is used to score the frequency of a problem encountered as a result of oral ill health. Frequency of impacts is calculated by summing the reported negative impacts (i.e., fairly often or very often) across the 49 statements. In addition, the measure allows assessment of perceived severity of each impact, as a weighting method (set-weights) for each statement has been derived using the Thurstone’s paired comparison technique. However, research has questioned the value of the using the weights despite its intuitive appeal (Allen and Locker, 1997). It is unclear as to whether this reflects the known limitation of weighting where there is a large number of items or indeed the particular method of deriving weights which are set weights and thus may not have universal applicability. The validity and reliability of OHIP has been evaluated in a wide range of settings and there is a general consensus that it demonstrates acceptable psychometric properties to support its use in oral health research. Numerous short form measure of OHIP have been developed in an attempt to reduce its number of item which in turn reduces burden on its administration and also to make it more sensitive to specific oral health states using various statistical methods (Allen and Locker, 2002; Slade, 1997; Wong et al., 2007). The short form OHIP-14 consisting of 14 items covering the similar seven theoretical domains as OHIP and which is scored in a similar manner has proved popular (Slade, 1997). However, there are concerns that a key reason of its widespread use has not been because of suitability for use in all scenarios but rather it has been frequently chosen because of the shorter number of items it contains compared to OHIP (Allen and Locker, 2002; Wong et al., 2007). Another commonly used measure is the General Oral Health Assessment Index: (GOHAI) formally known as the Geriatric Oral Health Assessment Index (Atchison and Dolan, 1990). It was developed in an attempt to estimate the degree of psychosocial impact associated with oral diseases in older populations. The items relate to physical function (eating, speaking and swallowing), psychological discomfort (worry, self-consciousness, social interaction) and symptoms (pain, discomfort). Three of the items reflect an assessment of the positive dimension of oral health (“ability to swallow comfortably,” how “pleased they are with their looks” and “being free of discomfort”). Each item can be scored on a Likert scale, in some studies a six point scale has been used with responses ranging from “always” experiencing the effect (score of 5) to “never” experiencing the effect (score 0). Subsequently a five and three point scale has been used in other studies (Atchison, 1997). Before calculating final scores, the responses to nine of the items (the negative impacts of oral health) are reversed, in other words those who respond to “never” score 5, thus allowing a higher final score to represent more positive oral health. Final scores are calculated by summing responses on the Likert scales since it is constructed a single scale of the impact of oral health on life quality. When the six-point scale is used scores range from 0 to 60. The index is not weighted and thus does not provide an assessment of the severity of events. It is reported to have somewhat more preferable qualities than OHIP-14 for use among older and/or frail study populations (Locker et al., 2002a,b).
Overview of Instruments Used to Assess Quality of Life in Dentistry
8
The Oral Impacts on Daily Performance (OIDP) measure was developed to provide an alternative socio-dental indicator which focused on measuring the serious impacts of oral health (Adulyanon and Sheiham, 1997). Its theoretical basis similar to OHIP’s but focused on the disabling and handicap impact of oral diseases and conditions. Items were selected from various socio-medical and > socio-dental indicators utilizing comparison tables of disability indices. Originally nine items relating to physical, psychological and social performances were selected but after pilot testing one item was excluded. The eight items include eating and enjoyment of food, speaking and pronouncing clearly, cleaning teeth, sleeping and relaxing, smiling; laughing and showing teeth without embarrassment, maintain usual emotional state without being irritable, carrying out major work or social role and enjoying contact with people. The items are scored according to frequency and severity. Frequency scores can be classified according to whether the impact is “regular” (those never affected in the past 6 months score 0, less than once a month score 1, once or twice a month score 2, once or twice a week score 3, 3–4 times a week score 4, every or nearly every day score 5). Frequency of effects can also be classified by “spell” patterns, which relates to the length of time people experience impacts (if the duration was for 0 days then the score is 0, for up to 5 days in total the score is 1, for up to 15 days in total the score is 2, for up to 30 days the score is 3, for up to 3 months the score is 4, for over 3 months in total the score is 5). Thus for example a person who twice experienced impacts on eating during the past 6 months for 5 days in total should receive a score of 2 according to period/ spell basis, rather than score of 1 according to regular/ periodic basis. The severity of the impact is also scored from 0 to 5, five representing “very severe” and 0 representing “none.” A total OIDP score can be calculated by multiplying the frequency of the impact score with the severity score. This then can be divided by the maximum possible score (200) to provide a proportional score. However, improvement by multiplying frequency and severity score is of questionable value, given the performance of the frequency scoring on its own (Adulyanon and Sheiham, 1997). The instrument has been shown to have acceptable psychometric properties for use in oral health service research although the reported prevalence of oral health impacts are relative low compared to when other measures of oral health related quality of life presumably because ultimate impacts are rare in most study populations. It is recommended that the preferable method of administration for OIDP is by interview rather than questionnaire based (Robinson et al., 2001). OHIP-14 is reported to perform better than OIDP when compared in different settings (Robinson et al., 2003). The United Kingdom Oral Health-related Quality of Life measure (OHQOL-UK) was developed based on the revised World Health Organization’s conceptual model of health, which reflects both positive (functioning) and negative (disabling) aspects of health status (WHO, 2001). The selection of “items” (questions) for inclusion in the measure was based on the UK public’s perception (a national survey) of the most important ways in which oral health (McGrath and Bedi, 2002). This measure consists of 16-items covering three domains: physical, social and psychological effects. Two forms of OHQOL-UK exist. In the weighted form respondents are asked to rate “what effect, if any, does the condition of your teeth, gums, mouth and/or denture have on your (1 of 16 key areas)?”: Good, none or bad; and then asked to rate “How would you rate the impact of this effect on your overall quality of life?”: None, little, moderate, great or extreme. Summing up responses from individual questions can produces overall OHQOL-UK(W)ß scores ranging from 16 (all bad effects of extreme impact) to 144 (all good effects of extreme impact). An alternative simpler unweighted version exists whereby subjects are asked to rate “What effect does your teeth, gums, mouth and/or false teeth have on each of the 16 key areas?”: “Very bad (score 1), bad (score 2), none (score 3),
151
152
8
Overview of Instruments Used to Assess Quality of Life in Dentistry
good (score 4) or very good (score 5).” Summing up responses from each of the 16-items can produce overall OHQOL-UKß scores ranging from 16 to 80. Domain scores can also be calculated for both the weighted and unweighted version of OHQOL-UK. The instrument has been shown to have acceptable psychometric properties in terms of reliability and validity in both clinical and epidemiological studies (McGrath and Bedi, 2001, 2003).
4
Condition Specific Oral Health-Related Quality of Life Measures in Dentistry
There are a whole range of oral diseases and conditions which affect oral health and thus describing the effect of a particular disease/ condition on QOL has been challenging. To this end a range of condition specific HQOL measures have been developed. The most commonly employed measures in dentistry are outlined in > Table 8-3. This has many potential advantages since the measures capture specifically the subtle difference in different oral health status and
. Table 8-3 Examples of condition specific health-related quality of life measure used in dentistry References
Name of measure
Bjordel et al. (1994)
European Organisation for Research and Treatment of Cancer Head and Neck questionnaire (EORTC: HN)
Thomson et al. (1999)
The Xerostomia Inventory (XI)
Henson et al. (2001)
The Xerostomia- Related Quality of Life Questionnaire (XeQOLS)
Rogers et al. (2002)
University of Washington Quality of life questionnaire (UWQOL)
Cunnigham et al. (2002)
The Orthognathic Quality of Life Questionnaire (OQLQ)
Field et al. (2003)
The Liverpool Sicca Index (LSI)
Pace-Balzan et al. (2004)
The Liverpool Oral Rehabilitation Questionnaire (LORQ)
Aggarwal et al. (2005)
Manchester Orofacial Pain Disability (MPDS)
moreover the subtle changes that different treatment modalities can bring about to a specific oral health status. However, using conditions specific measures limits comparisons to be made across different health status (as well as different oral health states) and makes it difficult to compare the contribution of different oral health care initiatives and therapies to life quality. One particular area where condition specific HQOL measures have been widely use is with respect to oral cancers. These in part reflects the rather different and severe influences oral cancers or cancers of the head and neck, have on QOL compared to the more common oral diseases and conditions. Moreover, within cancer research the use of condition specific cancer measures has been widely used. The European Organization for Research and Treatment of Cancer (EORTC) questionnaire has a specific module to be used in HQOL assessments in head -and neck cancer patients. EORTC head and neck (Bjordal et al., 1994) consists of 35 questions about symptoms and side-effects of treatment, and most of these are scored on a four-point response scale from 1 (not at all) to 4 (very much). The first 18 questions ask about symptoms such as pain, swallowing, taste, and appearance whilst the next 12 questions ask about functions such as
Overview of Instruments Used to Assess Quality of Life in Dentistry
8
eating, talking, social contact and sexuality. The last five “yes/no” questions are about analgesia, supplemental feeding and weight. Another commonly used measure with respect to head and neck cancer has been the University of Washington Quality of life questionnaire (Rogers et al., 2002). The current versions (version 4.0) covers 12 domains – pain, appearance, activity, recreation, swallowing, chewing, speech, shoulder function, taste, saliva, mood and anxiety. Each question is scaled from 0 (worst) to 100 (best) according to the hierarchy of response. There are also two global questions, each a six-point Likert scale, one asking about health-related and the other asking about overall quality of life during “the past 7 days.” The Liverpool Oral Rehabilitation Questionnaire (LORQ) is a health-related quality of life instrument assessing the impact of oral rehabilitation on patients’ HQOL following treatment for oral cancer (Pace-Balzan et al., 2004). Various revisions of the measure have been made; version 3, LORQv3 consists of 40 items divided into 2 primary sections, the first relating to oral function, oro-facial appearance and social interaction, and the second section relating to prostheses and patient denture/prosthetic satisfaction. Items refer to problems or symptoms experienced during the previous week and are rated on a 1–4 Likert scale, from “never” (1) to “always” (4). The LORQv3 has also been used with patients attending general dental practices for routine dental care, patients attending the oral rehabilitation clinic, and in a department of prosthodontics (Pace-Balzan et al., 2008). Another area in which use of condition specific HQOL measures is common relates to > dento-facial deformities. The Orthognathic Quality of Life Questionnaire – OQLQ (Cunningham et al., 2002) consists of twenty-two items contributing to four domains: facial aesthetics, oral function, awareness of dentofacial aesthetics and social aspects of dentofacial deformity and each items is rated on a 4 point scale. The measure has proved more useful than generic HQOL and generic OHQOL measures in describing the effects of dento-facial deformities to life quality (Lee et al., 2008). Other patient centered measures of the combined orthodontic and orthognathic treatment have been developed but less wieldy used. In assessing the impact of orofacial pain on QOL the Manchester Orofacial Pain Disability (MPDS) scale has been developed (Aggarwal et al., 2005). MPODS consist of 32-items of two constructs – physical and psychosocial disabilities. Disability scores have been shown to be higher pain intensity, pain duration and were greater amongst subjects who had sought a health care consultation. Acceptable levels of internal reliability have been reported. To assess the impact of xerostomia (dry-mouth) on life quality various measures have been developed. The Xerostomia Inventory consists of 11 items and an additional 4-items can be used in assessing Burning Mouth Syndrome (Thomson et al., 1999). The Xerostomia-related Quality of Life Questionnaire (XeQOLS) consists of 15 items covering four domains pain/ discomfort, physical functioning, personal/psychological functioning and social functioning (Henson et al., 2001). A measure to assess sicca-related symptoms in patients with primary Sjo¨gren’s syndrome consists of 28 items across five domains has also been developed (Field et al., 2003). Further validation of all measures is required. In other disciplines of dentistry the development of condition specific HQOL measures continue to emerge. It is beyond the scope of this chapter to describe them all in detail but to mention that for the most part the performances of these condition specific measure relative to other established condition specific measure or compared to generic HQOL or generic OHQOL measures has not be adequately assessed. In the end a plethora of patient centered measured are available but there is little evidence as to which measures are preferable to use in specific situations and this has ultimately led to much confusion within the field. This is not to deny that developing new measure can make a contribution to how different oral condition
153
154
8
Overview of Instruments Used to Assess Quality of Life in Dentistry
and their management impact QOL. However, there should be a clear rational behind the development of new measures, their performance should be evaluated and compared with existing measures if they are likely to have any added potential benefit in research and practice.
5
Assessing the Impact of Oral Health on the Quality of Life of Children
Assessing the impact of oral health on QOL of children is a somewhat more complex issue because childhood is a period of immense changes both physical and cognitive (McGrath et al., 2004). It is only in relatively recently times that assessments of HQOL have been conducted among child populations. For the most part it has been deemed necessary to develop child specific measures and for specific age groups of children because of obvious linguistic and cognitive difference between adults and children and among children of different ages. A list of commonly used generic HQOL and Generic OHQOL measures for assessing the impact of oral health on QOL of children is shown in > Table 8-4. . Table 8-4 Examples of health-related quality of life measure used in dentistry among children References
Name of measure
Landgraf et al. (1998)
Child Health Questionnaire (CHQ)
Varni et al. (1999)
Pediatric Quality of Life Inventory (PedsQL)
Jokovic et al. (2002)
Child Perception Questionnaire (CPQ)
Locker et al. (2002a,b)
Family Impact Scale (FIS)
Locker et al. (2003)
Parental Perception Questionnaire (PPQ)
Filstrup et al. (2003)
Michigan Child Oral Health related Quality of Life Scale (MCOHQOL)
Gherunpong et al. (2004)
Child Oral Impacts on Daily Performances (C-OIDP)
Broder and Wilson-Genderson (2007)
Child Oral Health Impact (COHIP) measure
Pahel et al. (2007)
Early Childhood Oral Health Impact Scale (ECOHIS)
In terms of generic HQOL measures the most popular measures have been the Child Health Questionnaire (CHQ). This instrument is comprised of scales specifically developed for children and adolescents 5 years of age and older. The CHQ assesses a child’s physical, emotional, and social well-being from the perspective of a parent or guardian (the 50 item CHQ-PF50 and the 28 item PF-28 (short form) or, in some instances, the child directly (the 87 item CHQ-CF87, for children 10 years of age and older). All CHQ forms yield a 14-concept health status and well-being concepts as well as reliable and valid summary (physical and psychosocial health) scores. Areas measured include: general health, physical functioning, limitations in schoolwork and activities with friends, behavior, mental health, emotional or time impact on the parent, family cohesion, change in health, bodily pain or discomfort, selfesteem and limitations in family activities (Landgraf et al., 1998). CHQ has been employed
Overview of Instruments Used to Assess Quality of Life in Dentistry
8
within dentistry (the 87 item CHQ-CF87 form) in assessing the impact of tempromandibular joint dysfunction (TMD) on the life quality of children (Jedel et al., 2007). Child patients with TMD pain more than once a week reported significantly lower scores in CHQ-CF87 when compared with a control group. Another popular measure has been the Pediatric Quality of Life Inventory (PedsQL) (Varni et al., 1999). Its measurement model is a modular approach to measuring HQOL in children and adolescents. It consists of a 23-item generic core and it has four scales: physical functioning, emotional functioning, social functioning and school functioning. Three summary scores can be derived: total scale score, physical health summary score and psychosocial scale. Different forms of the measure exist for use among children of different ages from 2 to 18. It has been tested in dentistry but preliminary investigations have questioned its value in assessing the impact of early childhood caries on quality of life (Lee et al., 2008). Generic Child OHQOL measures have also been developed (> Table 8-4). The most popular being the Child Oral Health-related Quality of Life measure (COHQOL) comprising the Child Perception Questionnaire (CPQ), the Parental Perception Questionnaire (PPQ) and the Family Impact Scale (FIS) life (Jokovic et al., 2002, 2003; Locker et al., 2002a,b). CPQ assesses children’s own perceptions of the impact of their oral and oro-facial conditions on their OHQOL (Jokovic et al., 2002). There are several forms of the CPQ: CPQ8–10 and CPQ11–14. CPQ8–10 is used among children aged 8–10, and CPQ11–14 is used for children aged 11–14. CPQ11–14 consists of 37 items covering four domains: oral symptom (6 items), functional limitation (9 items), emotional well being (9 items) and social well being (13 items). Each item of the CPQ is scored on a 5-point Likert scale to rate the frequency of occurrence of a particular event. The respond options are “never” = 0, “once or twice” = 1, “sometimes” = 2, “often” = 3, “everyday or almost every” = 4. Scores are calculated by summating respond code of each domain. The possible score of CPQ can range from 0 to 148. A high score represent a poor oral health related quality of life. It has been shown to exhibit acceptable psychometric properties. For younger children a 25 item CPQ8–10 exist covering four domains: oral symptom (5 items), functional limitation (5 items), emotional well being (5 items) and social well being (10 items). Each item of the CPQ8–10 is scored on a 5-point Likert scale to rate the frequency of occurrence of a particular event in the previous 4 weeks in relation to child’s oral condition (Jokovic et al., 2006). The response options can range from “never” (score 0) to “every day or almost every day” (score 4), same with the CPQ11–14. Scores are calculated by summating respond code of each domain. The possible score of CPQ8–10 can range from 0 to 100. A high score represent a poor oral health related quality of life. An analogous questionnaire to elicit parental perceptions of the impact of their children’s oral health on the life quality of their children exists (PPQ). PPQ consists of 31 items covering the similar four domain of CPQ and has been shown to have appropriate validity and reliability (Jokovic et al., 2003). An additional scale the FIS consist of 14 items covering four domains of family life: parental/ family activity (5 items), parental emotions (4 items), family conflict (4 items) and financial burden (1 item) which is scored similar to CPQ and PPQ and too has demonstrated acceptable psychometric properties (Locker et al., 2002a,b). Another generic OHQOL measure for use among children is a child version of OIDP the Child Oral Impacts on Daily Performances (C-OIDP). C-OIDP was developed through modifying the wordings, changing the sequence of questions, simplifying the severity and frequency ratings and shortening the recall period as well as using the pictures to assess ratings of performances (Gherunpong et al., 2004). The authors advocated that the Child-OIDP specifies the different clinical causes of each oral impact and the treatment need. Children are asked to
155
156
8
Overview of Instruments Used to Assess Quality of Life in Dentistry
assess oral impacts on 8 daily performances in the past 3 months. If a child reports an impact on any performance, the frequency and severity of impact on their daily life are scored on a 4-point Likert scale which ranges from 0 to 3. The score of each performance is obtained by multiplying severity and frequency score. Thus, a score of each performance can range from 0 to 9. The overall impact score is obtained by summing the scores of 8 performances divided by 72 (the possible maximum score) and multiplied by 100. The other method of reporting the severity of oral impact is to use the “intensity” and “extent” of impacts. The “intensity” refers to the most severe impacts on any of the 8 performances or the highest performance score. It is classified into six levels: none, very little, little, moderate, severe and very severe. The “extent” refers to the number of performances with impact (PWI) affecting a child’s quality of life over the past 3 months. It can range from 0 to 8 PWI. “Intensity” and “extent” of impacts are more straightforward and use a single score. C-OIDP has been reported to exhibit acceptable psychometric properties. A more recent developed in assessing Child OHQOL has been the Child Oral Health Impact (COHIP) measure (Broder and Wilson-Genderson, 2007). COHIP consists of 34 items forming five conceptually distinct subscales: oral health, functional well-being, social/emotional well-being, school environment and self-image (positive feelings). Responses are recorded for an event occurring in the past 3-months being: “never” = 0, “almost never” = 1, “sometimes” = 2, “fairly often” = 3, and “almost all of the time” = 4. Scoring of the 28 negatively-worded items are reversed. Higher COHIP scores reflect more positive OHRQOL while lower scores reflect lower OHRQOL. Additionally there were two items regarding treatment expectations and one global health perception item used for clinical studies only. COHIP is reported to exhibit excelled reliability and validity (Broder and Wilson-Genderson, 2007). A less commonly used generic measure is the Michigan Child Oral Health related Quality of Life Scale: MCOHQOL (Filstrup et al., 2003) consist of a Child and parental/proxy version for use among children aged 4–16. It consists of 7-itmes covering three domains: pain/discomfort, functioning and psychology. Its psychometric properties are reported to be acceptable. For use among younger children the Early Childhood Oral Health Impact Scale (ECOHIS) has been developed (Pahel et al., 2007). COHIS was developed from an item pool of 45 items and resulted in a final 13-item measure with two scales: Child Impact (9 items) and Family impact (4 items). The developers have suggested that because of the infrequent nature of oral health problems and the young age of children being considered, the parent was asked to consider the child’s entire life span when responding to the questions. Response categories for the ECOHIS were coded: 0 = never; 1 = hardly ever; 2 = occasionally; 3 = often; 4 = very often; 5 = don’t know. ECOHIS scores are calculated as a simple sum of the response codes for the child and family sections separately, after recoding all “Don’t know” responses to missing. The measure has demonstrated acceptable reliability and validity in other settings (Li et al., 2008). As assessments of the impact of oral health on the life quality of children is a relatively recent initiative the use of condition specific measure has yet to emerge.
6
Uses and Future Directions of Health-Related Quality of Life Measures in Dentistry
Quality of life is a dynamic concept because an individual’s perceptions of health and healthrelated quality of life can change over time. Great advances have been made from theoretical standpoints as to how oral health may impact on life quality. Nevertheless for the more part there has been a tendency to focus only on the negative impact oral health on life quality – the burden of oral health despite the evolution towards positive health (WHO, 2001). Failure to
Overview of Instruments Used to Assess Quality of Life in Dentistry
8
include positive dimension may underestimate the importance of oral health to people lives, underestimate the psychosocial impact of oral health, and the improvements to life quality brought about by oral health care interventions (McGrath and Newsome, 2007). Furthermore, there is need to explicitly test how clinical and non clinical variables direct and mediated linkages within the theoretical models of health related quality of life (Baker et al., 2007). Increasingly we live and work in multi-ethnic, multi-cultural and multi-linguistic environments. It is acknowledged that measures of health related quality of life developed in western English-speaking settings have severe limitations for use in different environments. Thus, there is a need to adapt (or develop new) health-related quality of life measures because of language and cultural differences. Furthermore to establish whether instrument developed in one setting and employed in a different setting exhibit consistent findings in the different settings. Within dentistry much research that has derived different language forms of commonly use health related quality of life measures (McGrath and Newsome, 2007). Regrettably these measures for the most part these have been direct translations with little (if any) consideration to ethnic and cultural issues. What is required is a more rigorous approach to trans and cross cultural health related quality of life research. There is little doubt that quality of life data is better obtained from self-reporting than from a proxy as it is more reliable and reduces the possibility of bias from an observers’ subjective internal standards in some instances. Nevertheless, in certain situations language barriers, problems with memory and cognitive development often limits the use of self-report data. For example children are in a sense “moving targets” because childhood is a period with immense changes in psychosocial awareness, and because the children’s dental and facial features change rapidly (McGrath et al., 2003). Thus it is widely accepted; particularly among young children that reliance on proxy report is a more practical and possibly desirable approach. In addition, among old and frail elderly particularly where they suffer from a medical upset that impedes their cognitive ability then caregivers may be a useful alternative in eliciting information about their health related quality of life. What is important to address is whether proxy ratings are alternative or complimentary sources of information and as to who is an appropriate proxy in the assessment process. A significant barrier to their use of health related quality of life measure both in population and clinical studies has been the large number of items in many measures currently available (Allen, 2003). Various methods exist to derive short from measures based on expert approach as well as numerous statistical approaches (including factor analysis and item impact processes). Already advances have been made to derive briefer measures for the more commonly employed measures and they have been shown to offer similar (albeit reduced in many cases) validity and reliability (Allen and Locker, 2002; Slade, 1997; Wong et al., 2007). Unfortunately, whilst the short from measures were derived for a specific entity there ahs been a tendency to use the briefer measures even when the indented use is different to that which the short from measures was intended to be used for. Moreover, while shorter versions have an intuitive appeal, the content validity and reliability of the instruments tends to decrease as items are omitted. Disappointingly there has been a paucity of use of health-related quality of life measures in the clinic setting despite their potential use in assessing needs, prioritizing care and evaluating outcomes for clinical “objective” clinical measure have severe limitations (Rogers et al., 1999; Pace-Balzan et al., 2007; McGrath and Newsome, 2007). Perhaps with the explosion of interest in health related quality of life research briefer (reduced number of items) measures will emerge based on sound theoretical models with acceptable validity and reliability to screen oral health needs and becomes useful tools which are sensitive and responsive to oral health care interventions.
157
158
8
Overview of Instruments Used to Assess Quality of Life in Dentistry
Summary Points There has been considerable interest in assessing the impact of oral health on QOL and the effects of oral health care intervention on improving QOL in recent decades.
Generic HQOL used in other disciplines of medicine have been used within dentistry but
have proven for the most part ineffective except for the more severe forms of oral diseases such as oral cancer. A number of generic OHQOL measures have been developed and several of them have proved popular in use as their validity. There is a move towards the use of condition specific HQOL measures within dentistry. However, for the most part the value of condition specific HQOL over generic HQOL and OHQOL has not been established. In recent times advancements have been made in assessing the impact of oral health status on QOL of children using age specific generic HQOL and OHQOL measures. There has been limited use of QOL measures in dental practice or in assessing outcomes from oral health care interventions.
References Adulyanon S, Sheiham A. (1997). In: Slade DG (ed.) Measuring Oral Health and Quality of Life. University of North Carolina, Chapel Hill. Aggarwal VR, Lunt M, Zakrzewska JM, Macfarlane GJ, Macfarlane TV. (2005). Community Dent Oral Epidemiol. 33(2): 141–149. Allen PF. (2003). Health Qual Life Outcomes. 1: 40–43. Allen PF, Locker D. (1997). Community Dent Health. 14(3): 133–138. Allen PF, Locker D. (2002). Int J Prosthodont. 15(5): 446–450. Allen PF, McMillan AS. (2003). Clin Oral Implants Res. 14(2): 173–179. Allen PF, McMillan AS, Walshaw D, Locker D. (1999). Community Dent Oral Epidemiol. 27(5): 344–352. Atchison KA, Dolan TA. (1990). J Dent Educ. 54(11): 680–687. Baker SR, Pankhurst CL, Robinson PG. (2007). Qual Life Res. 16(2): 297–308. Bergner M, Bobbitt RA, Pollard WE, Martin DP, Gilson BS. (1976). Med Care. 14(1): 57–67. Bjordal K, Ahlner-Elmqvist M, Tollesson E, Jensen AB, Razavi D, Maher EJ, Kaasa S. (1994). Acta Oncol. 33 (8): 879–885. Brennan DS, Spencer AJ. (2005). Community Dent Health. 22(1): 11–18. Brennan DS, Spencer AJ, Roberts-Thomson KF. (2007). J Dent Res. 86(8): 713–717. Broder HL, Wilson-Genderson M. (2007) Community Dent Oral Epidemiol. 35 (Suppl. 1): 20–31.
Buck D, Newton JT. (2001). Community Dent Oral Epidemiol. 29(1): 2–8. Cunningham SJ, Garratt AM, Hunt NP. (2002). Community Dent Oral Epidemiol. 30(2): 81–90. Dolan TA, Gooch BR, Bourque LB. (1991). Community Dent Oral Epidemiol. 19: 1–8. EuroQOL Group. (1990). The Health Policy. 16(3): 199–208. Field EA, Rostron JL, Longman LP, Bowman SJ, Lowe D, Rogers SN. (2003). J Oral Pathol Med. 32(3): 154–162. Filstrup SL, Briskie D, da Fonseca M, Lawrence L, Wandera A, Inglehart MR. (2003). Pediatr Dent. 25 (5): 431–440. Gherunpong S, Tsakos G, Sheiham A. (2004). Community Dent Health. 21(2): 161–169. Henson BS, Inglehart MR, Eisbruch A, Ship JA. (2001). Oral Oncol. 37(1): 84–93. Jedel E, Carlsson J, Stener-Victorin E. (2007). Eur J Pain. 11(5): 557–563. John MT, Patrick DL, Slade GD. (2002). Eur J Oral Sci. 110(6): 425–433. Jokovic A, Locker D, Guyatt G. (2006). Health Qual Life Outcomes. 4: 4–6. Jokovic A, Locker D, Stephens M, Kenny D, Tompson B, Guyatt G. (2002). J Dent Res. 81(7): 459–463. Jokovic A, Locker D, Stephens M, Kenny D, Tompson B, Guyatt G. (2003). J Public Health Dent. 63(2): 67–72. Kressin NR. (1996). J Dent Educ. 60(6): 501–507.
Overview of Instruments Used to Assess Quality of Life in Dentistry Landgraf JM, Maunsell E, Speechley KN, Bullinger M, Campbell S, Abetz L, Ware JE. (1998). Qual Life Res. 7(5): 433–445. Larsson P, List T, Lundstro¨m I, Marcusson A, Ohrbach R. (2004). Acta Odontol Scand. 62(3): 147–152. Lee S, McGrath C, Samman N. (2008). J Oral Maxillofac Surg. 66(6): 1194–1199. Li S, Veronneau J, Allison PJ. (2008). Health Qual Life Outcomes. 1(6): 9–11. Locker D. (1998). Community Dent Health. 5(1): 3–18. Locker D, Jokovic A, Stephens M, Kenny D, Tompson B, Guyatt G. (2002a). Community Dent Oral Epidemiol. 30(6): 438–448. Locker D, Matear D, Stephens M, Jokovic A. (2002b). Community Dent Health. 19(2): 90–97. McGrath C, Bedi R. (2001). Community Dent Health. 18(3): 138–143. McGrath C, Bedi R. (2002). Community Dent Health. 19(4): 211–214. McGrath C, Bedi R. (2003). J Public Health Dent. 63(2): 73–77. McGrath C, Broder H, Wilson-Genderson M. (2004). Community Dent Oral Epidemiol. 32(2): 81–85. McGrath C, Comfort MB, Lo EC, Luo Y. (2003). J Oral Maxillofac Surg. 61(7): 759–763. McGrath C, Newsome PR. (2007). Dent Update. 34(1): 41–42. Pace-Balzan A, Butterworth CJ, Dawson LJ, Lowe D, Rogers SN. (2008). J Prosthet Dent. 99(3): 233–242. Pace-Balzan A, Cawood JI, Howell R, Lowe D, Rogers SN. (2004). J Oral Rehabil. 31(6): 609–617.
8
Pahel BT, Rozier RG, Slade GD. (2007). Health Qual Life Outcomes. 1(5): 6–8. Reisine S. (1996). J Dent Educ. 60(6): 488–493. Reisine ST, Fertig J, Weber J, Leder S. (1989). Community Dent Oral Epidemiol. 17(1): 7–10. Robinson PG, Gibson B, Khan FA, Birnbaum W. (2001). Community Dent Health. 18(3): 144–149. Robinson PG, Gibson B, Khan FA, Birnbaum W. (2003). Community Dent Oral Epidemiol. 31(2): 90–99. Rogers SN, Gwanne S, Lowe D, Humphris G, Yueh B, Weymuller EA Jr. (2002). Head Neck. 24(6): 521–529. Rogers SN, Miller RD, Ali K, Minhas AB, Williams HF, Lowe D. (2006). Int J Oral Maxillofac Surg. 35(10): 913–919. Slade GD. (1997). Community Dent Oral Epidemiol. 25(4): 284–290. Slade GD, Spencer AJ. (1994). Community Dent Health. 11(1): 3–11. Thomson WM, Chalmers JM, Spencer AJ, Williams SM. (1999). Community Dent Health. 16(1): 12–17. Varni JW, Seid M, Rode CA. (1999). Med Care. 37(2): 126–139. Ware JE, Sherbourne CD. (1992). Med Care. 30(6): 473–448. Wong AH, Cheung CS, McGrath C. (2007). Community Dent Oral Epidemiol. 35(1): 64–72. World Health Organisation. (2001). International Classification of Functioning, Disabilities and Health. World Health Organisation, Geneva.
159
9 SQOR-V: A Patient Reported Outcome Specifically Dedicated to Chronic Venous Disorders J.-J. Guex . S. E. Zimmet . S. Boussetta . C. Taieb 1
Introduction: Chronic Venous Disorders (CVD), Definition . . . . . . . . . . . . . . . . . . . . . . 162
2
Specificities of Chronic Venous Disorders (CVDs) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
3
Why Is QOL So Important for CVD Assessment? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
4
Validation of the SQOR-V . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
5 Use of the SQOR-V . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164 5.1 Calculation of the SQOR-V Questionnaire Score . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165 5.2 Managing Missing Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171 6
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
#
Springer Science+Business Media LLC 2010 (USA)
162
9
SQOR-V: A Patient Reported Outcome Specifically Dedicated to Chronic Venous Disorders
Abstract: Chronic venous disorders (> CVD) are recognized on a number of clinical signs of variable severity, from telangiectasias to skin ulceration of the lower legs. Their symptomatology is highly variable as well. Evaluation of the global severity of the disease remains difficult, except when considering late stages and complications, which represent only a vary small group of patients. No biological or instrumental marker exists which could be correlated to improvement or worsening whether spontaneous or after treatment. All these characteristics of CVD explain why Patient Reported Outcome (> PRO) are becoming a popular part of phlebological research. The SQOR-V is a validated PRO designed for Chronic Venous Disorders (CVD). It allows a relevant and sensitive assessment of clinical features and quality of life of patients at all stages of Chronic Venous Disorders (CVD). It can be used in epidemiological and clinical trials. It is sensitive enough to assess CVD at early stages of the disease. Using the SQOR-V in clinical trials is possible for any kind of treatment for CVD: drugs, compression, surgery, or endovenous ablation. It adds a composite global assessment complementary to hemodynamic measurements. The SQOR-V can be used freely; provided it is applied according to its construction rules and that its authors are informed of the intended use. More studies are needed to validate versions in more languages and to better refine its practical use. List of Abbreviations: CVD, chronic venous disorders; QOL, quality of life; PRO, patient reported outcome
1
Introduction: Chronic Venous Disorders (CVD), Definition
CVD stands now for disorders when it previously stood for diseases. This term includes all manifestations related to a chronic impairment of the venous function, i.e., the return of blood to the heart. It can be caused by blood reflux (valvular incompetence), outflow obstruction, vein wall compliance reduction, venous wall cells dysfunction, and probably other mechanisms. Venous abnormalities can affect superficial, deep and perforating venous networks. CVD can be congenital, primary or secondary (=acquired, especially post-thrombotic). Chronic venous Insufficiency (CVI) refers to severe stages where tissular decompensation occurs, especially at skin or sub-cutaneous level.
2
Specificities of Chronic Venous Disorders (CVDs)
Clinical presentation of Chronic Venous Disorders (CVD) is extremely variable and no straight correlation can be observed between signs, and/or symptoms, and/or hemodynamic anomalies. Very large varicose veins can be completely painless for years – until occurrence of complications – when some patients experience early severe symptoms while lacking any visibly dilated vein (varicose veins are a sign of chronic venous disease). From the patient’s point of view, CVD can impair many different aspects of their life. CVD can be responsible for pain and various discomforts, poor cosmetic aspect, complications such as thrombosis or ulcers, reduction of social activities, and can be considered as a threat for their health, especially in case of severe family history of CVD (ulcers, superficial thrombo-phlebitis). From the instrumental point of view, the assessment of CVD is possible by Level 2 (noninvasive) investigations such as Duplex scanning (detecting and describing anatomic and hemodynamic abnormalities) or plethysmograms (evaluating functional impairment through dynamic limb volume measurement). However, due to the lack of correlation with clinical findings, no single method is sufficient or reliable enough to estimate the actual severity of the disease.
SQOR-V: A Patient Reported Outcome Specifically Dedicated to Chronic Venous Disorders
9
Finally, CVD are fortunately almost always non lethal and the incidence of severe events (ulcers, hemorrhage, pulmonary emboli, amputation) is delayed and limited, evaluation based on these outcomes is only possible in late (and less frequent) cases. Therefore, unlike many other diseases (for example arterial diseases where the ankle/ brachial index provides a simple method of assessment) there is no simple outcome usable to assess worsening or improvement of CVD, spontaneously or after treatment. Establishing the success of a CVD treatment has proven difficult since patient’s satisfaction may not be correlated with the improvement of hemodynamic or appearance. The only sure thing is that if the patient is not satisfied by the results, it means that the treatment has failed. Other types of failures are also possible, some of which are not detected by the patient (and/or the physician), and may sometimes lead to an early recurrence. In order to refine clinical evaluation, venous clinical severity score (> VCSS) and venous disability score (VDS) have been associated to the CEAP Classification (Eklo¨f, 2006). All are assessed by the physician and not by the patient; though used in several studies, they are still under revision and not appropriate for early stages of the disease.
3
Why Is QOL So Important for CVD Assessment?
Recently, like in many other medical fields (e.g., Haibel et al., 2004), more studies have used Patient Reported Outcomes (PRO) to assess some aspects of the severity of the disease (Guex et al., 2005, 2007). Currently used QOL scales are:
either generic: like SF 12 (Ware et al., 1996) and SF 36 (Ware and Sherbourne, 1992) or specific: Of venous symptoms and signs like the Edinburgh (Smith et al., 1999), CIVIQ (Launois et al., 1996) and > VEINES-QOL (Kahn et al., 2006) questionnaires
Of daytime sleepiness like the > EPWORTH questionnaire (Miletin and Hanly, 2003) Or of depression like the > CES-D (Fuhrer and Rouillon, 1989) Generic and specific questionnaires are generally scored in opposite ways: the higher the value of the generic questionnaire score, the better the health status, and conversely; the higher the value of the specific score the worst the QOL (> Figure 9-1). CVD can be described using the CEAP classification (Eklof, 2006; Porter and Moneta, 1995). The CEAP uses four descriptors (the four letters of the acronym: C (clinical), E (etiology), A (anatomy), P (pathophysiology). Each descriptor is delineated by a figure and/or a letter as indicated in > Table 9-1. In fact, CEAP is now the international standard for description of CVD. For example, a patient with uncomplicated symptomatic primary varicose veins of the great saphenous territory will be classified [C2s, Ep, As, PR2,3], and a patient with a painful post thrombotic (popliteal incompetence) active ulcer will be classified [C6s, Es, Ad, PR14]. However, CEAP classification is not an evaluation tool; it only describes the type of CVD, and does not evaluate its severity. Focusing the evaluation of CVD on clinical features, including all known patients’ complaints, through a composite, sound and understandable PRO is the most logical option. However, especially when evaluating efficacy of treatments, anatomical and hemodynamic modifications (suppression of veins, correction of reflux, improvement of outflow, etc. . .) must also be measured and reported. But the clinical appraisal, as reported by patients, should be the predominant method of assessing therapeutic success or failure.
163
164
9
SQOR-V: A Patient Reported Outcome Specifically Dedicated to Chronic Venous Disorders
. Figure 9-1 Variation of QOL score: difference between specific and generic questionnaires
We have developed the SQOR-V after long discussions among vein experts (Guex et al., 2007) and careful analysis of existing venous PRO, to which we acknowledge precedence and value. However, certain QOL questionnaires are not without cost, and we wanted to improve both the relevance of the questions and simplicity of the survey. The SQOR-V may be used without charge.
4
Validation of the SQOR-V
Field testing and validation of the SQOR-V PRO were carried out in France (in French) and have been reported (Guex et al., 2007). Languages for which we have obtained a cultural and linguistic validation (at proof review time, more in progress): French, English, Czech, Spanish (Spain and Argentina), Italian, Afrikaans, South African English. Two hundred and two questionnaires were analyzed after a first application and 152 after a second (test-retest). The Cronbach’s alpha coefficient was calculated at 0.96 and the structural analysis demonstrated excellent internal and structural coherence. Reproducibility has been verified by test-retest with a correlation of 0.80 on the global score. Clinical validity and convergent validity were established by comparison with symptoms, and with two PRO: SF12 and CES D (Fuhrer and Rouillon, 1989; Ware et al., 1996). Unlike severity scores like the VCSS, SQOR-V is able to demonstrate a satisfactory sensitivity between lower classes of the CEAP, with mean values of 38.89, 42.24, and 48.71 for classes C1, C2 and C3 respectively.
5
Use of the SQOR-V
Like other QOL questionnaires, the SQOR-V is auto administered and patients are left alone for answering questions. Questions must not be discussed with the physician. Minimal
SQOR-V: A Patient Reported Outcome Specifically Dedicated to Chronic Venous Disorders
9
. Table 9-1 Outline of CEAP classification of chronic venous disorders C Clinical signs
0: No sign of venous disease 1: Reticular veins and telangiectases 2: Varicose veins 3: Edema 4: Skin changes 4a: Eczema, pigmentation 4b: Lipodermatosclerosis, ‘‘atrophie blanche’’ 5: Healed ulcer 6: Active ulcer
C Clinical symptoms
A: Asymptomatic
E Etiology
P: Primary
S: Symptomatic S: Secondary C: Congenital N: No venous etiology identified A Anatomy
S: Superficial D: Deep P: Perforating veins N: No venous location identified
P Pathophysiology
R: Reflux (with # s indicating pathologic segment(s)) O: Obstruction RO: Reflux and obstruction N: No venous pathophysiology identifiable
Nomenclature of diseased vein segments: Superficial: 1. Telangiectases & reticular veins; 2. Great saphenous vein above knee; 3. Great saphenous vein below knee; 4. Small saphenous vein; 5. Non saphenous veins. Deep: 6. Inferior vena cava; 7. Common iliac vein; 8. Internal iliac vein; 9. External iliac vein; 10. Pelvic veins; 11. Common femoral vein; 12. Deep femoral vein; 13. Femoral vein; 14. Popliteal vein; 15. Deep leg veins; 16. Muscular veins. Perforating Veins: 17. Thigh; 18. Leg. For example, a patient with uncomplicated symptomatic primary varicose veins of the great saphenous territory will be classified [C2s, Ep, As, PR2,3], and a patient with a painful post thrombotic (popliteal incompetence) active ulcer will be classified [C6s, Es, Ad, PR14]
interaction provides better reliability. If questions are not understood it is better to leave the item blank (see ‘‘missing data’’ below) rather than completed with the physician’s or staff ’s help. The usual duration for questionnaire completion is less than 10 min.
5.1
Calculation of the SQOR-V Questionnaire Score
The SQOR-V questionnaire (Specific Quality of Life and Outcome Response – Venous) scoring comprises 45 items (> Figures 9-2 and > 9-3). Each of these items is rated from 1 to 5 (1 = None to 5 = Extremely). This questionnaire has been developed to explore five
165
166
9
SQOR-V: A Patient Reported Outcome Specifically Dedicated to Chronic Venous Disorders
. Figure 9-2 Questionnaire in French
SQOR-V: A Patient Reported Outcome Specifically Dedicated to Chronic Venous Disorders
. Figure 9-2 (continued)
9
167
168
9
SQOR-V: A Patient Reported Outcome Specifically Dedicated to Chronic Venous Disorders
. Figure 9-2 (continued)
broad dimensions affected by venous insufficiency, and each of the 45 items covers only one dimension:
Discomfort – annoyance – pain: 22 items Appearance – aesthetic aspect: 3 items Risk – threat to health: 4 items Restriction in movements – activities: 12 items Emotional problems: 4 items
The relationship between the items and the dimension that they represent is summarized in 9-2 and > 9-3. We have adopted a method of ‘‘normalization’’ by bringing each dimension back to 20, in order to facilitate interpretation of the dimension scores. The total score is then the sum of the values shown for each dimension. This total score will thus have a maximum of 100. Therefore, each dimension has the same impact on the global score, which serves to clarify the impact of each dimension. In order to refine the interpretation of the score and to make its usage more specific, we have created two impacts resulting from the five dimensions. One is a psychosomatic component and the other a physical component:
> Tables
Physical impact: resulting from the sum of the scores of the dimensions ‘‘discomfort,’’ ‘‘restriction in movement’’ and ‘‘risk,’’ with a total score of 100.
Psychosomatic impact: resulting from the sum of the scores of the dimensions ‘‘appearance’’ and ‘‘emotional problems,’’ with a total score of 100. The total score as well as the two impacts can thus vary from 20 to 100. The higher the score, the worse the quality of life. A reduction in the score or scores reflects an improvement in the quality of life.
SQOR-V: A Patient Reported Outcome Specifically Dedicated to Chronic Venous Disorders
. Figure 9-3 Questionnaire in English
9
169
170
9
SQOR-V: A Patient Reported Outcome Specifically Dedicated to Chronic Venous Disorders
. Figure 9-3 (continued)
SQOR-V: A Patient Reported Outcome Specifically Dedicated to Chronic Venous Disorders
9
. Figure 9-3 (continued)
5.2
Managing Missing Data
In order to avoid considerable loss of data, we have used a system for replacing missing information. This involves giving each missing dimension value the average value of this dimension obtained from the other patients in the study. Important Statement "
The authors allow the USE of the SQOR-V provided it is used with the above mentioned protocol and provided it is used in validated languages. Such validations are encouraged. Before constructing their studies, potential SQOR-V users are required to inform SQOR-V authors of their purpose and methodology. After completing their study, they should communicate their results and indicate where and when their study will be published or presented.
171
172
9
SQOR-V: A Patient Reported Outcome Specifically Dedicated to Chronic Venous Disorders
. Table 9-2 Items, dimensions, impacts in the SQOR-V, in French Nume´ro item
Items
Dimension
Valeur
Impact
1
Geˆne global (jambe Gauche)
2
Douleur (jambe Gauche)
Inconfort
1–5
Physique
3
Lourdeur (jambe Gauche)
Inconfort
1–5
Physique
Inconfort
1–5
Physique
4
De´mangeaisons (jambe Gauche)
Inconfort
1–5
Physique
5
Crampes nocturnes (jambe Gauche)
Inconfort
1–5
Physique
6
Gonflement (jambe Gauche)
Inconfort
1–5
Physique
7
Sensation de chaleur (jambe Gauche)
Inconfort
1–5
Physique
8
Picotements (jambe Gauche)
Inconfort
1–5
Physique
9
Elancements (jambe Gauche)
Inconfort
1–5
Physique
10
Jambes sans repos (jambe Gauche)
Inconfort
1–5
Physique
11
Aggravation avec la chaleur (jambe Gauche)
Inconfort
1–5
Physique
12
Geˆne global (jambe Droite)
Inconfort
1–5
Physique
13
Douleur (jambe Droite)
Inconfort
1–5
Physique
14
Lourdeur (jambe Droite)
Inconfort
1–5
Physique
15
De´mangeaisons (jambe Droite)
Inconfort
1–5
Physique
16
Crampes nocturnes (jambe Droite)
Inconfort
1–5
Physique
17
Gonflement (jambe Droite)
Inconfort
1–5
Physique
18
Sensation de chaleur (jambe Droite)
Inconfort
1–5
Physique
19
Picotements (jambe Droite)
Inconfort
1–5
Physique
20
Elancements (jambe Droite)
Inconfort
1–5
Physique
21
Jambes sans repos (jambe Droite)
Inconfort
1–5
Physique
22
Aggravation avec la chaleur (jambe Droite) Inconfort
1–5
Physique
23
Apparence global de la jambe droite affecte´e par les proble`mes veineux
Apparence
1–5
Psychosomatique
24
Apparence global de la jambe gauche affecte´e par les proble`mes veineux
Apparence
1–5
Psychosomatique
25
Proble`mes veineux conditionnent le choix Apparence des veˆtements
1–5
Psychosomatique
26
Proble`mes veineux conditionnent le choix Restriction des de vos activite´s mouvements
1–5
Physique
27
Restriction globale
Restriction des mouvements
1–5
Physique
28
Activite´s professionnelles
Restriction des mouvements
1–5
Physique
29
A la maison
Restriction des mouvements
1–5
Physique
SQOR-V: A Patient Reported Outcome Specifically Dedicated to Chronic Venous Disorders
. Table 9-2 (continued) Nume´ro item
Items
Dimension
Valeur
9
Impact
30
Activite´s de loisirs ou sportives
Restriction des mouvements
1–5
Physique
31
Station debout prolonge´e
Restriction des mouvements
1–5
Physique
32
Position assise prolonge´e
Restriction des mouvements
1–5
Physique
33
Lors de la marche
Restriction des mouvements
1–5
Physique
34
Utilisation de la marche
Restriction des mouvements
1–5
Physique
35
Au cours du sommeil
Restriction des mouvements
1–5
Physique
36
Activite´s sociales
Restriction des mouvements
1–5
Physique
37
Relations intimes ou sexuelles
Restriction des mouvements
1–5
Physique
38
Proble`mes veineux vous inquie`tent-ils?
Menace – Risque
1–5
Physique
39
Aggravation de la maladie veineuse vous inquie`tent-elle?
Menace – Risque
1–5
Physique
40
Complication de la maladie veineuse vous Menace – inquie`tent-elle? Risque
1–5
Physique
41
Le fait qu’un de vos proche souffre de maladie veineuse vous inquie`tent-il?
Menace – Risque
1–5
Physique
42
Conse´quences e´motionnelles globales
Proble`mes e´motionnels
1–5
Psychosomatique
43
A cause de mon proble`me veineux, je suis Proble`mes a` cran e´motionnels
1–5
Psychosomatique
44
A cause de mon proble`me veineux, je suis Proble`mes irritable e´motionnels
1–5
Psychosomatique
45
A cause de mon proble`me veineux, impression d’eˆtre un fardeau pour les autres
1–5
Psychosomatique
6
Proble`mes e´motionnels
Conclusion
Use of SQOR-V in studies of CVD is simple, fast and reliable. It brings a sound and sensitive assessment of patients’ clinical features and QOL, even in patients with early (limited) symptoms and signs, which increases the relevance and utility for both epidemiologic (descriptive), and comparative studies (see > Figure 9-4).
173
174
9
SQOR-V: A Patient Reported Outcome Specifically Dedicated to Chronic Venous Disorders
. Table 9-3 Items, dimensions, impacts in the SQOR-V, in English (suggested wording) Item number
Items
Dimension
Value
Impact
1
Overall discomfort (left leg)
Discomfort
1–5
Physical
2
Pain (left leg)
Discomfort
1–5
Physical
3
Heaviness (left leg)
Discomfort
1–5
Physical
4
Itching (left leg)
Discomfort
1–5
Physical
5
Night cramps (left leg)
Discomfort
1–5
Physical
6
Swelling (left leg)
Discomfort
1–5
Physical
7
Warm or burning sensation (left leg)
Discomfort
1–5
Physical
8
Tingling (left leg)
Discomfort
1–5
Physical
9
Stinging or stabbing sensation (left leg)
Discomfort
1–5
Physical
10
Restless legs (left leg)
Discomfort
1–5
Physical
11
Worse with heat (left leg)
Discomfort
1–5
Physical
12
Overall discomfort (right leg)
Discomfort
1–5
Physical
13
Pain (right leg)
Discomfort
1–5
Physical
14
Heaviness (right leg)
Discomfort
1–5
Physical
15
Itching (right leg)
Discomfort
1–5
Physical
16
Night cramps (right leg)
Discomfort
1–5
Physical
17
Swelling (right leg)
Discomfort
1–5
Physical
18
Warm or burning sensation (right leg)
Discomfort
1–5
Physical
19
Tingling (right leg)
Discomfort
1–5
Physical
20
Stinging or stabbing sensation (right leg)
Discomfort
1–5
Physical
21
Restless legs (right leg)
Discomfort
1–5
Physical
22
Worse with heat (right leg)
Discomfort
1–5
Physical
23
Overall appearance of your right leg affected by vein problems
Appearance
1–5
Psychosomatic
24
Overall appearance of your left leg affected by Appearance vein problems
1–5
Psychosomatic
25
Vein problems impacting clothing chosse
Appearance
1–5
Psychosomatic
26
Vein problems impacting activities chosse
Restriction in movements
1–5
Physical
27
Overall restriction
Restriction in movements
1–5
Physical
28
At work
Restriction in movements
1–5
Physical
29
At home
Restriction in movements
1–5
Physical
30
Sport or leisure activities
Restriction in movements
1–5
Physical
9
SQOR-V: A Patient Reported Outcome Specifically Dedicated to Chronic Venous Disorders
. Table 9-3 (continued) Item number
Items
Dimension
Value
Impact
31
Prolonged standing
Restriction in movements
1–5
Physical
32
Prolonged sitting
Restriction in movements
1–5
Physical
33
When walking
Restriction in movements
1–5
Physical
34
When using stairs
Restriction in movements
1–5
Physical
35
During sleep
Restriction in movements
1–5
Physical
36
Social activities
Restriction in movements
1–5
Physical
37
Intimate or sexual relations
Restriction in movements
1–5
Physical
38
Overall, do your vein problems worry you?
Risk – threat to health
1–5
Physical
39
Does the possible worsening of your vein disease worry you?
Risk – threat to health
1–5
Physical
40
Does the possibility of your condition causing Risk – threat to complications worry you? health
1–5
Physical
41
Does it worry you that someone related to you Risk – threat to suffers from vein disease? health
1–5
Physical
42
Overall emotional consequences
Emotional problems
1–5
Psychosomatic
43
Because of my vein problems, I am on edge
Emotional problems
1–5
Psychosomatic
44
Because of my vein problems, I am irritable
Emotional problems
1–5
Psychosomatic
45
Because of my vein problems, I feel like I am a Emotional burden to others problems
1–5
Psychosomatic
Summary Points The classification of CVD (CEAP classification) provides a detailed and relevant description but is not intended to serve as an evaluation tool.
Evaluation of CVD severity at early stages is mostly clinical and robust outcomes, such as skin changes and ulcers appear only late. Evaluation through a PRO allows a more precise assessment. Severity scores proposed in CVD are not quality of life questionnaires, nor PRO. They are physician reported and specific of late, severe stages of the disease. They are appropriate for Chronic Venous Insufficiency.
175
176
9
SQOR-V: A Patient Reported Outcome Specifically Dedicated to Chronic Venous Disorders
. Figure 9-4 Example of use of the SQOR-V (experimental study, not published yet). Variation of SQOR-V value according to Body Mass Index (BMI) in the general population (776 persons): (Higher SQOR-V global score means worse status)
SQOR-V: A Patient Reported Outcome Specifically Dedicated to Chronic Venous Disorders
9
SQOR-V provides with a multidimensional evaluation which needs to be associated to
instrumental assessments of hemodynamic abnormalities (plethysmography, duplex ultrasound). Evaluation of treatments cannot rely only on hemodynamic appraisal, even if they are absolutely necessary.
References Eklo¨f B. (2006). Classifying venous disease In: Bergan JJ (ed.) The Vein Book. Elsevier, San Diego. Fuhrer R, Rouillon F. (1989). Psychiatrie et Psychobiologie. 4:163–166. Guex JJ, Myon E, Didier L, Nguyen Le C, Taieb C. (2005). Int Angiol. 24: 258–264. Guex JJ, Zimmet SE, Boussetta S, Nguyen Le C, Taieb C. (2007). J Mal Vasc. 32: 135–147. Haibel H, Niewerth M, Brandt J, Rudwaleit M, Listing J, Sieper J, et al. (2004). Z Rheumatol. 63(5): 393–401. Kahn S, Lamping D, Ducruet T, Arsenault L, Miron M, Roussin A, et al. (2006). J Clin Epidemiol. 59: 1049–1056.
Launois R, Reboul-Marty J, Henry B. (1996). Qual Life Res. 5(6): 539–554. Miletin MS, Hanly PJ. (2003). Sleep Med. 4(3):195–199. Porter JP, Moneta GM. (1995). J Vasc Surg. 21:635–645. Smith JJ, Garratt AM, Guest M, Greenhalgh RM, Davies AH. (1999). J Vasc Surg. 30(4): 710–719. Ware JE Jr, Kosinski M, Keller SD. (1996). Med Care. 34: 220–233. Ware JE Jr, Sherbourne CD. (1992). Med Care. 30(6): 473–483.
177
10 The Uniscale Assessment of Quality of Life: Applications to Oncology E. Ballatori . F. Roila . B. Ruggeri . A. A. Bruno . S. Tiberti . F. di Orio 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180 2 Problems with the use of Psychometric Questionnaires . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181 3 The Uniscale Assessment of Quality of Life: An Italian Experience . . . . . . . . . . . . . . . . 183 4 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
#
Springer Science+Business Media LLC 2010 (USA)
180
10
The Uniscale Assessment of Quality of Life: Applications to Oncology
Abstract: The measurement of quality of life (QOL) is an important challenge in clinical research, not only because QOL is one of the two main endpoints of effectiveness of treatments, the other being overall survival, but also for taking clinical decisions shared with the patient. Today, QOL is generally assessed using psychometric questionnaires; nevertheless, these latter suffer from several shortcomings that often lead to unreliable results. About 14 years ago, the Italian Group for the Evaluation of Outcomes in Oncology (IGEO) planned a research program articulated in two phases. In the first phase, > domains of QOL and problems connected with it were defined performing a content analysis of the interviews of 248 Italian cancer patients, based on four areas related to the foundations of quality of life. The domains/problems referred by the patients were submitted to a large population of more than 6,000 Italian cancer patients so as to assign a frequency to the relevance of each domain and to the presence/absence of each problem. In this study, a uniscale evaluation of QOL, using a Visual Analogue Scale (> VAS), was also obtained. The rating of each patient was classified in ‘‘bad QOL’’ and ‘‘good QOL,’’ if the chosen point fell in the 0–30 or 70–100 interval, respectively; the other scores (30–70 interval) were not considered. The relationship between the uniscale assessment of QOL and the presence of each problem was investigated. The impact of each problem, adjusted for the presence of the others, and for the patient’s characteristics, was detected by a unifactorial analysis using logistic additive models, where ‘‘good’’ and ‘‘bad’’ QOL were in turn considered as dependent variables. Thirteen of 19 problems were significant, and this confirms the external validity of the uniscale assessment of QOL. AVAS can be considered a reference point in multidimensional QOL scales and should still be regarded as a useful and synthetic tool to investigate phenomena related to the patient’s QOL. In this perspective, more research on the psychometric properties of this instrument, in the context of the assessment of QOL, is still needed. List of Abbreviations: CRF, Case Record Form; > HRQL, Health-Related Quality of Life; IGEO, Italian Group for the Evaluation of Outcomes in Oncology; > KPS, Karnofsky Performance Status; > LP, Linear Predictor; NED, No Evidence of Disease; NHS, National Health System; OR, > Odds Ratio; QOL, Quality of Life; RR, > Relative Risk; SD, Standard Deviation; VAS, Visual Analogue Scale
1
Introduction
One of the most useful working definitions of quality of life is that reported by David Cella (Cella and Tulsky, 1990): ‘‘Quality of life refers to patients’ appraisal of and satisfaction with their current level of functioning as compared to what they perceive to be possible or ideal.’’ In fact, this dynamic definition incorporates the patient’s strategies of coping with changed health conditions due to the progression of disease. Moreover, it makes clear that HealthRelated Quality of Life (HRQL) is the goal of the evaluation. HRQL can be considered as an important endpoint in several clinical trials in oncology, especially when therapies giving a similar survival rate are compared, or when the survival rates are expected to be different but the treatment that assures the longest survival is the same that offers the worst HRQL. Its importance is not only in evaluating the effectiveness of alternative medical interventions (HRQL is one of the two efficacy endpoints in comparative clinical trials, the other being overall survival), but also in screening individual patients for
The Uniscale Assessment of Quality of Life: Applications to Oncology
10
possible psycho-social intervention, in monitoring the quality of care (Aaronson, 1990), and in making clinical decisions shared with the patient. Regardless of the type of instrument used, in assessing HRQL there are some points on which there is general agreement (Aaronson, 1990): 1. Self-assessment, because of the subjectivity of HRQL. 2. HRQL should be measured using a tool able to give quantitative responses; these measurements should be obtained repeatedly, so as to evaluate the variations of HRQL when a patient’s health condition changes. 3. HRQL has a multidimensional framework. A review of literature suggests that HRQL encompasses at least the following components (or domains, or dimension): physical functioning status, disease symptoms and treatment side effects, psychological status, social functioning (Aaronson, 1990). The HRQL of cancer patients is today generally assessed using a psychometric questionnaire able to explore each of several domains by means of one or more items. When the patient repeatedly fills it out, the level of each domain at any time and their modification during follow-up can be evaluated. A great number of papers on the evaluation of HRQL of cancer patients have been published during the last 30 years, but the majority of them were concerned with the > reliability and validity of the measurement instrument (Ballatori, 2001). More than 600 different quality of life questionnaires have been validated; the reasons for this are mainly the following: 1. The complexity of the problem. 2. The lack of a generally accepted theory. 3. The difficulties to obtain results that are useful in daily clinical practice. Another tool able to measure HRQL is the uniscale assessment, widely used in early studies (Coates et al., 1990; Hiratsuka and Kida, 1993). Generally it consists in a 100-mm Visual Analogue Scale (VAS) where the patient reports his/her response to the following question: ‘‘How would you rate your quality of life today?’’ (Aaronson, 1990). The VAS for the evaluation of HRQL mainly presents two shortcomings: (1) an unsatisfactory reproducibility, and (2) giving a global measure, it yields only a summary score, insufficient to understand the reasons for its variations. On the other hand, serious problems arise when using a psychometric questionnaire; they are discussed in the next section. Aims of this article are to obtain a further proof of validity of the uniscale assessment of HRQL and to stimulate a discussion on future perspectives.
2
Problems with the use of Psychometric Questionnaires
Different problems arise both in measuring HRQL and evaluating the results of an assessment of HRQL in cancer patients. They involve the patient, questionnaire, setting of administration, and their interactions. 1. Selection effects: A questionnaire should be self-administered by the patient. In our experience, about 12% of cancer patients, who gave their informed consent to participate in the study, did not fill out the basal questionnaire (IGEO, 1999a). Patients who refuse to fill out the HRQL questionnaire cannot be considered randomly selected from the
181
182
10
2.
3.
4.
5.
The Uniscale Assessment of Quality of Life: Applications to Oncology
experimental group. In fact, this percentage increases with age and stage of disease, and is particularly high in patients with low performance status and a low educational level (IGEO, 1999a). Furthermore, in longitudinal studies, the percentage of patients who do not fill out the HRQL questionnaire greatly increases during the follow-up, often with disease progression. For instance, one study (Kemmler et al., 1999) indicated that, in the sixth evaluation, about 4 months after the beginning of the trial, the percentages of responses were 36% and 42% in the two arms of treatment involved, respectively; about 60% of patients were not evaluated. In this latter case, a selection bias can occur, being the presence of dropouts frequently correlated with disease progression. Items: Psychometric questionnaires are designed to evaluate both the levels of each domain and their modifications during the follow-up; therefore, the same questionnaire is used in repeated assessments. In this way it is implicitly assumed that the same domains and the same items continue to be the most important even when disease is in progression. For instance, in the early stage of disease, limitations in doing household jobs (an item which often explores the domain ‘‘social role’’) might be perceived as important, but probably the concerns are different when the disease becomes disseminated. The use of the same questionnaire during the follow-up seems to be in contradiction to Cella’s definition of quality of life (see Introduction section), and may lead to a lack of > responsiveness. Geographic differences: Domains may not be independent of the country where the evaluation is carried out. The ways of life (i.e., the patient’s family could play a different role in the management of the cancer patient) as well as the mean educational level may be different in various countries. Moreover, different social and National Health Systems (NHS) could cause the patient to assign a different importance to the same domain (i.e., economic conditions may be more or less important depending on whether the cost of the disease is completely supported by the NHS or is totally or partially charged to the patient). In conclusion, in some countries domains different from those explored in a questionnaire produced and validated in another country may be the most important. Scoring: The score of each multi-item domain is obtained by calculating a simple mean. In this way, a weight system is implicitly assumed: all items are equally important. The choice of another weight system could highlight the different importance of the items in exploring the same domain. The score of each multi-item can be affected by the choice of the weights to assign to each item, but there is no agreement on how to obtain the best weight system. Setting: A patient’s answers could be affected by both the time and place in which the questionnaire is filled out. Perhaps a patient waiting to receive chemotherapy might give answers different from those that he/she would give after hearing reassuring news about his/her disease. Unfortunately, experimental studies on these aspects are lacking.
Furthermore, comparisons between scores of the same domain obtained using different HRQL questionnaires should be avoided. Even when the same patient fills out two questionnaires, correlation between the scores referring to the same domain is generally poor (Kemmler et al., 1999). Many of the above-mentioned problems affecting psychometric questionnaires may be overcome by using a simpler, single item overall measurement of HRQL such as the uniscale. Because the selection effects are at least in part due to the length and complexity of a multiitem questionnaire, perhaps a simpler tool might lead to greater patient compliance. Then, problems described in points (2–4) would automatically be overcome.
The Uniscale Assessment of Quality of Life: Applications to Oncology
10
The price to be paid consists in (1) more research on this old tool, and (2) the lack of analyticity, i.e., the impossibility to detect what the domains are that have induced a variation in a patient’s quality of life. In the next section, an Italian experience of the external validity of the uniscale is described, highlighting how easy it is for cancer patients to fill out the VAS.
3
The Uniscale Assessment of Quality of Life: An Italian Experience
Starting from the awareness that the available QOL instruments reflect more a medical perspective, rather than a patient’s perspective, the Italian Group for the Evaluation of Outcomes in oncology (IGEO) planned a study on the foundations of QOL from the patient’s point of view. In the first phase, Italian cancer patients were interviewed to give an empirical content to the concept of QOL as well as to define what domains and problems were involved in their idea of QOL. A large sample of cancer patients, equally distributed among different types and stages of disease, and geographic areas, was selected. They were asked to give articulated answers to four questions: 1. 2. 3. 4.
What does the term ‘‘quality of life’’ mean to you? In your opinion, what contributes to a good quality of life? What contributes to a poor or bad quality of life? Tell me what physical or psychological symptoms or problems influence your quality of life.
In addition, patients were asked to maintain a diary on ‘‘everything (positive and negative) changed in their quality of life by the illness or its treatment.’’ A study sample of 288 cancer patients with an equal distribution of the following characteristics was planned: 1. Place of residence (96 patients living in North, Center, and South, respectively) and for each place of residence, by 2. Primary cancer site (12 patients with cancer of the breast, lung, gastrointestinal, female genital organs, male genital organs, urinary system, head and neck tumor, and others, respectively), and for each cancer site, by 3. Stage of disease (six patients in follow-up with no evidence of disease or undergoing adjuvant chemotherapy, and six patients undergoing therapy for advanced disease or palliative care for each cancer site, respectively). The seven participating centers were asked to identify the patients with the requested characteristics. The patients’ answers were transcribed, and the transcript was submitted to a content analysis by a team including a psychologist, a nurse, and a physician (Costantini et al., 2000). Two hundred and forty eight patients were enrolled (86.1% of the 288 planned patients). Their characteristics are shown in > Table 10-1. Overall, a list of symptoms and 43 contents were identified, 19 problems, and 24 domains. In the second phase, an evaluation of the relevance of each domain and of the presence of each problem in a very large population of Italian cancer patients was carried out, so as
183
184
10
The Uniscale Assessment of Quality of Life: Applications to Oncology
. Table 10-1 Socio-economic and clinical characteristics of the 248 interviewed patients Characteristics
No. (%)
Geographic area North
76 (30.6)
Center
88 (35.5)
South
84 (33.9)
Age Mean sd*
53.2±14.7
Gender Male
140 (56.7)
Female
107 (43.3)
NR
1
Education (years) 5
79 (33.9)
6–8
75 (32.2)
9–13
62 (26.6)
>13
17 (7.3)
NR
15
Living status Alone 1 relative
17 (7.8) 55 (25.2) 146 (67.0)
NR
30
Cancer site Breast
36 (14.5)
Lung – pleura
34 (13.7)
Gastrointestinal
33 (13.3)
Female genital organs
25 (10.1)
Male genital organs
22 (8.9)
Urinary system
36 (14.5)
Head and neck
35 (14.1)
Others
27 (10.9)
Stage of disease NED+ or adjuvant chemotherapy
120 (40.4)
Advanced or palliative care
128 (51.6)
*sd = standard deviation; NR = Not reported (excluded in calculating the percentages) + NED = No evidence of disease In planning the sample, patients were balanced with respect to both geographic area and cancer site so as to have a widespread overview of both domains and problems affecting a patient’s QOL
The Uniscale Assessment of Quality of Life: Applications to Oncology
10
to associate a frequency to the importance of each dimension and to the presence of each problem. From July 6 to 12 1996, over a period of 1 week, consecutive cancer patients attending 76 Italian medical oncology and radiotherapy centers (9 in tertiary care, 18 in teaching hospitals, and 49 in common hospitals) were asked to fill out a self-administered questionnaire concerning the relevance of the above-mentioned domains as well as the presence of the listed problems. Moreover, patients were asked to fill out three VASs: the first, regarding a uniscale assessment of their QOL, the second and the third their perception of the severity and curability of their own disease, respectively (IGEO, 1999a; IGEO, 1999b). From this database we extracted information related to the uniscale assessment of QOL, the presence/absence of 19 problems, and the patient’s and disease characteristics recorded in the CRF (Case Record Form) so as to detect the problems that are particularly important in defining the levels of QOL (Ballatori et al., 2007). More precisely, patients were asked to mark a point in a 100-mm long horizontal line in which the extremes were labeled with 0 (the worst QOL) and 100 (the best QOL). Answers were classified as ‘‘bad QOL’’ if a point in the 0–30 mm interval was chosen on the VAS or as ‘‘good QOL’’ if a point in the 70–100 mm interval was selected (> Figure 10-1). The two cutoff values, ‘‘30’’ and ‘‘70,’’ were chosen because they are specific approximations of the two tertiles of the interval 0–100. Therefore, patients who gave a score between ‘‘30’’ and ‘‘70’’ were not considered, allowing us to obtain two variables that could be separately analyzed. Multifactorial analyzes were performed using additive logistic models, in which dependent variables in turn considered each of the two responses, ‘‘bad QOL’’ or ‘‘good QOL,’’ assuming as explanatory variables the following factors: sex, age (5 years), stage of disease (no evidence of disease, localized disease, disseminated disease, at diagnosis). Moreover, among the explanatory variables, binary information derived from the presence/absence of the 19 above-mentioned problems was considered. Results of unifactorial analyzes were provided as adjusted odds ratios and their 95% confidence intervals. The G-test (which is equal minus two times the logarithm of maximum likelihood ratio and asymptotically distributed as a w2) was used to evaluate the overall significance of each factor in the model. The z-test (signed square root of Wald’s test that is asymptotically distributed as a w2 with 1 degree of freedom), obtained dividing the parameter estimation by the estimation of the correspondent standard error, was used to assess the significance of differences between levels of each factor, adjusting for the other factors in the model. Of 6,918 patients, 820 (11.9%) did not fill out the questionnaire, and, therefore, 6,098 patients were evaluated. Patients’ characteristics, and the percentages of patients perceiving as ‘‘good’’ or ‘‘bad’’ QOL with respect to them, are displayed in > Table 10-2. The high number of valid responses assures that the VAS is easy to fill out and generally accepted by Italian cancer patients. Patients with a ‘‘good QOL’’ are mainly those who are in better physical condition (see KPS), outpatients, and patients with no evidence of disease; a ‘‘bad QOL’’ is more common among patients with a low KPS score, inpatients, and patients with disseminated disease. Present problems, and the percentages of patients perceiving as ‘‘good’’ or ‘‘bad’’ their QOL with respect to them are shown in > Table 10-3. QOL was considered ‘‘good’’ by 2,099 (34.4%) as opposed to 962 (15.8%) patients who found it ‘‘bad.’’ These percentages were influenced by the presence of problems. In fact, the percentage of patients who felt ‘‘bad’’ is highest among patients who have difficulties in ‘‘washing and getting dressed,’’ in ‘‘daily life,’’ and in ‘‘physical activities.’’ Patients who have these same problems less frequently than the others perceived their QOL as ‘‘good.’’ The results of the multifactorial analyzes (see > Tables 10-4–10-6) are shown for ‘‘good QOL’’ (> Table 10-4) and ‘‘bad QOL’’ (> Table 10-5) separately, in terms of z-test and corresponding significance level. Moreover, they were shown in > Table 10-6 in terms of odds ratios adjusted for the other factors in the models, assuming equal to 1 the odds ratio for each reference category. Only significant factors were reported. Gender, education, Karnofsky performance status, and setting are important in explaining the variability of the perception of a ‘‘good’’ QOL. The probability of perceiving their own QOL as ‘‘good’’ is highest for males, for patients with the lowest levels of education, for outpatients and patients with a KPS 90. Finally, the presence of each of 12 of 19 listed problems (trouble in concentrating, living in a particularly stressful, or anxious period, anxiety about follow up results, body changes due to illness, lack of optimism, difficulties in physical, working, and sexual activity, economic troubles, unsatisfactory communication with doctors, lack of desire of social relationship, and change in working skill) less frequently led patients to perceive their QOL as ‘‘good.’’ Assuming ‘‘bad QOL’’ as a response variable, the pattern of significant explanatory variables is similar but inversely correlated. In conclusion, in this extensive evaluation, among the 19 problems referred by cancer patients in the first phase of the study 12 were found to have a significant impact on defining as ‘‘good’’ or ‘‘bad’’ the patients’ QOL.
10
The Uniscale Assessment of Quality of Life: Applications to Oncology
. Table 10-2 Percentages of patients perceiving as ‘‘Good’’ or ‘‘Bad’’ their QOL with respect to patient’s and disease characteristics No. Patient’s characteristics Total
6098
Good QOL
Bad QOL
%
(No.)
%
(No.)
34.4
(2099)
15.8
(962)
Gender Male
2295
37.2
(855)
15.4
(353)
Female
3803
32.7
(1244)
16.1
(609)
5 years
849
31.5
(132)
15.5
(132)
NR*
222
30.6
(41)
18.5
(41)
Time from diagnosis
Disease extension No evidence of disease
2580
41.1
(1061)
10.2
(263)
Localized disease
1197
33.2
(398)
17.6
(211)
Disseminated disease
2088
26.8
(560)
21.3
(445)
233
34.3
(80)
18.5
(43)
At the time of the diagnosis or NR* *NR: Not reported
. Table 10-3 Percentages of patients perceiving as ‘‘Good’’ or ‘‘Bad’’ their QOL with respect to the presence of problems in their present condition No. Patients’ conditions TOTAL
6098
Good QOL
Bad QOL
%
(No.)
%
(No.)
34.4
(2099)
15.8
(962)
Trouble in concentrating
2446
26.4
(645)
22.0
(538)
Memory difficulties
2547
28.7
(730)
19.3
(493)
Living in a particularly stressful or anxious period
4244
28.6
(1213)
19.5
(828)
Anxiety about follow-up results
4303
30.6
(1317)
18.6
(802)
Body changes due to the illness
3624
27.9
(1011)
20.3
(736)
Lack of optimism
(476)
1722
22.3
(383)
27.6
Lack of support from relatives
443
27.7
(123)
19.2
(85)
Difficulties in washing and in dressing
392
19.6
(77)
38.1
(149)
Difficulties in daily life
995
17.0
(169)
37.5
(373)
Difficulties in physical activity
1953
18.3
(358)
29.6
(578)
Difficulties in working activity
2520
21.2
(535)
27.5
(693)
Difficulties in performing sexual activity
2859
25.2
(722)
23.8
(680)
Economic troubles
1033
23.5
(243)
27.7
(286)
Inadequate support from the health service personnel
268
27.3
(73)
24.6
(66)
Inadequate care and services
260
23.5
(61)
27.7
(72)
Unsatisfactory communication with doctors
311
19.6
(61)
27.1
(84)
10
The Uniscale Assessment of Quality of Life: Applications to Oncology
. Table 10-3 (continued) No. Patients’ conditions
Good QOL
Bad QOL
%
(No.)
%
(No.)
Lack of desire of social relationship
2375
30.3
(720)
18.7
(444)
Change in working skill with respect to the past
3706
26.1
(965)
21.2
(786)
934
31.3
(292)
20.8
(194)
Change in one’s faith
Percentages of patients who have a ‘‘good’’ or a ‘‘bad’’ QOL could be read comparing them with those reported for all the patients (i.e., ‘‘total,’’ independent of the presence of the problems; see the first line). Thus, the impact of each problem on patient’s QOL becomes evident. For instance, 15.8% of all patients feel they have a ‘‘bad QOL’’; the same proportion among those who have ‘‘difficulties in washing and in dressing’’ grows to 38.1%
4
Conclusions
In recent years QOL has been studied following a multidimensional approach. Many studies have demonstrated that several domains are concurrent in defining the concept of QOL. Each dimension must be explored using one or, more frequently, more items. Therefore, the QOL questionnaires currently in use are comprehensive of a number of items often varying from 30 to 60. Worthy of note is that this high number of items appears extraordinarily small when one thinks that each domain should be explored in three different dimensions: intensity, frequency, relevance (i.e., the impact on patient’s daily life). The most positively important aspect of the multidimensional evaluation consists in the possibility of separately evaluating each domain, so as to detect which of them are most affected by the growing severity of the disease or by the side effects of the therapies. However, this approach suffers from some shortcomings; in particular, a nonnegligible percentage of patients that refuse to fill out the first questionnaire in spite of having given their informed consent to participate in the study (about 12% in Italy (IGEO, 1999a)). Moreover, in longitudinal studies an increasing percentage of patients refuse to complete the questionnaire in subsequent sessions, mainly because of disease progression. Thus, a severe selection bias often affects the evaluation of QOL when, instead, it is most important that it be done correctly. In addition, for reasons of comparability, the same items are used in subsequent evaluations; in this way it is implicitly assumed that the same domains/items continue to be the most important, even with the worsening of the disease. For instance, at early stage disease, limitation in doing household jobs could be perceived as important by the patient, but probably the concerns are different when the patient is bedridden due to disseminated disease. Therefore, a psychometric questionnaire should be used only in parallel randomized clinical trials, where the obtained results can be regarded as reliable only if the number of dropouts is not large and the two arms are well balanced. At the beginning of the 1980s, a unidimensional approach was suggested, using a VAS. However, this procedure was soon abandoned because of much criticism, such as its low reliability, the dependence of the answer on the patient’s contingent conditions, and the impossibility to detect the domains that are most affected by changes in the patient. Some of these criticisms were founded on empirical research; others, instead, were based only on a theoretical point of view. Today, because of the serious difficulties arising from the use of questionnaires containing too many items, the old uniscale might be reevaluated as a basic tool in measuring quality of life. In fact, a VAS is very easy to fill out, and its reliability can be substantially improved in
189
190
10
The Uniscale Assessment of Quality of Life: Applications to Oncology
. Table 10-4 Multifactorial analyzes of the ‘‘good QOL’’ perception performed using logistic additive models: parameter estimations, z-test, and significance levels Factors
z-test
P
Table 10-5. LP = –4.159 – 0.238 (outpatient) + 0.347 (disseminated disease) + 0.545 (particularly stressful period) + 0.487 (difficulties in daily life) + 0.734 (in working activity) + 0.463 (economic troubles) = –1.823. Therefore, the probability to have a ‘‘bad QOL’’ (PY/.) is equal to P(Y/.) = e-1.821/(1+e-1.821) = 13.9%
different ways (i.e., by analyzing the responses, not as numerical values, but grouping them in classes). Unfortunately, there is no clear evidence that the use of a VAS allows us to obtain better patient compliance than the use of a psychometric questionnaire, but this should be the main aim of an ad hoc planned study. The results of our study (Ballatori et al., 2007), performed on a very large population of Italian cancer patients, can be considered as further proof of the uniscale external validity, showing that it is strongly influenced by unfavorable patient characteristics and by the problems related to their condition. These results are in accord with those reported in a recent paper (de Boer et al., 2004) where it was shown that, in patients affected by esophageal adenocarcinoma, the VAS used as the uniscale assessment of the QOL is an instrument with good convergent and discriminant validity and good responsiveness. Moreover, as shown in our paper, a VAS is able to summarize several adverse conditions, and their importance can be evaluated only by the patient: this overcomes the difficulties
191
192
10
The Uniscale Assessment of Quality of Life: Applications to Oncology
. Table 10-6 Multifactorial analyzes of the QOL perception performed using logistic additive models: odds ratios} (95% confidence interval) for each significant factor adjusted for the other factors in the models Factors
Good QOL
n.s.+
Gender Male* Female
1 0.80 (0.71–0.91) n.s.+
Education Primary school*
1
Middle school
0.80 (0.69–0.92)
High school/Degree
0.62 (0.54–0.72)
NR
Bad QOL
n.s. n.s.+
Karnofsky PS 80*
1
90
1.21 (1.05–1.40)
Treatment setting Inpatient*
1
Outpatient
1.27 (1.04–1.55)
1 0.79 (0.64–0.97)
Disease extension No evidence of disease* Localized disease Disseminated disease At the time of the diagnosis or NR Present condition
1 n.s.
1 +
1.35 (1.09–1.68)
0.76 (0.65–0.88)
1.41 (1.17–1.70)
n.s.+
1.67 (1.13–2.48)
1
1
No* Yes Trouble in concentrating
0.78 (0.69–0.88)
1.38 (1.18–1.61)
Living in a particularly stressful or anxious period
0.66 (0.58–0.75)
1.72 (1.39–2.13)
Anxiety about follow-up results
0.81 (0.71–0.92)
1.53 (1.25–1.87)
Body changes due to the illness
0.76 (0.68–0.86)
1.44 (1.21–1.72)
Lack of optimism
0.60 (0.52–0.69)
2.13 (1.82–2.49)
Difficulties in daily life
n.s.
+
1.63 (1.34–1.97)
Difficulties in physical activity
0.69 (0.58–0.81)
1.31 (1.08–1.60)
Difficulties in working activity
0.70 (0.60–0.82)
2.08 (1.71–2.54)
Difficulties in performing sexual activity
0.83 (0.73–0.93)
1.43 (1.21–1.70)
Economic troubles
0.70 (0.59–0.83)
1.60 (1.33–1.90)
Unsatisfactory communication with doctors
0.70 (0.52–0.95)
n.s.+
10
The Uniscale Assessment of Quality of Life: Applications to Oncology
. Table 10-6 (continued) Factors
Good QOL
Bad QOL
Lack of desire of social relationship
0.76 (0.67–0.86)
1.33 (1.14–1.55)
Change in working skill with respect to the past
0.70 (0.61–0.80)
n.s.+
*Reference category: For each reference category odds ratio is fixed equal to 1; +n.s. = Not significant; NR: Not reported} Odds ratio (OR) is a nonnegative measure of association. OR = 1 indicates the independence (i.e., when OR = 1, the probability of a ‘‘good QOL’’ is not affected by the variation of the other characteristic). When OR > 1, there is a positive association, with the highest values indicating the strongest positive association (i.e., when considering ‘‘bad QOL’’ and ‘‘Trouble in concentration,’’ OR = 1.38 means that the probability of having a ‘‘bad QOL’’ is superior in patients who have ‘‘Trouble in concentrating’’ than the others. OR < 1 indicates a negative association with the smallest values indicating the strongest negative association (i.e., when considering ‘‘Trouble in concentration’’ and ‘‘good QOL,’’ OR = 0.78 means that the probability of a ‘‘good QOL’’ is inferior in patients having ‘‘Trouble in concentration’’ than in those who do not have it
of requiring changes in both the domains and items when the disease becomes more severe. In fact, remembering Cella’s definition of QOL (‘‘Quality of life refers to patients’ appraisal of and satisfaction with their current level of functioning as compared to what they perceive to be possible or ideal’’), the uniscale evaluation of QOL is able to summarize the patient’s comparison between his/her actual condition with those he/she perceives ‘‘to be possible or ideal,’’ where the perception of ‘‘possible or ideal’’ conditions varies with disease progression. When there is the necessity to evaluate the changes in several QOL domains, a VAS could be used to explore each of them, so as to reduce the many items of the psychometric questionnaire (i.e., from 30 to 6). In this case, it would be sufficient to find a way to make the content of each domain easily understandable to the patient. In conclusion, a VAS can be considered as a reference point in multidimensional QOL scales and should still be regarded as a useful and synthetic tool to investigate phenomena related to the patient’s QOL. In this perspective, more research on the psychometric properties of this instrument is still needed.
Summary Points Measuring QOL is extremely important in clinical research not only because QOL is one of
the only two endpoints of effectiveness of treatments, but also for making clinical decisions shared with the patient. Today, assessment of QOL is made using psychometric questionnaires; nevertheless, these latter have several shortcomings that often lead to unreliable results. About 14 years ago, the Italian Group for the Evaluation of Outcomes in oncology (IGEO) planned a research program articulated in two phases. In the first phase, domains of QOL and problems connected with it were defined performing a content analysis of the interviews of 248 Italian cancer patients. The domains/problems reported by the patients were submitted to a large population of more than 6,000 Italian cancer patients so as to assign a frequency to the relevance of each domain and to the presence/absence of each problem. In this study, a uniscale evaluation of QOL was also obtained using a Visual Analogue Scale (VAS).
193
194
10
The Uniscale Assessment of Quality of Life: Applications to Oncology
From this databank were extracted the uniscale evaluation, information about the presence of the problems, and patient’s and disease characteristics recorded in the CRF.
Each answer to the VAS was classified as ‘‘bad QOL’’/‘‘good QOL’’ if a point belonging to 0–30/70–100 interval was marked. Answers in the 30–70 interval were not considered so as to assure independence to the chosen response variables. The relationship between each of these responses (dependent variable) and the presence of the 19 problems, and patient’s and disease characteristics (explanatory variables) were analyzed using multifactorial logistic models. 12/19 problems were significant in explaining the variability of the two responses. This confirms the external validity of the uniscale assessment of QOL. A VAS aimed at measuring QOL should still be regarded as a useful and synthetic tool to investigate phenomena related to a patient’s QOL. In the perspective of its further development, more research on the psychometric properties of this instrument is still needed.
Acknowledgments > Tables
10-2, 10-3, 10-6 were reproduced with kind permission of the Editor of Tumori. We thank Mrs Katherine Brandt for her helpful assistance in reviewing the text.
References Aaronson NK. (1990). Oncology. 4(5): 59–66. Ballatori E. (2001). Ann Oncol. 12(3): S11–S13. Ballatori E, Porzio G, Roila F, Ruggeri B, Mattei A, Cortesi E. (2007). Tumori. 93: 78–81. Cella DF, Tulsky DS. (1990). Oncology. 4(5): 29–38. Coates A, Glasziou P, Mc Neil D. (1990). Ann Oncol. 1(3): 213–217. Costantini M, Mencaglia E, Di Giulio P, Cortesi E, Roila F, Ballatori E, Tamburini M, Casali P, Licitra L, Candiis DD, Massidda B, Luzzani M, Campora E, De Placido S, Palmeri S, Angela PM, Baracco G, Gareri R, Martignetti A, Ragosa R, Zoda L, Ionta MT, Bulletti S, Pastore L. (2000). Qual Life Res. 9: 151–159.
de Boer AG, van Lanschot JJ, Stalmier PF, van Sandick JW, Hulscher JB, de Haes JC, Sprangers MA. (2004). Qual Life Res. 13(2): 311–320. Hiratsuka T, Kida D. (1993). Internal Med. 32: 832–836. IGEO (The Italian Group for the Evaluation of Outcomes in Oncology). (1999a). Tumori. 85: 92–99. IGEO (The Italian Group for the Evaluation of Outcomes in Oncology). (1999b). Ann Oncol. 10: 1095–1100. Kemmler G, Holzner B, Kopp M, Dunser M, Margreiter R, Greil R, Sperner-Unterweger B. (1999). Jco. 17(9): 2932–2940.
11 The Bone Metastases Quality of Life Questionnaire X. Badia . A. Vieta . M. Gilabert 1 1.1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196 Why is it Important to Assess Quality of Life in Patients with Bone Metastases? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
2 2.1 2.2
The BOMET-QOL Questionnaire . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198 What is the BOMET-QOL? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198 Criteria Applied in Developing the Questionnaire . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198
3 3.1 3.2 3.2.1 3.3 3.3.1 3.3.2
Development of the BOMET-QOL-10 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200 Phase 1. Item Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200 Phase 2. Item Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200 Identification of the Initial Group of Items . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200 Phase 3. Item Reduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201 First Reduction: Reduction of the BOMET-QOL from 35 to 25 Items . . . . . . . . . . . 201 Second Reduction: Reduction of the BOMET-QOL from 25 to 10 Items . . . . . . . . 202
4
Validation of the BOMET-QOL-10 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202
5
Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
#
Springer Science+Business Media LLC 2010 (USA)
196
11
The Bone Metastases Quality of Life Questionnaire
Abstract: The > Bone Metastases Quality of Life Questionnaire (BOMET-QOL) is the first > bone metastases-specific HRQOL measure. It is a simple, self-administered questionnaire intended for use in clinical research and in the routine monitoring of patients with malignant bone disease due to neoplasia. The BOMET-QOL questionnaire is unidimensional and consists of 10 items that can be answered according to a Likert scale with five categories scoring from 0 to 4. The timeframe refers to ‘‘last week.’’ The items refer to HRQOL topics such as pain, mobility, vitality, sex life, worries about the future, perceptions and daily activities. The global score is calculated by adding up the answers obtained in each item, and may range from 0 (worst HRQOL) to 40 (best HRQOL). These scores are standardized to make it easier to interpret the scoring, with the final scores ranging from 0 (worst HRQOL) to 100 (best HRQOL). The BOMET-QOL was developed in three phases. The item generation phase was performed by means of a literature search, a panel of experts and 15 semi-structured interviews with patients. An initial set of 179 expressions was identified. The item selection phase consisted of the identification of the initial group of items. The 15 experts carried out a qualitative and quantitative reduction of the 179 expressions according to their clarity, frequency and importance. This phase resulted in the 35-item version of the BOMET-QOL. The item reduction phase was carried out in two steps. The initial reduction yielded a 25-item questionnaire. This was administrated to a non-randomized sample of 92 patients with malignant bone disease due to neoplasia (MBDN) and the reduction was carried out via factorial analysis. Similarly, the BOMET-QOL-25 was reduced to an integrated version of 10 items by means of a sample of 263 oncology patients. It was then validated, showing high homogeneity, good reproducibility and significant correlations with the ECOG and the EORTC-QLQ-C30 questionnaire. The BOMET-QOL questionnaire is a feasible (easy and user-friendly), reliable and specific 10-item instrument for assessing HRQOL in patients with MBDN. List of Abbreviations: BM, > bone metastases; BOMET-QOL, bone metastases quality of life questionnaire; CARES, cancer rehabilitation evaluation system; FACT-G, functional assessment of cancer therapy; HRQOL, health-related quality of life; MBDN, > malignant bone disease due to neoplasia; PMI, > pain management index; QLQ- 30, quality of life questionnaire; RSCL, Rotterdam symptom check list
1
Introduction
1.1
Why is it Important to Assess Quality of Life in Patients with Bone Metastases?
Several developments in cancer treatment have promoted interest in health-related aspects of quality of life. Firstly, while medical interventions have historically been evaluated in terms of their effectiveness in reducing tumor progression and mortality, the major therapeutic goal for many patients is to reduce the symptoms of the illness. Secondly, concerns about potential treatment safety and tolerability make quality of life issues important components of patient decision-making when choosing between alternative courses of treatment. Thirdly, the ability of sophisticated medical technologies to extend lifespan raises significant questions about the quality of the life that is being prolonged. Finally, with an increased emphasis on humanizing healthcare, improving quality of life is considered an important outcome even if life is not prolonged (Avis et al., 1996).
The Bone Metastases Quality of Life Questionnaire
11
For these reasons, assessment of Health-Related Quality of Life (HRQOL) in oncology has increased in the last few years. HRQOL has become an important outcome measure of new therapies in clinical trials (Cella et al., 2002). The importance of assessing HRQOL in clinical practice is also widely acknowledged, since cancer treatment, in advanced cancer patients, is aimed mainly at relieving symptoms, especially pain. Means of standardizing assessment of the disease’s impact and treatments on HRQOL from the patient’s point of view are therefore needed (Badia et al., 2003). The skeleton, particularly the axial skeleton, is a frequent location for metastases, especially in patients with breast, prostate, thyroid, kidney or lung cancer or multiple myeloma. Specifically, bone metastases (BM) are the third most frequent metastatic site for primary tumors and the first in terms of morbidity and impact on HRQOL. The incidence of bone metastases in patients with advanced metastatic disease is approximately between 65 and 75% in breast and prostate cancer, 30–40% in lung cancer, 14–45% in melanoma, 20–25% in renal cell carcinoma, and 60% in thyroid and 40% in bladder neoplasia (Coleman, 2004). Bone metastases are a major cause of morbidity and significantly reduce a patient’s quality of life. Pain is the most frequent clinical sign and bone metastasis is likely to be the main cause of pain in oncology patients (Lipton, 2003; Martinez et al., 2003; Reddi et al., 2003). Aside from pain, BM frequently produces the following complications: pathological fractures, spinal cord compression, tumor-induced hypercalcemia and bone marrow infiltration. It is well established that these complications significantly reduce patients’ HRQOL, although this has not been properly studied (Bunting and Shea, 2001). Although patients can survive for long periods, treatment of bone metastasis tends to be only of a palliative nature. External radiotherapy and palliative chemotherapy are two of the most frequently used strategies to treat bone pain, along with the analgesic use and nerve blocking of regional anesthesia as palliative therapeutic options. In the last few years, therapy for malignant bone disease due to neoplasia (MBDN) has improved significantly due to its multidisciplinary approach and to new therapeutic treatments, especially to the second and third generation bisphosphonates (Djulbegovic et al., 2002). Initially used to treat tumorinduced hypercalcemia, these were later associated with hormone therapy and chemotherapy, showing benefits in the control of symptoms, reducing morbidity and bone progression. The aims of the adjunct treatment are to alleviate pain and stabilize the bone, as well as to ensure the patient’s well-being and independence (> Table 11-1). The assessment of quality of life in oncology patients requires the use of standardized instruments (usually self-administered questionnaires) specifically designed for this purpose. It is essential that the instruments used to measure quality of life meet the basic characteristics (> validity, > reliability and sensitivity to change) in order to guarantee the quality of the outcomes deriving from their application. Using these questionnaires means that patients’ quality of life can be quantified, using this as an outcome measure for treatments and medical interventions both in research and in clinical practice. Nowadays, there are various instruments to assess HRQOL in oncology, such as the Cancer Rehabilitation Evaluation System (CARES) (Ganz et al., 1992), > EORTC QLQ-30 Quality of Life Questionnaire (Aaronson et al., 1993), Functional Assessment of Cancer Therapy (FACT-G) (Cella et al., 1993), and the Rotterdam Symptom Checklist (RSCL) (Watson et al., 1992). Taking into account the general nature of these questionnaires, which cover common issues concerning impact on the quality of life in all types of neoplasia, they might be insufficiently sensitive to some changes exclusively associated with MBDN. Consequently, the 10-item BOMET-QOL questionnaire (BOMET-10) was designed in order to assess quality of life in patients with MBDN.
197
198
11
The Bone Metastases Quality of Life Questionnaire
. Table 11-1 Key features of bone metastases 1. Metastatic bone disease is a cancer starting in another location and spreading to the bones 2. The metastasis of tumor cells to bone requires a series of sequential steps, involving detachment from the primary tumor site, invasion of the vasculature, migration and adherence to distant capillaries of the bone, extravasation, and proliferation. Once tumor cells have invaded the bone matrix, they produce growth factors that can directly or indirectly stimulate osteoclasts that break down the bone 3. Bone metastases are generally classified as osteolytic, characterized by the destruction of normal bone, or osteoblastic, with the deposition of new bone, based upon the predominant radiological appearance 4. Bone is the third most common site of metastatic disease, most likely because of the favorable microenvironment of the bone matrix and its ample blood supply 5. The incidence of bone metastases in patients with advanced metastatic disease is approximately between 60 and 65% in breast and prostate cancer, 30–40% in lung cancer, 14–45% in melanoma, 20–25% in renal cell carcinoma, and 60% in thyroid and 40% in bladder neoplasia 6. Bone metastases are a major cause of morbidity and significantly reduce the patient’s quality of life, often accompanied by skeletal-related events including pathological fracture, spinal cord compression, hypercalcemia or pain 7. Treatment of bone metastasis consists of active chemotherapy, hormonal therapy treatment of the original cancer and the use of bisphosphonates This table lists the key facts in bone metastases including physiology, epidemiology, medical impact and clinical management
2
The BOMET-QOL Questionnaire
2.1
What is the BOMET-QOL?
The Bone Metastases Quality of Life Questionnaire (BOMET-QOL) is the first to measure bone metastases-specific HRQOL. It is a simple, self-administered questionnaire, intended for use in clinical research (trials) and in the routine monitoring of patients with bone metastases or myeloma. The items refer to HRQOL topics such as pain, mobility, vitality, sex life, worries about the future, perceptions and daily activities. The BOMET-QOL questionnaire is unidimensional and consists of 10 items that can be answered according to a Likert scale with five categories (always, almost always, sometimes, rarely and never), scoring from 0 to 4, and the timeframe refers to ‘‘last week’’ (> Figure 11-1). The global score is calculated by adding up the answers obtained in each item, and may range from 0 (worst HRQOL) to 40 (best HRQOL). These scores are standardized to make it easier to interpret the scoring, with the final scores ranging from 0 (worst HRQOL) to 100 (best HRQOL) (> Figure 11-1).
2.2
Criteria Applied in Developing the Questionnaire
Any instrument for assessing outcome research should be patient-oriented (describe the functional state, including daily activities), useful and simple (ideally should contain less than 10 items, be acceptable for daily clinical practice and easy to interpret). Also, certain basic
The Bone Metastases Quality of Life Questionnaire
11
. Figure 11-1 Scoring system and items on the BOMET-QOL-10 questionnaire. This figure presents the 10 items of the BOMET-QOL questionnaire, answered according to a Likert scale of five categories
measuring standards (validity and reliability) should be carried out on the instrument in order to ensure the quality of the results deriving from the questionnaire. The validity of a questionnaire is defined as its ability to measure and describe what it is supposed to measure and describe. Validity is evaluated in terms of cross-sectional (construct validity) and > longitudinal validity. Cross-sectional or construct validity addresses the issue of whether or not the instrument assesses the construct as defined in the measurement model. Longitudinal construct validity or sensitivity to change implies that even small changes in the order of appearance of questions, or question wording, may cause changes in the results. The validity of a questionnaire relies first on reliability. Reliability means the statistical reproducibility of measurements. This may be test-retest reliability, when repeating the questionnaire under the same conditions produces the same results, or reliability within a scale, when all the items designed to measure a particular attribute are definitely measuring the same attribute. In view of the above, the development of the BOMET-QOL to assess the quality of life in patients with MBDN focused on the following criteria: 1. The content of the questionnaire should be suitable and specific for patients with MBDN. The scale should be suitable for all illness severity levels, no matter what the causes are. 2. The questionnaire should be based on assessing the perceived problems (subjective) so they can be reported by the patient. The questionnaire should generate answers concerning problems that may have been important or activities that may have been stopped because of the illness.
199
200
11
The Bone Metastases Quality of Life Questionnaire
3. The questionnaire should have a high applicability. Age, gender or any other sociodemographic issue, such as the patient’s level of education, should not be a conditioning factor in its administration or a determining factor in the outcomes obtained. For these reasons, the questionnaire should be easy to understand and fast to complete. It should also be easy to administrate by the health carers and easy to process, analyze, and interpret. 4. The measurement properties of the questionnaire should be validated so that it can be applied rigorously both in a research context and in common clinical practice.
3
Development of the BOMET-QOL-10
The BOMET-QOL-10 questionnaire was developed in three phases:
3.1
Phase 1. Item Generation
This phase identified the areas of content or dimensions of the questionnaire. The bibliography was reviewed in order to identify the clinical manifestations and clinical characteristics of patients with bone metastases. A panel of experts (10 oncologists, 1 hematologist and 4 urologists) then identified, in an interview, the most relevant signs and symptoms for these patients. Lastly, exploratory interviews were carried out with 15 patients with breast, lung or prostate cancer affected by bone metastases (> Table 11-2). The aim of this phase was to analyze the impact of symptoms and their treatment on quality of life: wellness, satisfaction with the state of health, and functional level according to the patients’ age and preferences. Special interest was placed on how the illness might impact areas such as personal, social and cognitive functioning or rest. The interviews were semistructured and were carried out by two monitors trained by researchers with expertise in this area. The patients interviewed had different clinical and socio-demographic characteristics and met all the criteria defined by the panel of experts. The interviews were recorded and transcribed for subsequent qualitative and quantitative analysis. This phase yielded the clinical criteria required to identify the patients to test the questionnaire and its dimensions. nine domains were identified from this phase (physical, psychological, social, daily activities, symptoms, cognitive dimension, health perception, treatment perception, energy/vitality and pain) as well as an initial set of 179 expressions.
3.2
Phase 2. Item Selection
3.2.1
Identification of the Initial Group of Items
The 15 experts who were involved in phase 1 identified the initial group of items by means of a qualitative and quantitative analysis. For the qualitative analysis, the experts gathered the expressions obtained for the different dimensions from phase 1 and identified the ones that were too similar, ambiguous or inadequate, as these expressions could be converted into simple expressions suitable for a
The Bone Metastases Quality of Life Questionnaire
11
. Table 11-2 Characteristics of MBDN patients interviewed for the item generation phase Patient
Gender
Age
Education level
Primary tumor
Pain score (0–10)
1
F
36
Secondary level
Breast
6
2
F
53
Primary level
Breast
5
3
F
43
University level
Breast
7
4
F
62
Primary level
Breast
6
5
F
58
Secondary level
Breast
6
6
M
78
No education
Prostate
7
7
M
69
Primary level
Prostate
5
8
M
58
University level
Prostate
4
9
M
65
Primary level
Lung
3
10
F
72
No education
Myeloma
5
11
F
67
No education
Myeloma
6
12
M
56
Primary level
Myeloma
3
13
M
61
No education
Myeloma
7
14
M
49
University level
Myeloma
8
15
M
75
No education
Myeloma
2
This table summarizes the main patient characteristics, such as gender (F: female, M: male), level of education (no education, primary, secondary and university level), primary tumor site (breast, prostate, lung or myeloma) and the pain score from 1 to 10
self-administered questionnaire. The quantitative analysis consisted of recounting the frequency of the terms used. The experts scored the initial group of items according to the clarity of wording, frequency of occurrence, and importance using a 5-point Likert scale. As a result of this analysis, a sample of items or representative questions was identified that represent the perceptions of HRQOL in patients with bone metastases. This qualitative and quantitative analysis carried out by the experts resulted in the 35-item version of the BOMET-QOL.
3.3
Phase 3. Item Reduction
3.3.1
First Reduction: Reduction of the BOMET-QOL from 35 to 25 Items
This preliminary questionnaire (BOMET-QOL-35) was then administrated to a nonrandomized sample of 92 patients with MBDN from 21 centers throughout Spain. The baseline neoplasia was: breast (37%), prostate (23%), myeloma (15%), lung (21%) and other types (4%). Fifty percent of the patients were male. There were patients with different levels of education, aged from 30 to 80 years, diagnosed with MBDN for up to a maximum of 9 years, who had or had not undergone chemotherapy. Nineteen percent had concomitant osteoarticular disease and ninety percent a non-osteoarticular chronic disease. From the answers of the patients in this initial administration, the items were analyzed (> Rasch analysis) evaluating each one in terms of its discriminant validity, which is the
201
202
11
The Bone Metastases Quality of Life Questionnaire
questionnaire’s capacity to distinguish between patient subgroups. The > Rash analysis constructs a line of measurement with the items placed hierarchically and provides fit statistics to indicate just how well different items describe the group of subjects responding to a questionnaire (Rasch, 1993). Therefore, the items that showed the worst fit were eliminated from the questionnaire. Using the Factorial analysis from the Rasch analysis, the most adequate item grouping was established, as presented in the 25-item BOMET-QOL questionnaire. Criteria such as the internal consistency of the items (using item-total correlation) and reliability of the results were also taken into consideration. The questionnaire showed good internal consistency with a Cronbach’s Alpha coefficient of 0.94. This analysis resulted in the 25-item BOMET-QOL questionnaire (Adrover et al., 2005).
3.3.2
Second Reduction: Reduction of the BOMET-QOL from 25 to 10 Items
The BOMET-QOL-25 was reduced to an integrated version of 10 items (Sureda et al., 2007). The BOMET-QOL-25 was administered to 263 oncology patients from the oncology, urology and hematology units of 46 hospital centers in Spain. The mean age (SD) of the patients was 62.20 (12) years old and 42.9% were males. Patients diagnosed as having breast (38.8%), prostate (18.2%) or non-microcytic lung (16.7%) cancer, with MBDN or myelomas (26.3%), were included. The final reduction of the items in the BOMET-QOL questionnaire was carried out in two phases. In the first phase, a factor analysis was carried out with varimax rotation of the primary BOMET-QOL questionnaire’s items. The factor analysis showed the unidimensional nature of the questionnaire, with only one factor explaining 61.2% of the variance. In the second phase, each of the resulting factors was calculated using the Rasch Rating Scale model. This analysis reduced the questionnaire to 10 items. As a result of the reduction process, an easy and user-friendly questionnaire was obtained with only 10 items and one dimension (> Figure 11-2).
4
Validation of the BOMET-QOL-10
The psychometric properties of the BOMET-QOL-10 were assessed under common clinical practice conditions in an observational, prospective and multi-center study designed and carried out with the above sample of 263 patients. Patients were assigned to two groups: group A included stable patients with no expected changes in disease control over a 15-day period; and group B consisted of patients with an expected change in their health owing to their undergoing treatment of effective surgery. The characteristics of each group in terms of age, gender, level of education, type of baseline neoplasia, concomitant osteoarticular disease and non-osteoarticular chronic diseases were similar. Patients included in group A made two visits (baseline visit and 15 days after baseline visit) and patients included in group B made three visits (baseline visit, 3 months after and 6 months after baseline visit) (> Figure 11-3). The first visit gathered socio-demographic variables and clinical variables such as site and date of diagnosis for the primary tumor, MBDN sites, presence, number and duration of irruptive pain crises, concomitant chronic and
The Bone Metastases Quality of Life Questionnaire
11
. Figure 11-2 Diagram of the development of the BOMET-QOL 10 questionnaire. This figure presents the development of the BOMET-QOL-10 questionnaire. The BOMET-QOL was developed in three phases: item generation (179 expressions), item selection (BOMET-QOL 35) and item reduction (BOMET-QOL 10)
osteoarticular diseases and the treatment for the primary tumor and for the MBDN. The investigator also included the ECOG Performance Scale (Oken et al., 1982) and the Pain Management Index (PMI) (Cleeland, 1991), which relates the type of analgesic treatment received to the pain level declared by the patient. After completing the first visit, the patient fills out the Spanish version of the EORTC-QLQ-C30 questionnaire, the BOMET-QOL-25 questionnaire and the perception of general health status. Any changes in treatment, ECOG and PMI were recorded in the second and third visits. Likewise, changes in health status perceived by the patients were computed and the EORTC-QLQ-C30 and BOMET-QOL questionnaires (BOMET-QOL-25 and BOMET-QOL-10) were administered again (> Figure 11-4). The impact on HRQOL assessed by the BOMET-QOL-10 questionnaire and variables of irruptive pain, patients’ self-reported general health status, PMI and ECOG index (> ECOG
203
204
11
The Bone Metastases Quality of Life Questionnaire
. Figure 11-3 Graph of the observational study design. This figure presents the observational study design used in reducing the BOMET-QOL from 25 to 10 items, and the validation of the psychometric properties of the BOMET-QOL-10
. Figure 11-4 Socio-demographic, clinical and humanistic variables collected via the observational study. This figure presents all the variables collected in the observational study at the baseline visit and/or follow-up visits
index for performance status) were shown to be correlated. All the dimensions of the BOMETQOL-10 had a statistically significant correlation with the dimensions of the EORTC-QLQ-30 score (p < 0.01). No statistically significant differences were found regarding the HRQOL in terms of the time of neoplasia evolution (p = 0.12).
The Bone Metastases Quality of Life Questionnaire
11
The measurement properties analyzed of the BOMET-QOL-10 questionnaire were validity, sensitivity to change and reliability. Cross-sectional validity was assessed evaluating the correlation number and duration of the irruptive pain crisis. A correlation was found between the impact on HRQOL evaluated with the BOMET-QOL questionnaire and the irruptive pain variable. A lower BOMET-QOL score was observed when there were fewer irruptive pain episodes, when PIM = 0 (good pain control) or when the ECOG score was between 0 and 1, and Group B patients scored lower in the BOMET-QOL questionnaire (p < 0.001). Longitudinal validity (sensitivity to change) was determined by the changes in the question ‘‘general health status’’ between the baseline visit and the visit 3 months later. Changes in the ECOG index were correlated with the scores observed in the BOMET-QOL10 questionnaire (p < 0.01). The BOMET-QOL-10 questionnaire detected changes better than the ECOG index. Also the size of the effect was measured for patients in Group B at 3 and 6 months ( 0.84). This increased its score (better HRQOL) throughout the study (p < 0.001). In terms of reliability, good internal consistency was achieved (Cronbach’s a = 0. 92). No significant differences in the questionnaire scores were observed at the two moments of the study (baseline and 15 days visits). The Intraclass Coefficient Correlation (ICC) was 0.93. To conclude, the BOMET-QOL questionnaire of 10 items has shown good measurement properties for changes (> Table 11-3), being of great use both in clinical research and common clinical practice. > cross-sectional
. Table 11-3 Psychometric properties of the BOMET-QOL-10 questionnaire Psychometric properties Cross-sectional validity
Measurement
Relationship between BOMET-QOL-10 and the number of r = 0.293 irruptive pain crises (p < 0.01) Relationship between BOMET-QOL-10 and the mean duration of irruptive pain crises
Longitudinal validity/ sensitivity to change
Statistics
r = 0.226 (p < 0.01)
Relationship between a change in the ‘‘general health p < 0.01 status’’ question and a change in the scores between the baseline visit and the visit 3 months later Relationship between a change in the ECOG and a p < 0.01 change in scores between the baseline visit and the visit 3 months later Size of the effect between the baseline visit and the visit 6 months later
Reliability
0.84
Cronbach’s Alpha
0.93
Interclass correlation coefficient (ICC)
0.94
This table summarizes the measurement properties of the questionnaire: cross-sectional validity, sensitivity to change and reliability. ‘‘r’’ corresponds to Pearson’s correlation coefficient
205
206
11 5
The Bone Metastases Quality of Life Questionnaire
Conclusions
There are many reviews describing the importance of symptomatology in patients with MBDN, either in general or in relation to the primary tumor. Nonetheless, these are mostly descriptive and do not posses the proper methodology to evaluate HRQOL. Moreover, they do not deal with MBDN as a unique and differential syndrome (irrespective of the primary tumor), looking at the impact on HRQOL. This article has presented the development and validation of a specific questionnaire to measure HRQOL in patients with MBDN. The content of specific questionnaires is more adequate for patients’ problems than generic questionnaires and is more sensitive to significant clinical changes. HRQOL questionnaires enable standardized information to be obtained about the impact of the disease or treatment on the patient’s HRQOL. However, at present these questionnaires are not used in current clinical practice as many clinicians have difficulty in interpreting the results of the multidimensional information outcomes, or with the extension provided by the HRQOL questionnaires (Bezjak et al., 1998). For these reasons, emphasis has been placed on developing and validating this 10-item questionnaire. The BOMET-QOL-10 is user-friendly, easy to complete by the patient and to assess by the professional and has good measurement properties. The first step toward developing an HRQOL tool in patients with MBDN was the identification by experts of the domains directly influenced by MBDN with an impact on HRQOL. Afterwards, fifteen patients were asked to explain how the pathology influenced their HRQOL and answers from them matched the domains the experts had previously selected. It was obvious that MBDN has a well-known effect on HRQOL and patients welcomed the opportunity to verbalize their health-related limitations. Patients place a high value on their HRQOL and this fact is extensively reported in the literature. After identifying an initial group of items, the number was progressively reduced by carrying out an analysis based on sound methodology. The initial 179 expressions were reduced to 35 items depending on their clarity, importance and frequency. A sample of 92 individuals from different centers in Spain and at different stages of the illness was selected to test the 35-item questionnaire. The items were reduced further, resulting in a 25-item instrument with good consistency and feasibility. Lastly, questionnaire development was completed in an observational prospective study with 263 patients, and the 10-item instrument obtained showed good psychometric properties. The outcomes obtained with regard to internal consistency and test-retest reliability indicate high homogeneity and good reproducibility throughout the questionnaire. Regarding the validity, just like other specific questionnaires for advanced cancer, the BOMET-QOL-10 has shown significant correlations with the ECOG and the EORTC-QLQ-C30 questionnaire. The procedure used to reduce the questionnaire, on both occasions, was the Rasch analysis based on item response theory. This has some advantages over the classical test theory (CTT) since it considers each answer as a probabilistic function, with the lineal probabilistic interaction of a person’s ‘‘ability’’ and a question’s ‘‘difficulty,’’ and it constructs a line of measurement with the items placed hierarchically, as well as being able to find people according to their competence (e.g., health status) (Prieto et al., 2003). The BOMET-QOL-10 is not an instrument that replaces the specific cancer questionnaires we have today, but is complementary to them for assessing HRQOL in patients with MBDN in clinical trials and daily clinical practice.
The Bone Metastases Quality of Life Questionnaire
11
Summary Points The Bone Metastases Quality of Life Questionnaire (BOMET-QOL) measures HRQOL dimensions relevant to patients with Bone Metastases.
It is the first bone metastases-specific HRQOL measure. It is a self-administered questionnaire intended for use in clinical research (trials) and in the routine monitoring of patients with bone metastases.
It is unidimensional and consists of 10 items that can be answered according to a Likert
scale with 5 categories (always, almost always, sometimes, rarely and never), scoring from 0 to 4, and the timeframe refers to ‘‘last week.’’ The items refer to HRQOL topics such as pain, mobility, vitality, sex life, worries about the future, perceptions and daily activities. The global score is calculated by adding up the answers obtained in each item, and may range from 0 (worst HRQOL) to 40 (best HRQOL). These scores are standardized to make it easier to interpret the scoring, with the final scores ranging from 0 (worst HRQOL) to 100 (best HRQOL). It is user-friendly, easy to complete by the patient and easy to assess by the professional. It has shown good measurement properties (cross-sectional validity, discriminant validity, sensitivity to change, reliability, etc.).
References Aaronson NK, Ahmedzai S, Bergman B, Bullinger M, Cull A, Duez NJ, Filiberti A, Flechtner H, Fleishman SB, M. de Haes JCJ, Kaasa S, Klee M, Osoba D, Razavi D, Rofe PB, Schraub S, Sneeuw K, Sullivan M, Takeda F. (1993). J Natl Cancer Inst. 85: 365–376. Adrover E, Allepuz J, Sureda A, Domine M, Barnadas A, Constela M, Lluch A, Ruiz M, Piera J, Mayordomo JI, Morales A, Mun˜oz M, Alcover J, Colomer R, Llombart A, Massutti B, Carballido J, Garrido P, Garcı´a R, Badia X, Lizan L, Gilabert M. (2005). J Outcomes Res. 9: 15–27. Avis NE, Smith KW, Hambleton RK, Feldman HA, Selwyn A, Jacobs A. (1996). Med Care. 34: 1102–1120. Badia X, Muriel C, Gracia A, Nu´n˜ez-Olarte JM, Perulero N, Ga´lvez R, Carulla J, Cleeland CS, Vesbpi G. (2003). Med Clin (Barc). 120: 52–59. Bezjak A, Taylor KM, Macdonald K, DePetrillo AD. (1998). Cancer Pre Control. 2: 230–235. Bunting RW, Shea B. (2001). Cancer. 92(Suppl. 4): 1020–1028. Cella D, Chang CH, Lai JS, Webster K. (2002). Semin Oncol. 29(Suppl. 8): 60–68. Cella DF, Tulsky DS, Gray G, Sarafian B, Linn E, Bonomi A, Silberman M, Yellen SB, Winicour P and Brannon J. (1993). J Clin Oncol. 11: 570–579. Cleeland CS. (1991). Cancer. 67: 823–827. Coleman R. (2004). Oncologist. 9: 14–27.
Djulbegovic B, Wheatley K, Ross J, Clark O, Bos G, Goldschmidt H, Cremer F, Alsina M, Glasmacher A. (2002). Cochrane Database Syst Rev. 4: CD003188. Ganz PA, Schag CA, Lee JJ, Sim MS. (1992). Qual Life Res. 1: 19–29. Lipton A. (2003). Curr Treat Options Oncol. 4: 151–158. Martinez MJ, Roque M, Alonso-Coello P, Catala` E. (2003). Cochrane Database Syst Rev. 3: CD003223. Oken MM, Creech RH, Tormey DC, Horton J, Davis TE, McFadden ET, Carbone PP. (1982). Am J Clin Oncol. 5: 649–655. Prieto L, Alonso J, Lamarca R. (2003). Health Qual Life Outcomes. 28: 1–27. Rasch G. (1993). Probabilistic Models for Some Intelligence and Attainment Tests. Paedagogiske Institut, Copenhagen. Reddi AH, Roodman D, Freeman C, Mohla S. (2003). J Bone Miner Res. 18: 190–194. Sureda A, Isla D, Co´zar JM, Ruiz M, Domine M, Margelı´ M, Edrover E, Ramos M, Pastor M, Martı´n A, Llombart A, Massuti B, Mun˜oz M, Barnadas A, Ferna´ndez J, Colomer R, Allepuz C, Gilabert M, Badia X. (2007). JME. 10: 27–39. Watson M, Law M, Maguire GP, Robertson B, Greer S, Bliss JM, Ibbotson T. (1992). Psychooncology. 1: 35–44.
207
12 The Impact of Weight on Quality of Life Questionnaire J. Manwaring . D.Wilfley 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210
2
Why Develop Quality of Life Measures in Obesity? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211
3 IWQOL and IWQOL-Lite Development and Psychometric Properties . . . . . . . . . . . 212 3.1 IWQOL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212 3.2 IWQOL-Lite . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213 4
IWQOL in Comparison with Other Obesity-Specific Instruments . . . . . . . . . . . . . . . . 217
5
IWQOL and Weight Loss . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218
6
IWQOL Among Subgroups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220
7
Conclusion and Future Directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
8
Information on Use of the IWQOL-Lite . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224
#
Springer Science+Business Media LLC 2010 (USA)
210
12
The Impact of Weight on Quality of Life Questionnaire
Abstract: Obesity is a complex disease with a multifactorial etiology that is reaching epidemic proportions worldwide. Assessments are needed to adequately capture the decrements in health and psychosocial well-being that are so often observed in > obese individuals. Accordingly, researchers are realizing the importance of quality of life as a measure of health status and weight loss treatment outcome in obese individuals. While weight loss is the primary outcome of interest for most obesity treatments, measuring quality of life is a vital component as individuals evaluate their treatment outcome. The Impact of Weight on Quality of Life (IWQOL) questionnaire is the first questionnaire developed to specifically assess the effects of obesity on health-related quality of life. The psychometric properties of the IWQOL are excellent, but given its length, a briefer version was developed, the Impact of Weight on Quality of Life-Lite Questionnaire (IWQOL-Lite). The five identified scales (Physical Function, Self-Esteem, Sexual Life, Public Distress, and Work) and the Total score have demonstrated excellent > reliability (0.90–0.94 for the scales, 0.96 for Total score), > validity, and utility with various groups of obese individuals. In this chapter, the rationale for the development of an obese-specific quality of life measure will be briefly reviewed, followed by the background and development of the IWQOL and IWQOL-Lite (including their psychometric properties), weight loss research using these instruments, and future research directions. It is concluded that the IWQOL-Lite is a reliable, valid instrument that can be utilized in various obese populations, and may be particularly useful as a treatment outcome measure. List of Abbreviations: BAROS, Bariatric analysis and reporting outcome system; BED, Binge eating disorder; BMI, > Body mass index (kg/m2); DDFC, Duke University Diet and Fitness Center; HRQOL, Health-related quality of life; IWQOL, > Impact of Weight on Quality of Life questionnaire; IWQOL-Lite, > Impact of Weight on Quality of Life-Lite questionnaire; SF-36, Medical Outcomes Study Short-Form Health Survey
1
Introduction
Obesity is a prevalent disease with significant negative effects on physical and psychosocial functioning. As determined by > body mass index (BMI) (kg/m2), recent estimates indicate that 32% of U.S. adults are obese (BMI 30); this prevalence has increased approximately 50% per decade over the last 30 years (Flegal et al., 2002). Among its myriad health problems, obesity is associated with hypertension, diabetes, and cardiovascular disease, all of which are among the leading causes of death in developed countries (e.g., Pi-Sunyer, 1995). Further, the psychosocial consequences of obesity are devastating. Obese adolescents have been found to have lower income, education, and marriage rates as adults than individuals with chronic physical conditions such as asthma, diabetes, and musculoskeletal deformities (Gortmaker et al., 1993). Health related quality of life (HRQOL) has been defined as the “physical, psychological, and social domains of health, seen as distinct areas that are influenced by a person’s experiences, beliefs, expectations, and perceptions” (Testa and Simonson, 1996). Generic measures of quality of life were utilized by clinicians before disease-specific instruments were developed in response to the need for more sensitive measures. The medical field has come to realize the importance of assessing psychosocial factors such as quality of life, as this construct may more accurately convey treatment efficacy, more finely note treatment outcome differences, aid physicians and patients in making treatment decisions, and allow
The Impact of Weight on Quality of Life Questionnaire
12
patients to comment subjectively on the outcome of treatment in addition to the objective treatment outcome markers (e.g., disease status; Kolotkin et al., 2001a). As such, numerous assessments have been developed to tap the dimensionality of health. This chapter will focus on the Impact of Weight on Quality of Life (IWQOL) questionnaire, a measure developed in response to the growing prevalence of obesity and need for sensitive and specific quality of life measures.
2
Why Develop Quality of Life Measures in Obesity?
The idea that the majority of > overweight/obese individuals would suffer from a poor quality of life may be a simple presumption, but some factors may be more associated with an impaired quality of life than others, and these factors will vary between individuals (Kolotkin et al., 2004). For example, one woman with severe obesity may struggle with activities of daily living, be exposed to work discrimination, and have difficulty in her interpersonal relationships because of her weight while another woman with the same level of obesity may experience some physical limitations but have a supportive work environment and only a few minor complaints regarding her interpersonal life. Thus, the typical medical evaluation points of disease status, in obesity most often measured by BMI (kg/m2), would not provide an adequate picture of the overall health of these two women. In addition to providing a patient’s perspective of his/her health for a more multifaceted view of treatment outcome, > Health-Related Quality of Life (HRQOL) measures are clinically useful in providing a nuanced, comprehensive view of the individual to aid clinicians in treatment planning. Having the patient’s perspective of his/her quality of life allows clinicians to more adequately discuss treatment options, and then evaluate treatment outcome in comparison to the patient’s baseline level of quality of life. Both general and disease-specific measures have been used to study obesity. The most commonly used HRQOL assessment instrument for a variety of medical problems is the Medical Outcomes Study Short-Form Health Survey (SF-36; Ware et al., 1993), which assesses eight discrete constructs including bodily pain and social functioning. The benefit of general HRQOL measures such as the SF-36 is that they allow for comparisons of various medical conditions across studies. However, their disadvantage is that they may not be sufficiently sensitive to detect minor treatment effects (Fontaine and Barofsky, 2001). Thus, researchers have recommended disease-specific measures for clinical use, and both diseasespecific and general quality of life measures for clinical trials (e.g., Kolotkin et al., 2001a; Wadden and Phelan, 2002). Studies that have assessed quality of life in obesity using generic measures have found greater impairment among obese adults (especially women) compared to community norms (e.g., Fontaine et al., 2000); a dose-response relationship between degree of overweight and HRQOL (e.g., Fine et al., 1999); and more impairment in physical domains than mental health domains (e.g., Barajas et al., 1998). However, as previously mentioned, generic instruments are most useful for general survey research, or comparisons of disease states, and are unable to assess the specific aspects most relevant to a particular disease. In response to these limitations of generic quality of life instruments, Kolotkin and colleagues, the authors of the IWQOL, aimed to create a reliable and valid measure of the impact of weight on quality of life, and in doing so, establish the aspects of quality of life impacted by weight and the changes noted in quality of life following treatment (Kolotkin et al., 1995).
211
212
12
The Impact of Weight on Quality of Life Questionnaire
3
IWQOL and IWQOL-Lite Development and Psychometric Properties
3.1
IWQOL
Development. In the development of the IWQOL, Kolotkin et al. (1995) interviewed approximately 20 outpatients in treatment for obesity at the Duke University Diet and Fitness Center (DDFC) to ascertain how their overweight status impacts their everyday lives. From this information the authors compiled 74 items which were modified according to feedback from other patients, and then divided into eight scales: Health (14 items), Social/Interpersonal (11 items), Mobility (10 items), Self-Esteem (10 items), Comfort with Food (9 items), Work (7 items), and Sexual Life (6 items). Items in the IWQOL begin with the phrase “because of my weight” and the participant chooses between the responses “always true” (scored as 5), “usually true,” “sometimes true,” “rarely true,” or “never true” (scored as 1), with higher scores representing poorer quality of life. The IWQOL was then administered to 64 other outpatients (37 women, 27 men) who completed the questionnaire on Day 1 of their visit, Day 2 of their visit (to assess > test-retest reliability), and Day 28 of their visit (to assess treatment effects from Day 1). Psychometric Properties. Item test-retest reliability ranged from 0.53 (“little attention to how much I eat”; Comfort with Food) to 0.92 (“ankles and lower legs swollen”; Health), with an average reliability coefficient of 0.75. Scale test-retest reliability fared even better, ranging from 0.81 (Comfort with food) to 0.93 (Social/Interpersonal, Self-Esteem), with an average reliability score of 0.89. Day 1 responses were used to measure the internal consistency of each scale, which averaged 0.87 and ranged from 0.68 (Comfort with Food) to 0.93 (Social Interpersonal, Self-Esteem). To measure treatment effects, 37 of the 64 outpatients (58%) completed the IWQOL on Day 28; their scores showed significant improvement on seven of the eight scales (greatest decrease/improvement on Mobility, p ¼ 0.0001); on the eighth scale – Comfort with Food – outpatients demonstrated significantly less comfort with food posttreatment, perhaps not surprisingly given the increased vigilance taught in behavioral weight loss treatment programs such as those offered by the DDFC. The authors of the IWQOL note that their studies support the clinical utility of the questionnaire, its item and scale score stability, and internal consistency of the items. Further, all but one of the scales (Comfort with Food) changed significantly in the expected direction after treatment. However, the authors report that the Comfort with Food scale had the weakest psychometric properties, exhibiting the poorest internal consistency (0.68), and thus posit deletion unless further studies suggest otherwise. Limitations of the original evaluation of the IWQOL include the lack of generalizability given the study sample used of treatment-seeking patients who were mostly Caucasian, upper-middle class, and morbidly obese. A later report in 1997 by Kolotkin and colleagues supported the > construct validity and further supported the internal consistency and treatment sensitivity of the IWQOL (Kolotkin et al., 1997). The IWQOL total and scale scores of 394 DDFC outpatients (243 women, 151 men) were compared to the scores on five questionnaires measuring aspects of health, and to patients’ medical records. The total IWQOL score was significantly correlated with the authordesigned quality of life single-item scale (r ¼ 0.46, p < 0.0001), and with the DUKE General Health scale (Parkerson et al., 1990, r ¼ 0.54, p < 0.0001). The IWQOL scale scores were significantly correlated with other similar assessments, except for two comparisons between the IWQOL scales and measures of similar constructs: IWQOL Social/Interpersonal with
The Impact of Weight on Quality of Life Questionnaire
12
DUKE Social Health, and IWQOL Health with cholesterol (although the latter may be expected given the hidden symptoms of hypercholesterolemia). The construct validity of the subscales Activities of Daily Living, Sexual Life, and Work could not be assessed since similar measures of these constructs were not administered. Internal consistencies reliability scores averaged 0.87 and ranged from 0.76 (Comfort with food) to 0.95 (Mobility). There were no significant differences in internal consistency between the authors’ 1995 study and this 1997 report. In examining treatment effects, both 2-week and 4-week treatment groups demonstrated significant improvements on all IWQOL scales. For men in the 2-week treatment group, no treatment differences were demonstrated in the Work and Mobility subscales. As with the original 1995 study, generalizability was limited given the demographic characteristics of the DDFC outpatients.
3.2
IWQOL-Lite
Development. Despite its early demonstrated clinical and research utility, clinical researchers commented on the burdensome length of the 74-item IWQOL for participants (requiring approximately 15 min to complete, e.g., Mannucci et al., 1999). Thus, in 2001 Kolotkin and colleagues developed a briefer version of the IWQOL with only 31 items, called the IWQOLLite (Kolotkin et al., 2001b). The development of this version improved upon the original version by using data from 1,987 individuals (randomly divided into a development sample and cross-validation sample) from a variety of settings who were more ethnically and socioeconomically diverse than the previous normative sample (see > Table 12-1). Obese participants who completed the assessments were part of an open-label phentermine-fenfluramine trial, a day treatment program for weight loss, an outpatient weight reduction studies or . Table 12-1 Demographics for IWQOL-Lite development study Study Open-label phen-fen
Group Obese
N
Gender
211 32 males 179 females
Day treatment
Obese
834 322 males 512 females
Weight-reduction studies/ programs
Obese
668 91 males
Gastric bypass
Obese
51 11 males
Total
Community
BMI
47.8 10.3
42.0 8.7
44.4 9.3
40.7 7.1
52.1 14.3
41.7 11.1
50.2 17.6
37.0 9.1
47.4 9.4
35.5 6.2
43.9 11.3
35.0 6.1
46.6 4.5
50.6 10.0
38.5 9.7
51.0 12.9
223 159 males
37.2 11.4
27.7 3.6
64 females
39.0 14.1
26.5 10.1
1987 615 males
47.3 14.1
37.2 10.8
45.9 14.3
36.6 9.4
577 females 40 females
Employees/friends
Age (in years)
1,372 females Adapted from Kolotkin et al. (2001b, p. 104). BMI body mass index
213
214
12
The Impact of Weight on Quality of Life Questionnaire
programs, or were undergoing gastric bypass surgery or were volunteers from the community. The majority of the participants (96%) provided BMI data, and all participants completed the IWQOL. To construct the IWQOL-Lite, the authors used data from the prior IWQOL studies to select the items and scales for the IWQOL-Lite, deleting items that correlated poorly with the scale score, total score, BMI, or changes in BMI (Kolotkin et al., 2001b). Factor analyses were then performed, and further deletions were made for those items not loading adequately on the derived factors. The resulting IWQOL-Lite was comprised of five scales: Physical Function (11 items, e.g., feeling short of breath, getting up from chairs), Self-Esteem (7 items, e.g., afraid of rejection, avoid looking in mirrors), Public Distress (5 items, e.g., fitting through aisles, experience ridicule), Sexual Life (4 items, e.g., avoid sexual encounters, difficulty with sexual performance), and Work (4 items, e.g., do not receive recognition, afraid to go on interviews), and correlated highly with the original IWQOL (0.97). As in the IWQOL, items in the IWQOL-Lite begin with the phrase “because of my weight” and the participant chooses between the responses on a Likert scale (from “always true” ¼ 5 to “never true” ¼ 1), but with the change that higher scores represent better quality of life. Psychometric Properties. The IWQOL-Lite demonstrated excellent reliability, with an overall reliability (Cronbach a coefficient) of 0.96 (Kolotkin et al., 2001b). The scales’ reliability coefficients were 0.94 for Physical Function, 0.93 for Self-Esteem, 0.91 for Sexual Life, and 0.90 for Public Distress and Work. As can be seen in > Table 12-2, correlations between the items and its designated scale are higher than any other scale, suggesting discreteness between scales. When corrected for overlap, the correlations between individual scales and the total score were in the tight range of 0.66 (Sexual Life) to 0.73 (Public Distress), indicating comparable contributions of the scales to the total score. All five scales and the total score of the IWQOL-Lite correlated significantly (p < 0.01) with change in BMI after 1 year (average weight loss: ~18% body weight), with the highest correlation seen in Physical Function and Total score (Kolotkin et al., 2001b). Further, even small changes in BMI (< 10%) were correlated with decreased IWQOL-Lite scores (with the exception of Work), ranging from effect sizes of 0.20 (Sexual Life) to 0.50 (Physical Function). Thus, the IWQOL-Lite is sensitive to small decreases in weight, and these results suggest that even modest weight loss is correlated with an increased quality of life. The IWQOL-Lite’s sensitivity was also demonstrated by the large effect sizes seen when comparing groups at each end of the weight spectrum (BMI < 25 versus > 40): 1.76 for Total score and Public Distress, 1.70 for Physical Function, 1.34 for Self-Esteem, 1.07 for Sexual Life, and 0.97 for Work. See > Table 12-3 for further information on 1-year changes. Since the original 2001 report, the IWQOL-Lite has undergone further psychometric testing. Kolotkin and Crosby (2002) evaluated the test-retest, discriminant, and > convergent validity of the questionnaire amongst 494 adult participants (341 females, 153 males) recruited from the community (Kolotkin and Crosby, 2002a). The sample was racially diverse with an average age of 38.1 (range 18–90). The participants’ average BMI of 27.4 (SD: 7.1; range: 18.6–73.0) fell in the overweight category (BMI 25). Overall, scores for this community sample were consistently lower/less impaired than the previous normative treatment-seeking samples (Kolotkin et al., 2001b). The authors examined psychometric properties of the IWQOL-Lite using the overall sample, and the overweight/obese participants only, and found internal consistency reliability, test-retest reliability, convergent validity, and discriminative validity. In the full sample, internal consistency a coefficients ranged from 0.82 (Work) to 0.94 (Self-Esteem), and 0.96 for the Total score. Test-retest intraclass correlation coefficients
The Impact of Weight on Quality of Life Questionnaire
12
. Table 12-2 Item-to-scale correlations for IWQOL-Lite Physical function
Selfesteem
Sexual life
Public distress
Picking up objects
0.84
0.43
0.50
0.61
0.54
0.75
Tying shoes
0.84
0.44
0.49
0.61
0.51
0.74
Getting up from chairs
0.84
0.44
0.47
0.64
0.55
0.75
Using stairs
0.84
0.46
0.48
0.63
0.53
0.76
Dressing
0.80
0.45
0.48
0.56
0.57
0.73
Mobility
0.84
0.48
0.49
0.62
0.61
0.77
Crossing legs
0.76
0.47
0.46
0.63
0.44
0.71
Feel short of breath
0.66
0.45
0.42
0.48
0.47
0.64
Painful stiff joints
0.61
0.32
0.39
0.38
0.43
0.55
Swollen ankles/legs
0.61
0.31
0.36
0.43
0.39
0.54
Worried about health
0.44
0.40
0.32
0.38
0.37
0.48
Self-conscious
0.47
0.84
0.53
0.51
0.48
0.69
Self-esteem not what it could be
0.47
0.86
0.54
0.49
0.51
0.69
Unsure of self
0.48
0.84
0.52
0.54
0.55
0.71
Do not like myself
0.41
0.78
0.54
0.44
0.47
0.63
Afraid of rejection
0.41
0.74
0.48
0.56
0.49
0.63
Avoid looking in mirrors
0.44
0.73
0.50
0.52
0.42
0.63
Embarrassed in public
0.50
0.73
0.54
0.56
0.51
0.69
Do not enjoy sexual activity
0.41
0.48
0.80
0.33
0.41
0.55
Little sexual desire
0.51
0.56
0.79
0.37
0.48
0.63
Difficulty with sexual performance
0.57
0.54
0.77
0.49
0.49
0.68
Avoid sexual encounters
0.51
0.60
0.83
0.45
0.49
0.67
0.42
0.55
0.35
0.62
0.41
0.57
Work Total
Physical function (a = 0.94)
Self-esteem (a = 0.93)
Sexual life (a = 0.91)
Public distress (a = 0.90) Experience ridicule Fitting in public seats
0.66
0.48
0.41
0.83
0.47
0.70
Fitting through aisles
0.67
0.52
0.43
0.87
0.48
0.73
Worry about finding chairs
0.65
0.47
0.38
0.83
0.47
0.69
Experience discrimination
0.48
0.61
0.37
0.64
0.44
0.62
Trouble accomplishing things
0.58
0.53
0.46
0.49
0.78
0.67
Less productive than could be
0.55
0.54
0.50
0.45
0.79
0.66
Work (a = 0.90)
215
216
12
The Impact of Weight on Quality of Life Questionnaire
. Table 12-2 (continued) Physical function
Selfesteem
Sexual life
Public distress
Work Total
Do not receive recognition
0.57
0.47
0.49
0.49
0.85
0.66
Afraid to go on interviews
0.50
0.49
0.41
0.47
0.72
0.61
Correlations between each item and its designated scale are in bold type (larger correlation equals stronger relationship). Also in bold type are correlations between each item and Total score. All correlations are corrected for overlap. Adapted from Kolotkin et al. (2001b, p. 105)
. Table 12-3 One-year change effect sizes in IWQOL-Lite scores by percent BMI loss group Percent BMI loss 21% (n = 58)
Overall (n = 160)
Physical function
0.50
0.62
1.20
0.81
Self-esteem
0.43
0.65
0.95
0.73
Sexual life
0.20
0.36
0.73
0.46
Public distress
0.28
0.47
0.62
0.50
Work
0.09
0.19
0.44
0.26
Total score
0.46
0.65
1.12
0.79
Scale
Cohen (1988) recommends effect sizes as: 0.20 as small, 0.50 as medium, 0.80 as large (Cohen, 1988). Adapted from Kolotkin et al. (2001b, p. 108). BMI body mass index
ranged from 0.81 (Public Distress) to 0.88 (Physical Function), and 0.94 for the Total score. To test convergent validity, the IWQOL-Lite scores were compared to SF-36 scores (among other measures) in overweight/obese subjects (BMI 25). The IWQOL-Lite Total score was most highly correlated with the SF-36 general health (0.58) and physical summary score (0.54), with the lowest correlation between the total score and the mental summary score (0.33). Finally, to test discriminative validity, scores on the IWQOL-Lite were compared with scores on the Marlowe-Crowne, the most widely used measure of social desirability, in overweight subjects (Crowne and Marlowe, 1960). Only one scale correlated higher than the absolute value of 0.19 with the Marlowe-Crowne (Self-Esteem, r ¼ 0.29, 8.4% shared variance), indicating minimal overlap between these two scales. This study furthered the psychometric qualities of the IWQOL-Lite by providing test-retest reliability, and psychometric data in a community sample of both normal-weight and overweight/obese adults. The psychometric properties of the IWQOL-Lite were also evaluated in individuals with schizophrenia and bipolar disorder, due to the high rates of obesity in this population (Kolotkin et al., 2006a). In this study, 111 individuals diagnosed with schizophrenia and 100 individuals diagnosed with bipolar disorder completed the IWQOL-Lite, among other measures. The questionnaire demonstrated excellent internal consistency (0.87–0.97) and test-rest reliability (0.74–0.95), as well as construct validity. The correlation between BMI and the IWQOL-Lite scores were lower for the schizophrenic group compared to previous psychometric evaluations of this measure. A similar study with obese individuals with and without type 2 diabetes found high, comparable internal consistency reliabilities between these two groups (Kolotkin et al., 2003a).
The Impact of Weight on Quality of Life Questionnaire
12
Administration and Scoring. The IWQOL-Lite can be completed in 3 min, and is appropriate for adults ages 18 and over who can read at a 6.3 grade level. In administering the IWQOL-Lite, raw scores for each scale can only be computed if at least 50% of the items for that scale are answered, and for the total score only if 75% of the answers for all items are completed. Thus, the authors recommend checking for unintentional missing data after administration (Kolotkin and Crosby, 2002b). The IWQOL-Lite was originally criticized for lacking interpretive value, since the level of decreased score that could be deemed clinically meaningful was unknown. This was later addressed, when the authors of the IWQOL-Lite used advanced statistical techniques to establish a 7.7–12 point total score increase (improvement) as clinically meaningful (adjusting for an individual’s baseline score; Crosby et al., 2004). Further information on administration and scoring can be found at the end of the chapter. IWQOL-Lite Summary. In conclusion, the IWQOL-Lite demonstrates an adequate scale structure with excellent internal consistency reliability, test-retest reliability, convergent validity, discriminative validity, and good construct validity using hypothesis-driven confirmatory factor analysis. Further, the questionnaire appears to be sensitive to change, and is applicable to a wide range of BMIs and populations. The authors recommend the IWQOL-Lite over the original IWQOL because of its brevity, its development with a heterogeneous sample and separate development and cross-validations samples, and its superior psychometric development and results (Kolotkin et al., 2001b). Limitations of the IWQOL-Lite include ambiguity regarding the subjective relevance and clinical significance of the items to the individual as items were constructed based on group feedback (Kolotkin et al., 2001b; Wadden and Phelan, 2002). In other words, the IWQOL-Lite equally weighs the individual items in determining the total quality of life score, whereas similar scores could have different personal impact depending on the person’s values and expectations (Mannucci et al., 1999). However, this could be viewed as a weakness shared by the majority of self-report measures. Another suggested weakness of the IWQOL-Lite is that the introduction, “Because of my weight. . .” could be leading so that a comparison of the questionnaire with and without these instructions would be useful (Wadden and Phelan, 2002). Finally, the measure’s utility may be confined to morbidly obese individuals given some of the item content (e.g., difficulty getting up from a chair or picking up objects), and because the scales and the total score of the IWQOL-Lite have not been able to significantly distinguish between lower BMI groups ( Table 12-4 for further details). Thus, unless future studies determine that other measures perform better than the IWQOL-Lite in assessing the impact of weight on quality of life in obese individuals, the IWQOL-Lite has the psychometric data to warrant its recommendation.
5
IWQOL and Weight Loss
Non-Surgical. Weight loss among obese persons has consistently been found to improve the impact of weight on quality of life, as measured by the IWQOL or IWQOL-Lite (e.g., Heshka et al., 2003). An improved quality of life as measured by the IWQOL or IWQOLLite appears to depend on the amount of weight loss rather than treatment type (Heshka et al., 2003), to occur even with modest weight loss (5–10% body weight; Foster et al., 2004; Samsa et al., 2001), and to deteriorate again with weight regain (Engel et al., 2003). In one study, a lower perceived quality of life (IWQOL-Lite) at baseline in women appeared to predict poorer long-term outcomes, with the Self-Esteem and Work subscales being particularly tied to weight loss success and/or completion (Teixeira et al., 2004). In a report on four studies investigating the effects of sibutramine versus placebo on weight loss, the Health, Mobility, and Activity of Daily Living subscales of the IWQOL were more likely to show changes than the Social, Work, Self-Esteem, and Sexuality subscales; weight losses of 5.01–10.0% were associated with 10-unit changes in the total score (Samsa et al., 2001). Finally, other studies comparing psychotropic medication to placebo have reported similar results of improved IWQOL scores in the treatment group compared to the placebo group (though the improved subscales differ across studies), with even the drug dosage affecting the amount of change in the IWQOL-Lite scores (Lustig et al., 2006). Both pharmaceutical and non-pharmaceutical weight loss studies demonstrate improved scores on the IWQOL or IWQOL-Lite following even modest weight loss. Surgical. As might be expected given the serious nature of weight loss surgical techniques, the literature suggests that patients presenting for gastric bypass surgery are significantly more impaired on the IWQOL or IWQOL-Lite than obese individuals seeking residential treatment (Stout et al., 2007) or obese individuals not seeking treatment (Kolotkin et al., 2003b). A graded pattern is even seen, where obese individuals presenting for gastric bypass surgery exhibit the greatest impairment on the IWQOL-Lite, followed by: individuals in day treatment programs, individuals in outpatient weight-loss programs, and participants in clinical trials (Kolotkin et al., 2002c). After surgery, patients have demonstrated significant improvement on all subscales and Total score of the IWQOL or IWQOL-Lite, almost reaching scores of obese
20
WRSM
Few
Few
Few
Many
Few
Few
Few
Few
Few
Many
Few
Studies using methoda
þþ þþþ þþþ þþþ 0 þþ þþþ þþþ
þþþ þþ þ þþ þþþ þþþ
þþþ
þþþ
þþþ
þþþ
þ
þþþ
þþþ
Results
þþþ
Testing thoroughness
þþþ
þþþ
þþ
þþ
þþþ
þþþ
þþþ
þ
þþ
þþþ
þþþ
Testing thoroughness
Validityb
þþ
þþ
þþ
þþ
þþ
þþ
þþ
þ
þþ
þþþ
þþ
Results
þ
þ
0
þþþ
þþþ
þþþ
þþþ
0
þþþ
þþ
þþþ
Testing thoroughness
þ
þ
0
þ
þþ
þ
0
0
þ
þþ
þþ
0
0
þ
þ
þ
0
0
0
0
þþ
0
Results Interpretabilityd
Responsivenessc
Modified from Duval et al. (2006, p. 358) a Few: 1–4 published studies have used the method; Many: 9þ or more different studies b Testing thoroughness: 0, no reported evidence of reliability or validity; þ, very basic information only; þþ, several types of test, or several studies have reported reliability/validity; þþþ, all major forms of reliability/validity testing reported. Results: 0, no numerical results reported; þ, weak reliability/validity; þþ, adequate reliability/validity; þþþ, excellent reliability validity c Testing thoroughness: 0, no reported evidence of responsiveness; þ, pre-posttreatment; þþ, controlled studies, but not randomized controlled trial; þþþ, randomized controlled trial. Results: 0, no changes; þ, few changes; þþ, many changes d 0, little information available on the interpretability of the scores; þ, moderate information available on the interpretability of scores; þþ, excellent information available on the interpretability of the scores Lewin-Tag, HSP the Health-related quality of life, health state preference (Mathias et al., 1997); OSQOL the Obese Specific Quality of Life (Le Pen et al., 1998); ORWELL 97 the Obesity Related Well Being (Mannucci et al., 1999); OAS-SF the Obesity Adjustment Survey-Short Form (Butler et al., 1999); OP-scale the Obesity-related Psychosocial problems scale (Karlsson et al., 2003); BAROS the Bariatric Analysis and Reporting Outcomes System (Oria and Moorehead, 1998); M-AQolQII the Moorehead-Ardelt Quality of Life Questionnaire II (Moorehead et al., 2003); OWLQOL the Obesity and Weight Loss Quality of Life Questionnaire (Niero et al., 2002); WRSM The Weight-Related Symptom (Niero et al., 2002)
6
7
BAROS
17
8
OP-scale
OWLQOL
20
OAS-SF
M-AQolQII
11
18
55
Lewin-TAG, HSP
ORWELL 97
31
OSQOL
74
IWQOL-Lite
Number of Items
IWQOL
Questionnaire
Reliabilityb
. Table 12-4 Comparison of the obesity-specific quality of life questionnaires
The Impact of Weight on Quality of Life Questionnaire
12 219
220
12
The Impact of Weight on Quality of Life Questionnaire
individuals from the community (Adami et al., 2005; Boan et al., 2004), with most improvements continuing 3 years post-surgery (Adami et al., 2005). Additionally, comorbid health conditions among patients seeking gastric bypass surgery were found to be significantly associated with the IWQOL-Lite subscales Physical Function, Sexual Life, and Total score, with patients having three or more conditions exhibiting poorer quality of life (Kolotkin et al., 2003b). In this study, only group status (seeking surgery vs. non-treatment-seeking controls), BMI, gender, and depression accounted for unique baseline variance in the IWQOL-Lite scores. Further demonstrating its sensitivity, a cross-sectional study that administered general and specific measures of quality of life (SF-36, IWQOL-Lite, and Bariatric Analysis and Reporting Outcome System (BAROS); Oria and Moorehead, 1998) to patients seeking gastric bypass surgery found improvements already evident on the Physical Functioning subscale of the IWQOL-Lite several weeks after surgery (Dymek et al., 2002). Between this point and 6 months post-surgery, all subscales and Total score of the IWQOL-Lite demonstrated significant improvement, and between 6 months and 1-year post-surgery, the Physical Functioning, Self-Esteem, Public Distress, Sexual Life (on a trend level), and Total score of the IWQOLLite exhibited significant improvements. The BAROS also showed significant differences on all subscales during these time periods. Notably, no subscales of the SF-36 demonstrated significant differences in this time period, suggesting the more sensitive nature of the IWQOL-Lite and BAROS. Although the American Society for Bariatric Surgery recommends a thorough assessment (including a quality of life measure such as the IWQOL-Lite) for patients seeking life-altering gastric bypass surgery (LeMont et al., 2004), surveys have shown that only 6–18.5% of clinical programs actually administer a quality of life measure (Bauchowitz et al., 2005; Walfish et al., in press). Clearly, this would be indicated given the impaired scores demonstrated by presurgery patients and the improvement on scores post-surgery; including this information in pre- and post-surgery protocols would provide clinicians with valuable treatment recommendations and outcome evaluation.
6
IWQOL Among Subgroups
Binge Eating Disorder. The IWQOL has been evaluated in subgroups of individuals who struggle with their weight. For example, individuals with binge eating disorder (BED) engage in frequent binge eating without compensatory behavior (e.g., vomiting); thus, BED patients are typically obese. The prevalence of BED among obese individuals seeking weight loss treatment ranges from 16 to 30% (de Zwaan, 2001). Because individuals with BED typically exhibit greater psychopathology than comparable obese individuals without BED, investigators have examined whether a diagnosis of BED impacts an individual’s quality of life as well. Two studies have administered the IWQOL-Lite to individuals with BED and treatmentseeking obese individuals without BED (de Zwaan et al., 2003; Rieger et al., 2005). One study found that individuals with BED seeking gastric bypass surgery were more impaired on the Self-Esteem, Sexual Life, and Work subscales (de Zwaan et al., 2003), while the other study found that individuals with BED were more impaired on all subscales except the Physical Function subscale (Rieger et al., 2005). In contrast, a study that controlled for both BMI as well as depression and overall psychopathology found that BED (as assessed by questionnaire self-report) did not significantly contribute any independent variance to the IWQOL-Lite
The Impact of Weight on Quality of Life Questionnaire
12
scores, and that the impact of BED on quality of life appeared, at least in this study, to be accounted for by the increased comorbid psychopathology in this group (Kolotkin et al., 2004). However, a gender-by-BED interaction was significant for Total score, Physical Function, and Public Distress, with women’s quality of life more adversely affected by the presence of BED. BMI, Gender, and Age. In the original 1995 IWQOL study, the relationship between the IWQOL scores and BMI and gender was examined among 181 DDFC outpatients (117 women, 64 men) with a mean BMI of 38.3 (Kolotkin et al., 1995). Women demonstrated significantly lower Self-Esteem than men (which was replicated in the 1997 study by the authors (Kolotkin et al., 1997). After controlling for the effects of BMI on gender, women evidenced significantly greater impairments than men on Self-Esteem and Sexual Life. Interestingly, these effects differed according to the severity of weight. In the lowest weight group (BMI < 32.7), significant differences emerged between the sexes on both Self-Esteem and Sexual Life; when BMI was between 32.7 and 39.8 significant differences were found between the sexes only on Self-Esteem; and when BMI > 39.8 no significant differences emerged on IWQOL scale scores between men and women. These gender differences have also been shown to vary according to treatment modalities (see > Table 12-5; Kolotkin et al., 2002c; Kolotkin et al., 2006b). In examining BMI and gender using the IWQOL-Lite, BMI at baseline was significantly correlated with all scale and Total scores (p < 0.001; Kolotkin et al., 2001b), with specific correlations as follows: Sexual Life, 0.30; Self-Esteem, 0.34; Work, 0.35; Physical Function, 0.61; Public Distress, 0.68; and Total score, 0.59. When divided into groups by BMI (40), all groups differed significantly from each other on the scales and Total score (p < 0.01) except for the following: Figure 12-1). However, it should be noted that BMI contributes the largest amount of variance to the IWQOL-Lite scores of all potential moderators examined thus far (Kolotkin et al., 2002c). When examining the effect of age, the authors found that as age increased, both sexes reported less effect of weight on quality of life (as measured by the IWQOL) in Self-Esteem and Social/Interpersonal Life; greater effect of weight on quality of life in Mobility; and women reported greater effect of weight on quality of life in Health. Researchers and clinicians thus need to take these gender and age differences into account when interpreting IWQOL or IWQOL-Lite scores from their patients.
7
Conclusion and Future Directions
The assessment of psychosocial correlates in obesity has increased almost as dramatically as the prevalence of obesity itself. With the increased interest in assessment and development of new measures comes a responsibility to “measure the measures” and ensure that frequentlyused assessments possess adequate reliability and validity. As such, the IWQOL and especially the IWQOL-Lite have consistently demonstrated excellent reliability and validity through
221
222
12
The Impact of Weight on Quality of Life Questionnaire
. Table 12-5 IWQOL-Lite scores by gender and treatment modality IWQOLLite scale
Community volunteers
Clinical trials
Outpatient weight-loss programs/ studies
Day treatment
Gastric bypass
All treatment modalities by gender
69.7 19.8
68.5 20.1
58.9 26.2
46.7 29.0
67.0 23.11
54.6 26.5
35.6 30.4
57.2 26.4
44.8 29.4
52.7 26.3
46.8 27.4
64.7 26.9
43.3 29.7
57.6 27.2
46.2 27.8
63.1 29.7
46.8 28.8
69.4 25.6
35.2 32.9
65.7 28.3
44.7 29.8
Physical function Women
80.4 21.4
Men
90.7 12.0
All 86.0 17.7 Subjects
77.4 18.7 a
72.3 19.7
70.7 18.7 a
68.8 19.9
b
c
72.5 24.72 c
Self-esteem Women
76.5 24.2
Men
92.7 14.3
All 85.3 21.1 Subjects
62.2 26.1
60.1 25.7
81.4 18.6 a
68.5 25.6
68.9 26.7 b
61.3 26.0
c
d
60.3 26.81 77.3 24.32 d
Sexual life Women
89.8 17.4
Men
97.4 8.4
All 94.0 13.8 Subjects
72.0 25.2
69.1 26.6
85.6 16.7 a
76.5 23.6
75.1 26.4 b
70.0 26.6
c
c
69.8 27.51 81.7 23.32 d
Public distress Women
93.0 15.3
84.1 18.8
80.8 22.1
73.2 26.0
43.2 25.6
Men
97.3 9.7
89.7 15.1
82.2 21.3
69.3 27.6
32.5 24.6
71.6 26.8
41.4 25.7
All 95.3 12.7 Subjects
a
86.0 17.9
a
81.0 22.0
b
c
80.0 23.6 83.8 23.2 d
Work Women
91.6 14.9
82.4 18.3
80.0 21.6
71.6 24.4
40.5 27.5
Men
95.1 11.8
84.0 18.0
78.6 20.8
72.6 22.3
40.1 32.7
72.0 23.5
40.4 28.4
61.9 20.7
45.3 22.0
63.4 20.6
37.4 22.5
62.5 20.7
44.0 22.2
All 93.5 13.4 Subjects
a
82.9 18.2
b
79.6 21.5
c
d
78.3 23.1 81.8 21.4 e
Total score Women
84.1 16.3
Men
93.6 9.5
All 89.3 13.9 Subjects
72.3 17.0
70.1 18.0
82.2 14.4 a
75.6 16.9
73.8 18.8 b
70.6 18.2
c
d
69.3 19.91 77.8 20.42 e
Adapted from Kolotkin et al. (2002c, p. 752). Cell entries represent unadjusted mean SD (lower score is more impaired). Treatment modality means with different letter superscripts (a–e) are significantly different at p < 0.05 with Bonferroni correction after controlling for age and body mass index. Gender means with different numerical superscripts (1 and 2) are significantly different at p < 0.05 after controlling for age and body mass index
numerous studies using male and female patient populations from differing ethnicities (e.g., White et al., 2004), treatment groups (e.g., Fontaine, 2002), BMIs (e.g., Kolotkin and Crosby, 2002a), and psychiatric profiles (e.g., Kolotkin et al., 2006a; Rieger et al., 2005). Its brevity benefits clinicians, researchers, and patients alike.
The Impact of Weight on Quality of Life Questionnaire
12
. Figure 12-1 BMI and IWQOL-Lite subscale scores for race and gender groups. From White et al. (2004, p. 952). Group means with a common letter do not differ from one another; group means without a common letter differ significantly (p < 0.05). Contrasts for IWQOL-Lite scores conducted after controlling for BMI (where appropriate). BMI body mass index
While the medical field has advanced significantly in its recognition of the psychosocial impact of obesity, the challenge remains for clinical researchers to increase awareness of obesity-specific quality of life measures such as the IWQOL-Lite among community physicians and psychologists who infrequently utilize these measures (Bauchowitz et al., 2005). Obesity treatment would benefit from a more subjective understanding of a patient’s experience and a more comprehensive picture of weight loss treatment outcome. While it is important to introduce obesity-specific measures such as the IWQOL-Lite to obese patients to aid in treatment planning and outcome assessment, clinicians and researchers also need to consider each patient’s unique constellation of presenting complaints in order to gauge the areas of greatest concern to the patient. Another future research direction is in the assessment of the impact of obesity on quality of life across the developmental spectrum, warranted by the increasing prevalence of overweight in children and adolescents, and the accompanying psychosocial sequelae. In these regards, it would be beneficial to adapt the IWQOL for children. The authors of the IWQOL have recognized the importance of a developmental perspective, as the IWQOL-Lite has recently been used to develop a similar questionnaire targeting adolescents ages 11 to 19 called the IWQOL-Kids (Kolotkin et al., 2006c). It appears promising, demonstrating good internal consistency and sensitivity to BMI and treatment group, but this measure should be examined further (Kolotkin et al., 2006c). For more information on the developmental perspective of obesity, please see the accompanying chapters in this publication, “Obesity stigmatization and quality of life in adolescents” and “Health-related quality of life in obese adolescents.”
223
224
12
The Impact of Weight on Quality of Life Questionnaire
Developing norms for the IWQOL-Lite appears indicated, given the differences seen across ages, gender, and BMI status. In addition, it will be vital for future research to explicate whether patients most value the amount of weight lost, the consequent improvement in comorbidities or quality of life, or some combination of these factors (Ballantyne, 2003). Knowing this answer, and how it may differ across gender, age, ethnicities, treatment groups, and BMI, will further the development, utility, and validity of the IWQOL-Lite and other obesity-specific assessments for clinicians and researchers alike.
8
Information on Use of the IWQOL-Lite
For scoring and other information on the IWQOL or the IWQOL-Lite, the reader should consult the IWQOL/IWQOL-Lite Manual, which can be obtained from the authors. For permission to use the IWQOL-Lite (licensing fee varies according to use), copyrighted by Duke University, contact: Ronette L. Kolotkin, [email protected] (telephone 919 493 9995) or Dennis Thomas, Ph.D., Associate Director, Office of Licensing & Ventures, Duke University, [email protected], telephone (919) 681–7580. Copyright of the IWQOL-Kids is owned by Ronette L. Kolotkin and Cincinnati Children’s Hospital Medical Center.
Summary Points Given the increasing prevalence of obesity worldwide, and the associated psychosocial
consequences of this disease, obesity-specific measures of the quality of life provide a better method of assessment for obese individuals than general measures of the quality of life. The 74-item IWQOL was developed by Kolotkin and colleagues, and was the first obesityspecific measure to assess one’s quality of life. While the IWQOL demonstrated excellent reliability and validity, Kolotkin and colleagues developed the 31-item IWQOL-Lite in response to concerns about the length of the IWQOL. The IWQOL-Lite has demonstrated excellent reliability, validity, and treatment sensitivity for various groups of obese individuals in dozens of studies. The IWQOL-Lite is a recommended measure for obesity researchers and clinicians who wish to assess the impact of weight upon a patient’s quality of life, especially in regards to treatment outcome. It is most sensitive and useful for obese patients (BMI 30).
References Adami G, Ramberti G, Weiss A, Carlini F, Murelli F, Scopinaro N. (2005). Behav Med. 31: 53–60. Ballantyne G. (2003). Obes Surg. 13: 954–964. Barajas G, Robledo M, Garcia T, Cuesta S, Garcia M, Cerrada S. (1998). Rev Esp Sal Pu´b. 72: 221–231. Bauchowitz A, Gonder-Frederick L, Olbrisch M, Azarbad L, Ryee M, Woodson M, Miller A, Shirmer B. (2005). Psychosom Med. 67: 825–832.
Boan J, Kolotkin R, Westman E, McMahon R, Grant J. (2004). Obes Surg. 14: 341–348. Butler G, Vallis T, Perey B, Veldhuyzen van Zanten S, MacDonald A, Konok G. (1999). Int J Obes Relat Metab Disord. 23: 505–511. Cohen J. (1988). Statistical Power Analysis for the Behavioral Sciences. Lawrence Erlbaum, Hillsdale, NJ.
The Impact of Weight on Quality of Life Questionnaire Crosby R, Kolotkin R, Williams G. (2004). J Clin Epidemiol. 57: 1153–1160. Crowne D, Marlowe D. (1960). J Consult Psychol. 24: 349–354. de Zwaan M. (2001). Int J Obes Relat Metab Disord. 25: S51–S55. de Zwaan M, Mitchell J, Howell L, Monson N, SwanKremeier L, Crosby R, et al. (2003). Compr Psychiatry. 44: 428–434. Duval K, Marceau P, Perusse L, Lacasse Y. (2006). Obes Rev. 7: 347–360. Dymek M, Le Grange D, Neven K, Alverdy J. (2002). Obes Res. 10: 1135–1142. Engel S, Crosby R, Kolotkin R, Hartley G, Williams G, Wonderlich S, Mitchell J. (2003). Obes Res. 11: 1207–1213. Fine J, Colditz G, Coakley E, Moseley G, Manson J, Willett W, Kawachi I. (1999). JAMA 282: 2136–2142. Flegal K, Carroll M, Ogden C, Johnson C. (2002). JAMA 288: 1723–1727. Fontaine K. (2002). Obes Res 10: 854–855. Fontaine K, Barofsky I. (2001). Obes Rev. 2: 173–182. Fontaine K, Bartlett S, Barofsky I. (2000). Int J Eat Disord. 27: 101–105. Foster G, Phelan S, Wadden T, Gill D, Ermold J, Didie E. (2004). Obes Res. 12: 1271–1277. Gortmaker S, Must A, Perrin J, Sobol A, Dietz W. (1993). N Engl J Med. 329: 1008–1012. Heshka S, Anderson J, RL A, Greenway F, Hill J, Phinney S, Kolotkin R, Miller-Kovach K, Pi-Sunyer FX. (2003). JAMA 289: 1792–1798. Karlsson J, Taft C, Sjostrom L, Torgerson J, Sullivan M. (2003). Int J Obes. 27: 617–630. Kolotkin R, Binks M, Crosby R, Ostbye T, Gress R, Adams T. (2006b). Obesity. 14: 472–479. Kolotkin R, Crosby R. (2002a). Qual Life Res. 11: 157–171. Kolotkin R, Crosby R. (2002b). The Impact of Weight on Quality of Life-Lite (IWQOL-Lite): User’s manual. Obesity and Quality of Life Consulting, Durham, NC. Kolotkin R, Crosby R, Corey-Lisle P, Li H, Swanson J. (2006a). Qual Life Res. 15: 587–596. Kolotkin R, Crosby R, Kosloski K, Williams G. (2001b). Obes Res. 9: 102–111. Kolotkin R, Crosby R, Pendleton R, Strong M, Gress R, Adams T. (2003b). Obes Surg. 13: 371–377. Kolotkin R, Crosby R, Williams G. (2002c). Obes Res. 10: 748–756. Kolotkin R, Crosby R, Williams G. (2003a). Diabetes Res Clin Pract. 61: 125–132. Kolotkin R, Head S, Brookhart A. (1997). Obes Res. 5: 434–441. Kolotkin R, Head S, Hamilton M, Tse C. (1995). Obes Res. 3: 49–56.
12
Kolotkin R, Meter K, Williams G. (2001a). Obes Rev. 2: 219–229. Kolotkin R, Westman E, Ostbye T, Crosby R, Eisenson H, Binks M. (2004). Obes Res. 12: 999–1005 Kolotkin R, Zeller M, Modi A, Samsa G, Quinlan N, Yanovski J, Bell S, Maahs DM, Gonzales de Serna D, Roehrig HR. (2006c). Obesity. 14: 448–457. Le Pen C, Levy E, Loos F, Banzet M, Basdevant A. (1998). J Epidemiol Commun Health. 52: 445–450. LeMont D, Moorehead M, Parish M, Reto C, Ritz S. (2004) Suggestions for the Pre-surgical Psychological Assessment of Bariatric Surgery Candidates. Allied health Sciences Sections Ad Hoc Behavioral Health Committee, American Society for Bariatric Surgery. Lustig R, Greenway F, Velasquez-Mieyer P, Heimburger D, Schumacher D, Smith D, Smith W, Soler N, Warsi G, Berg W, Maloney J, Benedetto J, Zhu W, Hohneker (2006). Int J Obes. 30: 331–341. Mannucci E, Ricca V, Barciulli E, Bernardoa M, Travaglinia R, Cabras P, Rotella C. (1999). Addict Behav. 24: 345–357. Mathias SW, CL, Colwell H, Cisternas M, Pasta D, Stolshek B, Patrick D. (1997). Qual Life Res. 6: 311–322. Moorehead M, Ardelt-Gattinger E, Lechner H, Oria H. (2003). Obes Surg. 13: 684–692. Niero M, Martin M, Finger T, Lucas R, Mear I, Wild D, Glauda L, Patrick D. (2002). Clin Ther. 24: 690–700. Oria H, Moorehead M. (1998). Obes Surg. 8: 487–499. Parkerson G, Broadhead W, Tse C. (1990). Med Care 28: 1056–1072. Pi-Sunyer X. (1995). Medical complications of obesity. In: Fairburn C, Brownell K (eds.) Eating Disorders and Obesity: A Comprehensive Handbook. The Guilford Press, New York, pp. 401–405. Rieger E, Wilfley D, Stein R, Marino V, Crow S. (2005). Int J Eat Disord. 37: 234–240. Samsa G, Kolotkin R, Williams G, Nguyen M, Mendel C. (2001). Am J Man Care. 7: 875–883. Stout A, Applegate K, Friedman K, Grant J, Musante G. (2007). Surg Obes Relat Dis. 3: 369–375. Teixeira P, Going S, Houtkooper L, Cussler E, Metcalfe L, Blew R, Sardinha L, Lohman T. (2004). Int J Obes. 28: 1124–1133. Testa M, Simonson D. (1996). N Engl J Med. 334: 835–840. Wadden T, Phelan S. (2002). Obes Res. 10(S1): 50S–57S. Walfish S, Vance D, Fabricatore A. (2007). Obes Res. 17(12): 1578–1583. Ware J, Snow K, Kosinski M, Gandek B. (1993). SF-36 Health Survey: Manual and Interpretation Guide. The Health Institute, New England Medical Center, Boston. White M, O’Neil P, Kolotkin R, Byrne T. (2004). Obes Res. 12: 949–955.
225
13 The Quality in Later Life Questionnaire S. Evans 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228
2
QOL Measurement and Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229
3
Why a New Measure? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229
4
The QuiLL and Its Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231
5 QuiLL Development and Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235 5.1 Qualitative Interviews . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 238 5.2 Instrument Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239 6
What Did We Learn? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240
7
Applications of the QuiLL Assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241
8
The QuiLL’s Status as a QOL Measure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 244
#
Springer Science+Business Media LLC 2010 (USA)
228
13
The Quality in Later Life Questionnaire
Abstract: Having the ability to appropriately measure the life quality of > older people has potential benefit for all of those engaged in the delivery of health and social care, including policy makers, planners, clinicians and other practitioners, and patients/service users and their carers. While there are a plethora of instruments available for the measurement of quality of life (QOL) very few relate specifically to older people. To date studies have focused almost exclusively on treated populations in hospital or residential settings and upon aspects of life quality that can be attributed directly to illness or treatments, rather than on the wider impact on family and social life, safety, housing and finances etc. The Quality in Later Life (QuiLL) Assessment was designed to fill this gap. It is a short > operational measure that assesses objective life circumstances and subjective feelings relating to nine > life domains (including health and mental health) that were identified as essential attributes of life quality by older people, carers, professionals and academics in the field; it has good psychometric properties and can be used routinely in operational practice, in evaluations of services and clinical or treatment interventions, and in health services and epidemiological research. The measure has been well received in the field and is already used widely in research and practice. In this chapter the author summarizes the QuiLL’s development and validation process, outlines the lessons learned about QOL and its measurement in older people populations and describes the instrument’s potential applications in routine health and social care practice. Examples of its service and research relevance, findings and presentational attributes are also given. List of Abbreviations: DALY, disability adjusted life year; HRQOL, health related quality of life; QALY, quality adjusted life year; QOL, quality of life; QuiLL, quality in later life
1
Introduction
The development of interest in quality of life assessment in older adults can be seen as one part of a more general movement to focus on positive aspects of ageing, with the aim of promoting well-being for future generations (Bowling, 1995). Enabling older people to live longer either independently in their own homes or with their carers is a common objective of health and social policy. Nevertheless, achieving these objectives depends in part upon the ability of services to assess, monitor, support and review the quality of life (QOL) of the person and their carer, as well as on their responsiveness in accommodating changing health and > social care needs. Having the ability to appropriately measure the life quality of older people has potential benefit for clinicians and other practitioners, service managers and planners, as well as patients/service users and their carers (Chesterman et al., 2001). At an individual level, focusing on the monitoring and maintenance of life quality, may prevent deterioration of both physical and mental health, and could reduce the need for more intensive and expensive forms of care and treatment. At an aggregate level, QOL data could be used to compare the life quality of older people in different community, residential or treatment settings in order to understand the relative cost and outcomes benefits of different interventions. Also, since clinical and > service interventions can impact upon individual QOL the availability of these types of data could be used to inform service developments that promote productive and successful ageing.
The Quality in Later Life Questionnaire
2
13
QOL Measurement and Models
The development of QOL indicators is important to policy and service development (Hagerty et al., 2001), evaluation and planning in facilitating effective measurement of the impact of new initiatives, treatments and interventions. Nevertheless, the term QOL has multiple conceptualizations and is associated with a wide range of theoretical models and measures. In health, interest in QOL stemmed from a desire to evaluate the impact of clinical interventions and to assess the relative merits of different health systems using patient-centered indicators rather than more traditional service outcome measures such as morbidity, mortality or the number of patients treated (O’Connor, 1993). This interest was stimulated by the recognition that alleviation of symptoms and prolongation of patients’ lives were not the only criteria for success (Cummins et al., 2004; Salek, 1998). As a result, many of the QOL models and measures used in the evaluation of health treatments and interventions focus on those aspects of life quality that can be attributed directly to illness or treatments, and as such are best described as healthrelated quality of life (HRQOL) (Namjoshi and Buesching, 2001; Spiro and Bosse, 2000). Others focus on specific aspects of functioning rather than on the general effects of a disease (e.g., asthma or prostate cancer) and are referred to as disease-specific QOL measures; highly specific measures known as quality-adjusted life years (QALYs) and disability-adjusted life years (DALYs) are also used in outcomes evaluation, epidemiology and cost-effectiveness analyses. In contrast, those interested in the social construction of illness and disease often regard QOL in more generic terms as a broad concept that incorporates all aspects of life, referring to the sense of well-being and satisfaction experienced by people (Andrews and Withey, 1976; Cameron et al., 2006; Campbell et al., 1976) under their current life (or health) conditions (Lehman, 1983). There is evidence that for some specific diseases (such as rheumatoid arthritis for instance) such generic measures are able to demonstrate impacts on overall life-quality almost as well as disease-specific measures (Anis et al., 2005). It is this generic conceptualization of life quality on which the QuiLL (Quality of Life in Later Life) assessment is based. The conceptual model on which the QuiLL is founded derives from the works of Campbell et al. (1976) and Lehman (1983). The model developed by the author (presented in > Figure 13-1) assumes that overall life quality is determined by a combination of personal characteristics and attributes, material or objective circumstances and domain-specific > subjective well-being. The ideas that underpin this conceptualization of life-quality mirror the concerns of normal community, family, social and economic life (Cameron et al., 2006; Oliver et al., 1996). The concept usually incorporates the basic essentials of life whilst also recognizing the importance of less tangible values like sense of achievement, fulfillment of potential, reward for efforts made and for stimulation and security in life (Campbell et al., 1976).
3
Why a New Measure?
While there are a plethora of instruments available for the measurement of QOL, at the time of the QuiLL development (1999–2002) very few related to older people and in the interim very few new measures have been introduced that focus specifically on older people (see Chapter 24 in this volume). Much of the early work about QOL in older people was conducted on treated populations in hospital settings or > nursing homes (Courtney et al., 2003), and tended to focus entirely
229
230
13
The Quality in Later Life Questionnaire
. Figure 13-1 An alternative model of life-quality (Evans, 2004). In this model of life quality single arrows are used to illustrate direct and indirect associations between personal characteristics, personal and clinical attributes (e.g., depression and disposition) and objective life circumstances, and domain-specific and ‘‘general’’ or ‘‘overall’’ life quality. Double arrows illustrate reciprocal relationships between domain-specific and ‘‘general’’ or ‘‘overall’’ life quality
on health-related quality of life taking no account of other aspects such as family, social life, housing or finances. LEIPAD, an internationally applicable instrument to assess quality of life in the elderly had been published at this time (De Leo et al., 1998) but as it included more health and functioning domains than social or generic domains, it did not fulfill the requirements of our favoured QOL model. Many studies used disease-specific instruments designed for use with general adult populations rather than older people. Studies often elicited staff rather than the views of older people themselves and tended to exclude people with dementia because staff believed that these individuals did not have the capacity to respond, despite recent evidence to the contrary (Mozley et al., 1999; Trigg et al., 2007a, 2007b). Like others (Farquhar 1995; Hyde et al., 2003) the development team (including the current author, Professor Peter Huxley, Professor Sube Bannerjee, Claire Gately and Alyson Smith) felt that the QOL concept should not be limited to health alone, and were convinced by evidence based on comparisons between health-related and global QOL measures that assumptions about QOL could not be based upon measures of health status (Covinsky et al., 1999). Therefore, they sought to develop a generic QOL assessment that would complement rather than substitute for measures of disease-specific or health-related QOL. This decision has been supported by subsequent research, for instance a study of older people’s own views by Xavier and colleagues (2003), which concluded that: "
Health seems to be a good indicator of negative quality of life, though an insufficient indicator of successful elderliness (p 31)
Relatively recent research by Michalos and colleagues (2007), confirms that health-related concepts are related to but are not synonymous with life satisfaction:
The Quality in Later Life Questionnaire "
13
Among other things,.. (our).. results clearly show that respondents’ ideas about a generally healthy life are different from, but not independent of, their ideas about a happy, satisfying or contented life, or about the perceived quality of their lives or their subjective wellbeing (p 127)
Given that the meaning of QOL is thought to change over a life-span (Faden and German, 1994), can vary by elderly age-group, place (Farquhar, 1995; Mollenkopf et al., 2004; Knesebeck et al., 2007) and ethnicity (Bajekal et al., 2004), and is influenced by available social support (Blane et al., 2004) and activities (Menec, 2003; Mozley et al., 1999; Wilhelmson et al., 2005) it is important that any instrument claiming to the assess QOL of older people reflects older peoples’ understanding of the term and the factors that contribute to or detract from life quality in later life. Therefore, the views of older people and those concerned with the quality of life of older people were central to the development of the QuiLL, ensuring that the measure captured issues that were of direct relevance to older people and was capable of assessing the impact of ageing and associated illnesses and diseases on older people’s lives. The need to retain the views of the older person themselves in the understanding and measurement of the quality of their life is now well documented (Bowling and Gabriel, 2004a, b; Westerhof et al., 2001; Wilhelmson et al., 2005; Xavier et al., 2003). Finally, it was clear from literature reviews that most of the available instruments had been designed for research rather than operational use, and that there was no standardized and systematic way of assessing older people’s QOL in routine practice. On the basis of their previous experience of quality of life assessment in mental health services the research team felt that the development of a new QOL measure for use routinely in older people’s services would be valuable in individual treatment and care planning and reviews, for clinicians and practitioners, older people and their carers, and managers and service planners. Therefore, the QuiLL was developed as a reliable life quality assessment that can be used easily by care staff as part of their assessment and monitoring procedures; while it was developed primarily for use in community based care services it can also inform decisions about the future needs of people being discharged from hospital to home or residential care settings. A number of studies of care interventions have demonstrated improved quality of life, using generic measures (Brandi et al., 2004; Chan et al., 2005a, b; Chesterman et al., 2001) and there is a growing interest in the way in which the immediate physical care environment may be an important determinant of quality of life (Barnes, 2002; Kearney and Winterbottom, 2005; Sugiyama and Thompson 2005; Tang and Brown 2005).
4
The QuiLL and Its Properties
The Quality in Later Life (QuiLL) Assessment is a short operational measure of life quality aimed at people aged 65 and over. It includes demographic details and 27 questions covering objective life circumstances and subjective feelings relating to the nine life domains shown in > Table 13-1; these life domains were identified as essential attributes of life quality by older people, carers, professionals and academics in the field. The QuiLL incorporates previously validated items wherever possible and includes the Andrews and Withey (1976) seven-point ‘‘delighted-terrible’’ scale as a self-rated indicator of subjective well-being, on which a low score indicates poor life quality and a high score indicates a good level of life-quality; the inclusion of this validated scale facilitates comparison with numerous other studies of life-quality in treated and general populations.
231
232
13
The Quality in Later Life Questionnaire
. Table 13-1 QuiLL structure and content Subjective well-being (seven-point scale) How do you feel about . . .? Demographic details your life as a whole, today?
Disposition, Role of religious faith Objective life circumstances
how you occupy your time
Daily activities
your friendships
Friends, Loneliness
your financial situation
Difficulty paying bills
your accommodation
State of repair of home
your living arrangements
People you live with
your personal safety
Safety and security at home
your family relationships
Family contact
your marriage/relationship your physical health
Long standing illness or disability
your mental health
Depression
Your neighborhood
Access local amenities
The amount of independence you have The amount of influence you have (over your own life) The QuiLL contains items relating to personal characteristics, personal and clinical attributes, categorical indicators of objective life circumstances in a life domains, and domain-specific and general indicators of subjective life quality, assessed using a seven-point scale (1 = terrible; 7 = delighted)
For those who are interested, the QuiLL’s psychometric properties are good (as reported in detail elsewhere by Evans et al. (2005); the alpha coefficient for subjective well-being was 0.88 (n = 1,044) and in tests of inter-rater reliability kappa values of one were common; reliability between research and professional staff was not acceptable and requires further work, but criterion validity with the Spitzer Quality of Life Index (Spitzer et al., 1981) was reasonable at 0.64. Unlike some measures, for which there are concerns about the ability to demonstrate relationships between objective circumstances and subjective reporting of well-being (Atkinson et al., 1997) the QuiLL and other similar measures are capable of demonstrating associations between objective and subjective indicators at domain specific and global levels (Evans et al., 2005, 2007). Examples based on the QuiLL (presented later in > Table 13-3) demonstrate the QuiLL’s ability to show associations between overall QOL and living in a safe environment, having access to local amenities, having friends to turn to, being able to pay bills, longstanding illness/disability, family contact and the state of repair of one’s home; similarly at a domain level the QuiLL was able to demonstrate associations between subjective feelings about social life and having friends, being lonely, being able to access local amenities, family contact and living in a safe environment. In the original QuiLL study it was difficult to demonstrate changes over time in subjective QOL, possibly due to the lack of actual objective change (60% of cases did not experience any change in objective conditions) that
The Quality in Later Life Questionnaire
13
occurred in the follow-up period, which might have been too short to allow measurable changes to emerge. While some might argue that this finding suggests the QuiLL is not responsive to change we do not believe that to be the case, for two main reasons: first analyses of QuiLL data showed that where the domain-specific objective circumstances of service users had improved subjective QOL ratings were statistically significantly higher and where circumstances had deteriorated subjective ratings were statistically significantly lower than for those not experiencing these changes (Evans et al., 2002); secondly the Andrews and Withey (1976) subjective well-being items that the QuiLL incorporates and other measures using the same scales have a proven ability to demonstrate changes in treated and untreated groups (Huxley et al., 2001; Evans et al., 2007). At an individual level the objective items included in the QuiLL provide a clear insight into an older person’s current situation in terms for example of social, accommodation or financial credits or deficits, but it is only when examined alongside the individual’s subjective assessment of that aspect of their life that the individual’s need can be determined. For example, a person might have many friends but crave something more in relation to their social relationships, which might be accessed via support groups for people with the same clinical needs as them or engagement in some other form of social network. Alternatively, a person might not be able to access local amenities independently but be very pleased to be living in their neighborhood, where they might be surrounded by friends and family who can support them in some way. At an aggregate level, it is possible using QuiLL data to explore the nature of the associations between for example loneliness and feelings about overall life quality, to demonstrate the impact of treatments and service interventions on life quality, and to explain what factors are most influential in maintaining high levels of QOL in later life. The following examples illustrate some of the ways in which QuiLL data can be used, providing comparisons between older people receiving care services in two areas of South-East England (one a generic service for older people in London and the other a service dealing with older people with mental health problems in a rural area) and older people in the general population in South London. The data presented in > Figure 13-2 relate to self-reported feelings of loneliness and indicate high levels of loneliness among older people, particularly those in receipt of care services living in London. More than half of this group reported feeling lonely compared to 38% of older people in the general population in a nearby area of South London, and 42% of older people receiving care services in a rural area of South East England (w2 = 6.55 (df 2) p < 0.001). > Figure 13-3 illustrates subjective QOL profiles for the three older people samples and illustrates that while the three older people samples did not differ in terms of general life quality (top row) or quality of life in family, living arrangements and safety domains, the groups did differ in other aspects of life quality. The similarity in shape of the profiles for older people in receipt of care services in London and the other South East region indicates that these samples were more alike each other (as one might expect) than they were like older people in the general population. The most significant between-sample differences were observed in finance, health and occupation of time domains. Between sample differences in finance (F = 59.5, p < 0.001) related to older people in receipt of care services being more satisfied with their financial situation (London: mean 4.9 (sd 1.1) p = 0.016; other South East: mean 5.3 (sd 0.8) p < 0.001) than older people living in the community in South London (mean 4.2, sd 1.4;), perhaps
233
234
13
The Quality in Later Life Questionnaire
. Figure 13-2 Reported loneliness in care-service and general population samples of older people. This figure compares the proportion (%) of older people reporting feelings of loneliness in three samples – people receiving care services in London, people receiving care services in Essex and older people in the general population in London. While most older people did not report feeling lonely, more than half of those receiving care in London were lonely
. Figure 13-3 Subjective QOL by sample. This figure illustrates the QOL profiles based on subjective ratings of domain and general life quality for the three older people samples – a score of 1 indicates poor QOL and 7 a good QOL. Scores of 5 and above indicate at least some degree of satisfaction (i.e., mostly satisfied)
The Quality in Later Life Questionnaire
13
because the recipients of care were more likely to be receiving their full benefit entitlement. Sample differences in the neighborhood domain (F = 9.65, p < 0.001) related to the general population being less satisfied with their neighborhood (mean 5.2, sd 1.0, p = 0.010) than older people in receipt of care in the rural location (mean 5.7, sd 1.0), understandably given the deprived nature of the urban area, but the general population’s scores did not differ significantly from those of care recipients in their own area (mean 5.2 (sd 1.2) cf 5.5 (sd 1.4)). Understandably, the general population sample were more satisfied with their health (F = 56.3, p < 0.001) and the way they were able to occupy their time (F = 37.36, p < 0.001) than people in receipt of care (Health: general population mean 4.9 (sd 1.2) cf London cases mean 4.3 (sd 1.3; p = 0.001) and non-London cases mean 4.4 (sd 1.2; p = 0.013); Occupy time: general population mean 5.0 (sd 1.2) cf London cases mean 4.4 (1.3; p = 0.001) cf non-London cases 4.5 (sd 1.2; p = 0.026)). The general population sample were also more satisfied in relation to the self domain (independence and influence) than the recipients of care in London (mean 5.2 (sd 1.1) cf 4.6 (sd 1.3); p < 0.005) but not the non London cases (mean 4.8 (sd 1.1). Given that self-reported loneliness and subjective QOL ratings differed significantly between samples (as did depression – not presented here), that many subjective QOL ratings were statistically significantly lower for people who reported feeling depressed or lonely (also not presented here) and that the shape of QOL profiles was very similar for people who reported feeling depressed and those reporting feeling lonely, it was important to examine the nature of these associations with QOL, when other intervening variables were controlled for. Linear regression models were used to determine what explained domain-specific, general and ‘‘overall’’ (calculated as the mean of domain ratings) life quality, using research interviews for recipients of care in London and the rural area, and survey data. Age, sex, marital status, ethnicity, living status, disability, depression, loneliness, disposition, having friends to turn to, difficulty with household bills, state of repair of home, safety in the home, access to local amenities, sum of activities undertaken and sample group (survey and research interviews) were entered as independent variables. In many of the domains, substantial amounts of variance were explained (compared to most other studies of this type), and the ‘‘overall’’ model explained 43% of the variance (> Table 13-2). Loneliness and to a lesser extent depression variables appeared most frequently in the regression models (see > Table 13-3), although the magnitude of associations with the various aspects of life quality were sometimes similar. Loneliness made the major contribution to the explanation of variance for ‘‘life in general,’’ ‘‘life overall’’ and in family, living situation, occupation of time and ‘‘self ’’ domains; depression made the major contribution to the explanation of variance in the health domain. The models also showed that contact with family or with friends was important, as was having access to local amenities. Therefore, improving people’s ability to access local amenities and reducing feelings of depression and loneliness is likely to impact on people’s QOL. These data suggested that social interventions such as providing people with welfare benefits advice, helping them to improve their home security or improving the state of repair of their home may have an important role to play in improving people’s QOL.
5
QuiLL Development and Testing
The development and validation process (which has been described elsewhere (Evans et al., 2005) and is summarized here) involved several stages including literature review and
235
236
13
The Quality in Later Life Questionnaire
. Table 13-2 Regression Models of generic and domain life quality QOL Domains Life Final model
Variables Loneliness
R2 (AR2) 0.31 (0.30)
Beta
95% CI
P
0.33
1.20 to
0.60 Tables 15-5 and > 15-6. Most domains had a negative skew, with distributions showing a ‘‘pile-up’’ of respondents scoring at the positive pole of quality of life. A mild ceiling effect was observed for the environmental domain (12.7%). Other ceiling
15
Development and Assessment of Chinese General Quality of Life Instrument
. Table 15‐3 The proportion of total variance attributed to principal component factors for WHO-QOL-100, SF36, and Chinese QOL-35 using factorial analyses (N = 135) % of total variance distributed by principal component factors Instrument
First
Second
Third
Fourth
Fifth
Sixth
Seventh
Accumulated, %
QOL-35
18.1
12.5
9.3
8.5
6.9
6.3
4.9
66.5
WHO-QOL-100
27.1
5.5
5.3
3.4
3.2
2.9
2.7
50.3
SF-36
31.9
9.6
6.4
5.7
4.7
3.6
3.5
65.3
. Table 15-4 The Spearman correlation coefficients of each domain of Chinese QOL-35 with corresponding domains of WHO-QOL-100 and SF-36 Domains
WHO-QOL-100
SF-36
General
0.577
0.605
Physical
0.730
0.579
Independence
0.668
0.657
Psychological
0.633
0.535
Social
0.669
0.307
Environmental
0.587
a
quality of life transition
a
0.374b
Total
0.805
0.745
a
There was no item on quality of life transition in WHO-QOL-100 and no item on environmental domain in SF-36 Scores for quality of life transition in SF-36 were replaced by the scores for health status transition
b
. Table 15-5 The ceiling effect of Chinese QOL-35 among 1,356 participants Statistics in score of quality of life Mean (SD)
Range
General
61.7 (19.4)
0.0–100.0
0.4
8.6
50.0
52.0
79.0
Physical
77.0 (18.2)
5.0–100.0
0.1
9.0
65.6
81.4
91.6
Independence 85.9 (15.4)
0.0–100.0
0.1
6.7
81.5
90.8
95.9
Psychological
69.3 (17.3)
0.0–100.0
0.1
2.8
58.4
71.3
81.2
Social
75.7 (16.1)
7.1–100.0
0.1
6.3
64.3
76.6
89.6
Environmental 65.6 (20.4)
0.0–100.0
0.2
12.7
50.0
64.5
79.0
76.9 (12.1) 16.4–100.0
0.1
0.1
70.7
78.7
85.5
Total a
The proportion of those with min score The proportion of those with max score
b
% Mina % Maxb 25th %tile 50th %tile 75th %tile
Domains
275
276
15
Development and Assessment of Chinese General Quality of Life Instrument
effects were less than 10%. Floor effects were negligible for all domains of QOL-35. Cronbach’s a for the QOL-35 ranged from 0.88 for independence domain to 0.71 for environmental domain; mean a was 0.72.
6.2
Construct Validity
Construct validity was reassessed using principle components analyses. About 59.4% of the total variance could be contributed by the seven selected principle factors, which was close to 66.5% in the small sample (135 participants). So the construct validity is considered stable for the Chinese QOL-35 (see the > Tables 15‐3 and > 15‐7).
6.3
Discriminatory Validity
Discriminatory validity, also named known-groups validity or sensitivity, is based on the principle that certain specified groups of patients may be anticipated to score differently from others, and the instrument should be sensitive to these differences. The prevalence of several . Table 15-6 The Cronbach Alpha of Chinese QOL-35 in small and large samples Domains
Small sample (n = 135)
General
Large sample (n = 1,356)
0.79
0.72
Physical
0.83
0.75
Independence
0.88
0.88
Psychological
0.83
0.74
Social
0.79
0.77
Environmental
0.68
0.71
Total
0.93
0.91
. Table 15‐7 The Variance explained by the seven principal components extracted from the items in Chinese QOL-35 in 1,356 participants Extraction sums of squared loadings Principal component factors
Total
% of variance
Cumulative %
1
9.263
27.243
27.243
2
3.989
11.732
38.974
3
2.089
6.145
45.119
4
1.405
4.132
49.251
5
1.222
3.595
52.846
6
1.155
3.397
56.244
7
1.072
3.153
59.397
Extraction method principal component analysis
Development and Assessment of Chinese General Quality of Life Instrument
15
. Table 15‐8 Prevalence rate of severe chronic diseases by quartiles of total quality of life score using Chinese QOL-35 (Wu et al., 2005) Quartiles of total score of quality of life
Prevalence Rate of Diseases, %
N
Total
Stroke
CHD
Respiratory Liver Endocrine symptoms disease Kidney Tumors disease
Quartile 1
339 53.1
8.8
9.4
26.3
6.2
8.6
8.6
3.8
Quartile 2
341 33.1
2.9
6.7
13.2
5.6
2.9
5.0
4.4
Quartile 3
337 26.4
3.3
4.5
11.3
3.0
1.5
4.7
1.5
Quartile 4
339 25.1
3.2
2.7
9.1
5.6
2.1
4.7
1.8
Z values for trend test
–
7.86
3.25
3.98
6.14
0.82
4.52
2.07
2.24
P values
–
Figure 61-4). A similar analysis stemming from categorized BMI would only describe an average risk of headache within each
Obesity’s Final Toll
61
. Figure 61-4 Results on Body Mass Index and risk of headache United States women surveyed in NHIS 2003 (Keith et al., 2008), BMI stands for boy mass index
arbitrary classification (say above vs. below the detected threshold of BMI = 20), thus giving little information regarding the patterns of risk over the complete range of BMI.
5.6.5
Accounting for Uncertainty
We have described how complex summary statistics, such as YLL, have been computed to estimate the effects of obesity on duration of life and risk of death. Here we will discuss how to account for uncertainty in these estimates. First, recall that the estimates require input from various data sources in terms of statistics that have their own uncertainty (i.e., standard errors) that affects the overall level of uncertainty about the estimate output. Therefore, it is important to begin analyses by carefully ascertaining correct estimates and standard errors for input. In cases where the input statistics (e.g., hazard ratios that estimate relative risks) are estimated from complex samples (e.g., the NHANES series), specialized software such as SUDAAN (Research Triangle Institute, Research Triangle Park, NC) is required to account for the nonrandom sampling design, apply sampling weights, and provide adjusted parameter estimates and standard errors (Korn and Graubard, 1999). Once input estimates have been computed, one approach to accounting for uncertainty in complex output estimates (e.g., YLL) is by either nonparametric resampling (i.e., bootstrap or jackknife procedures) within the original data or by parametric resampling (i.e., Monte Carlo simulations) (Korn and Graubard, 1999) to generate hypothetical distributions of input statistics that may be used to generate a hypothetical distribution of output estimates from which standard errors and confidence intervals may be obtained. Another more direct and comprehensive approach has been developed by Parmigiani (2002). He describes how to build complex decision models for explicitly differentiating and modeling all sources of uncertainty whether they stem from individual-to-individual variability within a given population or limitations in the level of knowledge of information input applied to the model.
1101
1102
61 5.6.6
Obesity’s Final Toll
Choosing Reference Categories and Cut Points
Whenever one calculates a statistic like YLL for people at a particular body size or degree of fatness, there is an implicit ‘‘counterfactual.’’ The counterfactual is the alternative condition that could, hypothetically, have existed. In the case of the YLL calculations and attributable death calculations with respect to obesity, the counterfactual is some body size or BMI level that a person could hypothetically have had that is different than the one they do have. For example, if we say that a person with a BMI of 40 is estimated to and X YLL, we mean that their life expectancy with a BMI of 40 is X years less than it would be if they had had some other BMI, R. We use R to denote ‘‘reference’’ BMI or BMI category. This invites the question, to which value should we set R? The answer is important because the choice of R can have major effects on the estimated YLL. In our work (Allison et al., 1999c; Fontaine et al., 2003) we have typically used values of R for BMI between 23 and 25. Our rationale has been several-fold. First, such values are relatively near the center of the distribution of BMI suggesting that they are attainable by many people and, therefore, perhaps not too unreasonable goals to aim for. Second, such values are just below the BMI cutoff for overweight (i.e., 25) indicating a value suggested as ‘‘healthy’’ by leading authoritative bodies and yet also not extreme. Third, these are values that are roughly associated with the lowest MR’s across many studies. In similar research projects, others have sometimes used different values of R. One popular choice has been to use a reference category that includes BMIs between 18.5 and 25. We think this is misguided. Our reasoning involves the frequent observation that there is a U or J-shaped relation between BMI and MR. Specifically, at BMI values below 23, MRs tend to increase. Therefore, when obese oroverweight BMIs will appear less deleterious if compared to a reference category of 18.5–25 than,forexample23–25.Now,letusassumethattheelevationinMRatBMIvaluesbelow23are‘‘real,’’ i.e., that low BMIs truly cause increases in MR. Then surely we would not want to implicitly orexplicitly advocate that people strive for have such low BMIs. In that case we would be comparing one unhealthy range of BMI (overweight and obesity) to another unhealthy range (low BMIs). This makes little sense. Alternatively, instead of assuming that the elevation in MR at BMI values below23is‘‘real,’’i.e., thatlowBMIs trulycauseincreases inMR,wemightassumethat thiselevation is simply due to confounding by occult disease. If that is the case, one is assuming that the elevated values are not valid. If one assumes they are not valid, why would one include them in one’s analysis?Forthesereasons,wethinkthatuseofthebroadreferencerangeofBMI18.5–25ismisguided and reference ranges around 23–25 are far more reasonable.
6
Summary and Recommendations
Although obesity associates with myriad health problems and appears to reduce lifespan, there are a host of methodological and statistical issues that preclude the exact calculation of the toll obesity takes, yet reasonable estimates are available. In this chapter we have highlighted the most salient issues, including confounding, reverse causation, secular changes, and the choice of reference categories. However, because there are reasonable arguments that can be made to underlie a variety of decisions concerning these issues, there is little consensus of ‘‘best practices’’ on how to analyze data to estimate the association of obesity to mortality indices. Nonetheless, we believe the following recommendations will move us closer toward developing the best possible estimates of the mortality consequences imposed by obesity.
Obesity’s Final Toll
61
Develop methods for correcting reverse causation, other than excluding smokers and participants with obvious serious illness.
Search for factors other than smoking and serious illness that could cause confounding. Explore the possibility that advanced aging causes reverse-causation. Investigate the effects on the relation between adiposity and MR of correcting for measurement-error in the adiposity predictor and covariates measured with error.
Model the functional relationship between adiposity and MR as accurately as possible. This may call for new statistical methods of calculating YLL based on models of continuous (not categorized) BMI.
Summary Points The prevalence of obesity has tripled in many countries since the 1980s, and developing countries such as Nigeria are also experiencing increased obesity rates.
Obesity is an excess of body fat that can be one of total body lipid (fat) or any anatomical depot of fat or adipose tissue.
Obesity appears to contribute to serious medical conditions such as type-2 diabetes, heart disease, stroke, and many forms of cancer.
That obesity decreases lifespan has been hypothesized, if not fully documented, for millennia. With few exceptions, BMI-mortality studies indicate that the association is U- or J-shaped, though the association can vary as a function of factors such as age, sex, and race.
The extent to which the elevated mortality rate with low BMIs represents causation or merely association remains the subject of debate and inquiry.
Estimates of obesity-attributable deaths have been published for the United States and several other countries.
Obesity appears to reduce life expectancy at the individual and population level. The association of obesity, measured by indices such as circumference measures, skinfolds, and bio-impedance, to indices of mortality has recently become an area of inquiry.
BMI may not capture adequately the effect of adiposity on MR despite its high correlation with adiposity.
Reverse causation, confounding, regression dilution, accounting for uncertainty, and choosing reference categories and cut-points are among the statistical issues that make it challenging to generate definitive estimates of the association of obesity to indices of mortality.
References Adams KF, Schatzkin A, Harris RB, Kiphis V, Mouw T, Ballard-Barash R, Hollenbeck A, Leitzmann M. (2006). N Engl J Med. 355: 763–778. Allender S, Rayner M. (2007). Obes Rev. 8: 467–473. Allison DB, Downey M, Atkinson RL, Billington CJ, Bray GA, Eckel RH, Finkelstein EA, Jensen MD, Tremblay A (2008). Obesity. 16: 1161–1177.
Allison DB, Faith MS, Heo M, Kotler DP. (1997c). Am J Epidemiol. 146: 339–349. Allison DB, Faith MS, Heo M, TownsendButterworth D, Williamson DF. (1999b). Obes Res. 4: 342–354. Allison DB, Fontaine KR, Manson JE, Stevens J, VanItallie TB. (1999c). JAMA 282: 1530–1538.
1103
1104
61
Obesity’s Final Toll
Allison DB, Gallagher D, Heo M, Pi-Sunyer FX, Heymsfield SB. (1997a). Int J Obes Relat Metab Disord. 21: 424–431. Allison DB, Heo M, Flanders DW, Faith MS, Carpenter KM, Williamson DF. (1999a). Ann Epidemiol. 9: 132–142. Allison DB, Heo M, Flanders DW, Faith MS, Williamson DF. (1997b). Am J Epidemiol. 146: 672–680. Banegas JR, Rodriquez-Artalejo F, Graciani A, Villar F, Herruzo R. (2003). Eur J Clin Nutr. 51(Suppl): S18–S21. Banegas JR, Lopez-Garcia E, Guiterrez-Fisac JL, GuallarCastillon P, Rodriquez-Artalejo F. (2003). Eur J Clin Nutr. 57: 201–208. Bigaard J, Frederiksen K, Tjonneland A, Thomsen BL, Overvad K, Heitmann BL, Sorensen TI. (2005). Int J Obes (Lond). 29: 778–784. Bigaard J, Frederiksen K, Tjonneland A, Thomsen BL, Overvad K, Heitmann BL, Overvad K, Heitmann BL, Sorensen TI. (2004). Obes Res. 12: 1042–1049. Bouchard C. (2007). Int J Obes. 31: 1552–1553. Butler RN, Allison DB, Ludwig DS. (2005). N Engl J Med. 352: 1138–1145. Calle EE, Teras LR, Thun MJ. (2005). N Engl J Med. 353: 2197–2199. Carmelli D, Halpern J, Swan GE. (1991). J Clin Epidemiol. 44: 1341–1351. Carroll RJ, Ruppert D, Stefanski LA, Crainiceanu CM. (2006). Measurement Error in Nonlinear Models: A Modern Perspective, 2nd ed. Chapman & Hall/CRC, Boca Raton, FL. Clarke R, Shipley M, Lewington S, Youngman L, Collins R, Marmot M. (1999). Am J Epidemiol. 150: 341–353. Dhurandhar N, Lijoi MC, Grilo CM, DeLuca M, Westfall AO, Allison DB. (2006). Int J Obes. 30: 1585–1594. Durazo-Arvizu R, McGee D, Li Z, Cooper R. (1997). J Am Stat Assoc. 92: 1312–1319. Dyer A, Stamler J. (1990). Lancet. 335: 765–774. Flegal KM, Graubard BI, Williamson DF, Gail MH. (2005). JAMA. 293: 1861–1867. Flegal KM, Graubard BI, Williamson DF, Gail MH. (2007). Am J Epidemiol. 166: 975–982. Fontaine KR, Allison DB. (2003). In: Bray G, Bouchard C (eds.) Handbook of Obesity, 2nd ed. Dekker and Co., New York, pp. 767–785. Fontaine KR, Redden D, Wang C, Westfall A, Allison DB. (2003). JAMA. 289: 187–193. Galton F. (1885). J Antropol Inst 15: 246–263. Gardner J. (1875). Longevity: The means of prolonging life after middle age. Boston, W.F. Gill and Company. Garfinkel L. (1986). Cancer. 58: 1826–1829. Greenberg J. (2005). JAMA. 294: 552. Greenberg JA. (2006). Obesity. 14: 2071–2079. Greenberg JA. (2007). Int J Obes. 25: 1071–1078. Greenberg JA, Fontaine K, Allison DB. (2007). Int J Obes (Lond). 31: 1449–1455.
Gregg EW, Cheng YJ, Cadwell BL, Imperatore G, Williams DE, Flegal KM, Venkat Narayan KM, Williamson DF. (2005). JAMA. 293: 1868–1874. Hu FB, Willett WC, Li T, Stampfer MJ, Colditz GA, Manson JE. (2004). N Engl J Med. 352: 2694–2703. Kahn HS, Williamson DF. (1994). Int J Obes Relat Metab Disord. 10: 686–691. Katzmarzyk PT, Ardern CI. (2004). Can J Public Health. 95: 16–20. Keith SW, Fontaine KR, Greenberg JA, Allison DB. (2007). Obesity. 15(Suppl): A111. Keith SW, Redden DT, Katzmarzyk P, Boggiano MM, Hanlon EC, Benca RM, Ruden D, Pietrobelli AMD, Barger J, Fontaine KR, Wang C, Aronne LJ, Wright S, Baskin M, Dhurandhar N, Lijoi MC, Grilo CM, DeLuca M, Westfall AO, Allison DB. (2006). Int J Obes. 30: 1585–1594. Keith SW, Wang C, Fontaine KR, Cowan CD, Allison DB. (2008). Obesity 16: 377–383. Korn EL, Graubard BI. (1999). Analysis of Health Surveys. Wiley & Sons, New York. Kragelund C, Omland T. (2005). Lancet. 366: 1589–1591. Kuk JL, Katzmarzyk PT, Nichaman MZ, Church TS, Blair SN, Ross R. (2006). Obesity (Silver Spring). 14: 336–341. Landi F, Onder G, Gambassi G, Pedone C, Carbonin P, Bernabei R. (2000). Ann Intern Med. 160: 2641–2644. Lee IM, Manson JE, Hennekens CH, Paffenbarger RS. (1993). JAMA. 270: 2823–2828. Luo W, Morrison H, de Groth M, Waters C, Desmeules M, Jone-Mclean E, Ugnat A-M, Dejardins S, Lin M, Mao Y. (2007). Chronic Dis Can. 27: 135–144. MacMahon S, Peto R, Cutler J, Collins R, Sorlie P, Neaton J, Abbott R, Godwin J, Peto R. (1976). In: Fletcher CM, Peto R, Tinker CM, Speizer FE (eds.) The Natural History of Chronic Bronchitis and Emphysema. Oxford University Press, Oxford, pp. 218–223. Manson JE, Willett WC, Stampfer MJ, Colditz GA, Hunter DJ, Hankinson SE, Hennekens CH, Speizer FE. (1995). N Engl J Med. 333: 677–685. Menotti A, Descovich GC, Lanti M. (1993). Prev Med. 22: 293–303. Molinari N, Daures JP, Durand JF. (2001). Stat Med. 20: 237–247. Mokdad AH, Marks JS, Stroup DF, Geberding JL. (2005). JAMA. 293: 293–294. Mokdad AH, Marks JS, Stroup JL, Gerberding JL. (2004). JAMA. 291: 1238–1245. Ni Mhurchu C, Turley M, Stefanogiannis N, Lawes CM, Rodgers A, Vander Hoorn S, Tobias M. Public Health Nutr. (2005). 8: 402–408. Peeters A, Barendregt JJ, Willekens F, Mackenbach JP, Al Mamun A, Bonneux L. (2003). Ann Intern Med 138: 24–32.
Obesity’s Final Toll Olshansky SJ, Passaro D, Hershow R, Layden J, Carnes BA, Brody J, Hayflick L, Butler RN, Allison DB, Ludwig DS. (2005). N Engl J Med. 352: 1138–1145. Parmigiani G. (2002). Stat Methods Med Res. 11: 513–537. Rhoads GG, Kagan A. (1983). Lancet. 8323: 492–495. Seidel JC, Verschuren WMM, van Leer EM, Kromhout D. (1996). Arch Intern Med. 156: 958–963. Sempos CR, Durazo-Arvizu R, McGee DL, Cooper RS, Prewitt TE. (1998). Ann Epidemiol. 8: 289–300.
61
Simpson JA, MacInnis RJ, Peeters A, Hopper JL, Giles GG, English DR. (2007). Obesity 15: 994–1003. Spearman C. (1904). Am J Psychol. 15: 72–101. Vague J. (1947). Presse Med. 30: 339–340. von Stru¨mpell A (Published 1893). A Text-book of medicine for students and practitioners. (Translated by Herman Frank Vickery and Philip Coombs Knapp). D. Appleton and Co. Willett WC, Hu FB, Colditz GA, Manson. (2005). JAMA. 294: 551.
1105
62 Financial Impact of Obesity L. Barrios . D. B. Jones 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1108
2 2.1 2.2 2.3 2.4 2.5 2.6
Main Text . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1108 The Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1108 Children and Adolescents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1110 Impact of Lifestyle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1110 Social Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1111 Economics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1111 Weight Loss Surgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1112 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1117
#
Springer Science+Business Media LLC 2010 (USA)
1108
62
Financial Impact of Obesity
Abstract: > Obesity is associated with serious medical conditions, psychosocial problems, and major health costs. Medical illnesses associated with obesity include diabetes mellitus, hypertension, coronary artery disease, stroke, cancer, hyperlipidemia, gastroesophageal reflux disease, obstructive sleep apnea, degenerative joint disease, and back pain. Socially, obese individuals may suffer from discrimination at work, lower wages, inability to find a love companion, and overall poor quality of life. As the prevalence of obesity soars, the number of bariatric operations performed in the United States has increased dramatically over the last decade. Weight loss surgeries such as the laparoscopic Roux-en-Y gastric bypass, laparoscopic adjustable > gastric band placement, and > sleeve gastrectomy are being performed more frequently in adults, and are now also being performed in children and adolescents. This article summarizes findings from recent studies discussing the financial implications of these trends. Specifically, we address the high medical costs associated with weight related co-morbid illnesses, and we discuss the indirect costs to employers of the obese workforce. List of Abbreviations: > BMI, Body Mass Index; BPD, Biliopancreatic Diversion; > CEA, Cost-Effectiveness Analyses; CMS, Centers for Medicare and Medicaid Services; LAGB, Laparoscopic Adjustable Gastric Band; > NHANES, National Health and Nutrition Examination Survey; > QALY, Quality Adjusted Life Year; > RYGBP, Roux-en Y Gastric Bypass; SG, Sleeve Gastrectomy
1
Introduction
Obesity is a multifactorial disease of global proportions, and is associated with serious medical conditions, psychosocial problems, and major health costs (Obesity by Numbers). Currently, approximately one-third of Americans and an estimated 1.7 billion individuals worldwide are obese. (Buchwald et al., 2004) Due in large part to this increase in obesity prevalence, the number of bariatric operations performed in the United States has risen dramatically over the last several years. The purpose of this article is to summarize findings from recent studies which elucidate the financial implications of these dual trends. Specifically, this article reviews the high medical costs associated with weight-related co-morbid illnesses and discusses the indirect costs to society of obese employees.
2
Main Text
2.1
The Problem
In adults, obesity is defined by a BMI (Body Mass Index) of 30 kg m 2 or greater. BMIs greater than or equal to 40 kg m 2 or, when accompanied by significant comorbidities (e.g., obstructive sleep apnea, diabetes mellitus, and hypertension) as 35 kg m 2 are considered morbid obesity (> Table 62‐1). Obesity is directly related to a number of serious medical conditions including gastroesophageal reflux disease, weight related arthropathies, depression, and cancer. Morbid obesity has been further associated with early mortality – for example, a 25 year old morbidly obese man will live approximately 12 years fewer than his normal weight counterpart (Buchwald et al., 2004).
Financial Impact of Obesity
62
Over the last 20 years, obesity has steadily increased in the United States; overall obesity is estimated to have doubled from 1986 to 2000, while extreme obesity increased 400% from 1983 to 2000 (Sturm, 2002). According to Buchwald, 65% of the US adult population is overweight or obese, 30.6% are obese, and 5.1% are extremely obese (Buchwald et al., 2004). Results from the 2003–2004 National Health and Nutrition Examination Survey (NHANES), using measured heights and weights, indicate that an estimated 66% of U.S. adults are either overweight or obese, as shown in the graph on > Figure 62‐1 (from CDC. gov-National Health and Nutrition Examination Survey (NHANES)).
. Table 62‐1 Comorbid conditions Diabetes Mellitus Hypertension Coronary Artery Disease Stroke Cancer Hyperlipidemia Gastroesophageal Reflux Disease Obstructive Sleep Apnea Degenerative Joint Disease Back Pain Medical illnesses associated with obesity
. Figure 62‐1 Prevalence of adult overweight and obese has been increasing over the last 20 years
1109
1110
62 2.2
Financial Impact of Obesity
Children and Adolescents
Alarmingly, the prevalence of obese children and adolescents in the US has also increased significantly, rising by as much as 50–60% in a single generation (Ogden et al., 2006). Currently, approximately 2 million children suffer from extreme obesity (Collins et al., 2007). Data from NHANES in > Table 62‐2 depicts the national increase in overweight . Table 62‐2 Increasing prevalence of obesity in children and adolescents There has been an increase in the prevalence of obesity in children and adolescents of all age groups over the last 30 years
children and adolescents from 1971 to 2004, where overweight is defined as those children whose BMI > 95 percentile for age. All age groups demonstrate clearly significant increases in percentage of overweight children; overweight adolescents (ages 12–19) in particular rose from 6.1% in 1974 to 17.4% overweight in 2004. Although all ethnicities showed increases in adolescent/child overweight and obesity, non-Hispanic white males, and non-Hispanic black females show the greatest prevalence in overweight in 2004, at 19.1 and 25.4% respectively.
2.3
Impact of Lifestyle
Environmental changes have contributed greatly to the current obesity problem. For example, over the last 30 years, the price of food has dropped by 12–14%, and more people than ever have access to fast food establishments, and lifestyles have become more sedentary (Lakdawalla, 2006). The cost of unhealthy food has decreased the most, while healthy meals have become relatively more expensive. The bottom line is that sweets and fats both dropped in price, 45 and 35% respectively, while fish, fruit, vegetables and dairy products have increased in price by more than 50% (Herper, 2006; Lakdawalla, 2006). Undoubtedly, today’s sedentary lifestyle has also contributed to the development of an obese nation. The level of physical activity has diminished; more people drive longer distances to work, and fewer people are exercising. A significant portion of the population sits in front of a computer all day, and comes home to relax and watch television. The media, with increasing advertisements for unhealthy foods, promotes the consumption of such products. In the UK, approximately £522 million was spent in 2003 for food, soft drinks, and restaurant
Financial Impact of Obesity
62
advertisement on television (Boyce, 2007). Children spend more time playing video games as opposed to playing sports, and their food selection is influenced by what is observed on television (Boyce, 2007). Several studies have shown that children who spend more than 1 h per week watching TV, are at increased risk of becoming obese, with an increased prevalence by up to 2% for every additional hour viewed (Kaiser, 2004).
2.4
Social Issues
The socioeconomic consequences of morbid obesity are difficult to assess. Several studies have shown that obese adolescents tend to have higher high school drop out rates, less marriage, and higher rates of household poverty compared to their non-overweight peers. These social issues not only permeate the lives of obese children and adolescents, but also will burden them throughout their entire lives. Some of these issues include discrimination at work, which may explain lower wages, inability to find a love companion, and overall poor quality of life. The Medical Outcomes Study Short-Form Health Survey evaluated aspects such as physical functioning, social functioning, mental health, and pain, and found obese adults to score poorly in all these areas (Martin et al., 1998). Several studies have also shown a causal relationship between poverty and obesity. In the ‘‘food choice constraint model,’’ the ability to purchase healthy foods is limited to those who are impoverished because healthier choices are generally more expensive. Their budget will only allow them to purchase the generally cheaper high-calorie foods (Martin, 2005). Therefore, treating obesity has become not only a major health care priority, but also an important social concern.
2.5
Economics
Medical costs alone are significantly higher for obese individuals, especially those with associated chronic medical conditions, such as type 2 diabetes, cardiovascular disease, fatty liver, or other metabolic disorders. Indeed, obese and morbidly obese patients are reported to have 14–38% more physician visits, 48% more inpatient days and 1.8 times the number of annual pharmacy dispenses – particularly for diabetic and cardiovascular medications (Finkelstein et al., 2005b; Thompson et al., 1998). A Healthcare for Communities survey reports that financially, this translates into higher annual medical (36%) and medication (77%) costs for obese patients compared to normal weight controls (Sturm, 2002). Similarly, Finkelstein et al.’s study used data from the 1998 Medical Expenditure Panel Survey (MEPS) and Behavioral Risk Factor Surveillance System (BRFSS) survey and found an increase in annual medical costs in the order of 37% ($1,486) for Medicare patients and 39% ($864) for Medicaid patients (Finkelstein et al., 2004). In sum, these figures represent roughly 5–7% of the total US annual health care expenditures – or $75–93 billion per year, $17 billion of which is financed by Medicare and $21 billion by Medicaid (Finkelstein et al., 2004; Gates et al., 2006; Herper, 2006). Intriguingly, obese individuals also experience significantly more non-medical expenditures compared to normal weight controls, which is most apparent in the form of workplace absenteeism. Obese employees are reported to be 1.7 times more likely to have seven or more absences due to illness during a 6-month period than their leaner counterparts (Finkelstein et al., 2005b; Thompson et al., 1998; Tucker and Friedman, 1998). Finkelstein found that an obese male employee costs approximately $670 more annually, and an obese female cost an average of $1,200 more annually than normal weight workers (Finkelstein et al., 2005a, b). On a grander
1111
1112
62
Financial Impact of Obesity
scale, Thomson et al. report that such obesity-attributable absenteeism cost employers roughly $2.4 billion in 1998; others estimate this figure to be up to $4 billion annual dollars (Finkelstein et al., 2005b; Herper, 2006; Thompson et al., 1998). At this point in time, obesity-related health care costs exceed spending for smoking and problem drinking (Sturm, 2002). Unforeseen weight-related costs have been identified in fuel requirements for transport: airplanes are now required to use an extra 350 million gallons of fuel each year for the added weight of obese and overweight flyers, costing approximately $275 million, according researchers at Cornell University (Herper, 2006). By the same token, an obese person will spend an extra 5 cents on gasoline per year compared to a normal weight individual (Herper, 2006). Part and parcel, perhaps, of weight-related comorbidities, a large portion of disability claims now are secondary to obesity related conditions. One of the largest US disability carriers, Unum Provident, reported a ten-fold increase in the incidence of obesity-related disability claims; according their databases, the average annual health care cost for a disabled, obese individual is $51,023 ($30,567 medical + $8,720 disability payments + $11, 736 morbid medical costs) (> Bariatric Surgery Policy Guidance). Other studies have shown similar results: Arena and colleagues evaluated the association between BMI and short-term disability amongst 19,061 employees of a large financial services corporation. Their group found that BMI is an independent predictor for short term disability events, resulting in loss of productivity and increased costs to the companies (Arena et al., 2006). The overall message is that with enhancement of employee health, there should be fewer disability claims and a decrease in the costs incurred by employers. It is no surprise that obese individuals also have more injuries at work and increased claims to workers’ compensation. In evaluating the relationship between BMI and the number and type of workers’ compensation claims in 11,728 Duke University employees, Ostbye et al. reported a link between higher BMI and greater number of claims, mostly related to lower extremity, wrist, hand, and back injuries from falls or slips, lifting, and exertion. This study also identified a linear relationship with BMI and lost workdays, medical claims costs ($51,091 vs. $7,503), and indemnity claims costs ($59,178 vs. $5,396) (Ostbye et al., 2007). A similar investigation in a US aluminum manufacturing company noted that, of 7,690 employees, at least 85% of injured workers were considered overweight or obese. Occupational injuries to the leg or knee were the most common among workers with higher BMI’s (Pollack et al., 2007). Given the strong association between worker’s injuries and BMI, employers should consider it a priority to provide alternatives for employees to attain a healthy weight. Several studies have examined the impact of obesity on occupational choice and wages and observed that a higher incidence of obese individuals, particularly white obese women, have relatively low paying occupations, are less likely to obtain managerial or professional positions, and have higher rates of poverty when compared to same age normal weight females (Averett and Korenman, 1999; Cawley, 2000; Finkelstein et al., 2005b; Pagan JA, 1997). Cawley et al. noted that an increase in weight of two standard deviations (roughly 65 pounds) is associated with a 7% decrease in wages of white women (Cawley, 2000). Clearly, the economic ramifications of obesity are numerous, and have permeated all layers of society.
2.6
Weight Loss Surgery
According to the 1991 National Institute of Health Consensus Statement, indications for bariatric surgery include those patients who have a BMI greater than 40, or greater than 35
Financial Impact of Obesity
62
with serious comorbid conditions (NIH, 1991). For such morbidly obese individuals, bariatric surgery is a very valuable treatment option, providing not only an effective weight loss therapy, but frequently leading to an amelioration of many comorbid conditions, reducing mortality, and ultimately decreasing health care costs. In a 2004 meta-analyses of 136 studies, totaling 22,094 patients, Buchwald et al. showed that weight loss surgery resulted in the resolution of diabetes in 76% of subjects; hypertension was eliminated in 61.7%; obstructive sleep apnea in 85.7%; and high cholesterol levels decreased in more than 70% of patients who underwent bariatric surgery (Buchwald et al., 2004). Similarly, Christou et al. reported an absolute mortality reduction of 5.49% (P < .001) when comparing 1,035 patients who underwent weight loss surgery with 5,746 controls in a 5 year follow-up (Christou et al., 2004). Oster et al. found that a sustained weight loss of around 10% initial body weight reduced hypertension, hypercholesterolemia, and type 2 diabetes, decreased the expected lifetime incidence of heart disease and stroke, increased life expectancy by 2–7 months, and reduced expected lifetime medical care costs by $2,200–5,300 (Oster et al., 1999). Paralleling adult trends, bariatric procedures performed in obese children and adolescents have become increasingly frequent in an effort to decrease associated health problems and
. Figure 62-2 Roux-en-Y Gastric Bypass (RYGB) involves the creation of a small ( duodenal switch (BPD) are permitted by Centers for Medicare and Medicaid Services (CMS). > Figure 62‐2–62‐4 However, the mean cost per operation for all payers has increased 21% from $12,872 in 1998 to $15,533 in 2003 (HCUP). It is interesting to note that 80–90% of all bariatric cases performed are the gastric bypass, and that, while this procedure can be performed either open and laparoscopically, the laparoscopic approach
. Figure 62-4 Biliopancreatic Diversion with Duodenal Switch (BPD) creates malabsorption by maintaining a flow of bile and pancreatic juice through the biliopancreatic limb. The common-channel length of small intestine in the duodenal switch severely limits caloric absorption. The extent of malabsorption is thought to be a function of the length of the common channel. Illustrations reprinted with permission from Atlas of Metabolic and Weight Loss Surgery, Jones et al., Cine-Med, 2008
1115
1116
62
Financial Impact of Obesity
seems to be more cost effective – the average total cost of laparoscopic RYGB falls in the range of $17,660, in comparison to the higher cost of the open gastric bypass, which averages $20,443 (Paxton and Matthews, 2005). > Figure 62‐2 Other studies report similar results, where the laparoscopic adjustable gastric band appears to be the more expensive ($25,355), and the laparoscopic gastric bypass the most economical ($19,794) – even when compared to the open approach ($22,313) (Livingston, 2005). A relatively newer procedure, the sleeve gastrectomy, fairs between the laparoscopic adjustable gastric band and the laparoscopic gastric bypass in terms of cost. This procedure is performed by removing a large portion of the greater curve of the stomach. > Figure 62‐5 Clinical pathway development has also been shown to decrease costs to providers involved in weight loss surgery. . Figure 62-5 Sleeve gastrectomy (SG) SG consists of the restrictive component of the duodenal switch (DS), a vertical resection of the greater curvature of the stomach creating a long tubular stomach along the lesser curvature. The pylorus and part of the antrum are preserved. Illustrations reprinted with permission from Atlas of Metabolic and Weight Loss Surgery, Jones et al., Cine-Med, 2008
Several researchers have studied the cost-effectiveness of weight loss surgery in the Unites States and other countries. In the US, Cost-Effectiveness Analyses (CEA) have shown that gastric bypass has provided net savings in the order of $35,000 per QALY (Quality Adjusted Life Year), and appears to be more cost effective for women than for men, for individuals with a BMI greater than 40, and for younger individuals (Craig and Tseng, 2002). Similarly, Clegg et al.’s study out of the United Kingdom reported that in comparison with nonsurgical
Financial Impact of Obesity
62
management of obesity, weight loss operations were indeed more cost effective at £11,000 per QALY (Clegg et al., 2003). Snow and colleagues reported a savings of approximately $240,566.04 dollars per year on medications by patients undergoing gastric bypass (Snow et al., 2004). Improvement or elimination of comorbid conditions, decreased requirement for medications of chronic disease, such as diabetes and hypertension, and increase in productivity, are all considered in the cost-effectiveness analysis of weight loss surgery. In summary, obesity is a major health, social, and financial issue permeating all layers of society in the US and abroad. This disease is not exclusive to adults, and has similar, if not worse implications for children and adolescents. Weight loss surgery is a viable and effective alternative, and has been shown to improve quality of life, decrease morbidity and mortality, and decrease overall health costs to individuals and their employers. As a nation, we must increase awareness of this major health and socio-economic problem, and promulgate the extensive benefits of weight loss surgery.
Summary Points
Prevalence of obesity is increasing in adults and children. Weight loss surgery is a viable alternative for obese individuals. Medical and non-medical costs in the obese population are growing. Weight loss surgery is associated with a decrease or eradication of medical problems, increased life expectancy, and improved morale. Weight loss will ultimately result in decreased costs to the individual and society.
Acknowledgments We would like to thank Ms. Shannon Fischer for her editorial assistance with this manuscript.
References Arena VC, Padiyar KR, Burton WN, Schwerha JJ. (2006). J Occup Environ Med. 48: 1118–1124. Averett S, Korenman S. (1999). Int J Obes Relat Metab Disord. 23: 166–173. Bariatric Surgery Policy Guidance. Betsy Lehman Center for Patient Safety and Medical Error Reduction. Commonwealth of Massachusetts, Expert Panel on Weight Loss Surgery. Boyce T. (2007). Obes Rev. 8(Suppl 1): 201–205. Buchwald H, Avidor Y, Braunwald E, Jensen MD, Pories W, Fahrbach K, Schoelles K. (2004). JAMA. 292: 1724–1737. Cawley J. (2000). Body Weight and Women’s Labor Market Outcomes. NBER Working Paper Paper no. 7841. National Bureau of Economic Research, Cambridge, U.S.
Christou NV, Sampalis JS, Liberman M, Look D, Auger S, Mclean AP, Maclean LD. (2004). Ann Surg. 240: 416–423; discussion 423–424. Clegg A, Colquitt J, Sidhu M, Royle P, Walker A. (2003). Int J Obes Relat Metab Disord. 27: 1167–1177. Collins J, Mattar S, Qureshi F, Warman J, Ramanathan R, Schauer P, Eid G. (2007). Surg Obes Relat Dis. 3: 147–152. Commonwealth of Massachusetts Betsy Lehman Center for Patient Safety and Medical Error Reduction Expert Panel on Weight Loss Surgery: executive report. Conference on 12/12/07. Craig BM, Tseng DS. (2002). Am J Med. 113: 491–498. Finkelstein E, Fiebelkorn C, Wang G. (2005a). Am J Health Promot. 20: 45–51.
1117
1118
62
Financial Impact of Obesity
Finkelstein EA, Fiebelkorn IC, Wang G. (2004). Obes Res. 12: 18–24. Finkelstein EA, Ruhm CJ, Kosa KM. (2005b). Annu Rev Public Health. 26: 239–257. Gates D, Brehm B, Hutton S, Singler M, Poeppelman A. (2006). Aaohn J. 54: 515–520. HCUP Cost and utilization project (HCUP). Agency for Healthcare Research and Quality. Rockville, MD. http://www.ahrq.gov/data/hcup/ Herper M. (2006). The Hidden Cost of Obesity. Available at: Forbes.Com. Inge TH, Xanthakos SA, Zeller MH. (2007). Int J Obes (Lond). 31: 1–14. Kaiser. (2004). The Role of Media in Childhood Obesity. The Henry J. Kaiser Family Foundation, Washington, D.C. Lakdawalla D. (2006). Cheap Food, Societal Norms And the Economics of Obesity. Wall St J. http://online. wsj.com/public/article/SB115634907472843442_xrNV2M1Pwf8pAcQYUWEBITP1LQ_20060901. html Livingston EH. (2005). Am J Surg. 190: 816–820. Martin LF, White S, Lindstrom W Jr. (1998). World J Surg. 22: 1008–1017. Martin SS. (2005). From Poverty to Obesity: Exploration of the Food Choice Constraint Model and the Impact of an Energy-Dense Food Tax. http://www. allbusiness.com/accounting/1086324-1.html National Health and Nutrition Examination Survey (NHANES). NHANES can be accessed at: www. cdc.gov NIH. (1991). Gastrointestinal surgery for severe obesity. Proceedings of a National Institutes of Health Consensus Development Conference. Bethesda, MD.
Obesity by Numbers. Obesity in America. http://www. obesityinamerica.org/ Ogden CL, Carroll MD, Curtin LR, Mcdowell MA, Tabak CJ, Flegal KM. (2006). JAMA. 295: 1549–1555. Ostbye T, Dement JM, Krause KM. (2007). Arch Intern Med. 167: 766–773. Oster G, Thompson D, Edelsberg J, Bird AP, Colditz GA. (1999). Am J Public Health. 89: 1536–1542. Pagan JA, Davila A. (1997). Social Science Quarterly. 78: 756–770. Paxton JH, Matthews JB. (2005). Obes Surg. 15: 24–34. Pollack KM, Sorock GS, Slade MD, Cantley L, Sircar K, Taiwo O, Cullen MR. (2007). Am J Epidemiol. 166: 204–211. Santry HP, Gillen DL, Lauderdale DS. (2005). JAMA. 294: 1909–1917. Snow LL, Weinstein LS, Hannon JK, Lane DR, Ringold FG, Hansen PA, Pointer MD. (2004). Obes Surg. 14: 1031–1035. Sturm R. (2002). Health Aff (Millwood). 21: 245–253. Sugerman HJ, Sugerman EL, De Maria EJ, Kellum JM, Kennedy C, Mowery Y, Wolfe LG. J Gastrointest Surg. 2003 Jan; 7(1): 102–107; discussion 107–108. Thompson D, Edelsberg J, Kinsey KL, Oster G. (1998). Am J Health Promot. 13: 120–127. Tsai WS, Inge TH, Burd RS. (2007). Arch Pediatr Adolesc Med. 161: 217–221. Tucker LA, Friedman GM. (1998). Am J Health Promot. 12: 202–207.
63 Burden of Disease Attributable to Obesity and Overweight: Korean Focus Seok-Jun Yoon . Jae-Hyun Park 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1120
2 Measuring Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1123 2.1 Disease Selection Related to Overweight and Obesity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1123 3 Relative Risk (RR) Assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1123 3.1 The Burden of Disease Attributable to Overweight and Obesity . . . . . . . . . . . . . . . . . . 1126 3.2 Disease Burdens Attributable to Overweight and Obesity . . . . . . . . . . . . . . . . . . . . . . . . . . 1127 4
Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1129 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1133
#
Springer Science+Business Media LLC 2010 (USA)
1120
63
Burden of Disease Attributable to Obesity and Overweight: Korean Focus
Abstract: This chapter estimates the burden of disease attributable to > overweight and > obesity in Korea using disability adjusted life-years (DALYs). First, overweight- and obesity-related diseases and their > relative risks (RRs) were selected by systematic review. Second, population-attributable fractions (PAFs) were computed by using a formula that included RR and the prevalence of exposure (Pe) of overweight and obesity. Third, the DALYs of overweight and obesity-related diseases in Korea were estimated. Finally, the attributable burden (AB) of disease due to overweight and obesity was calculated as the sum of the products resulting from the multiplication of the DALYs of overweight and obesity-related diseases by their PAFs. In 2002, the disease burden attributable to overweight was 827.1 PYs overall: 732.6 for men and 922.9 for women per 100,000 persons. The disease burden attributable to obesity was 260.0 PYs overall: 144.2 for men and 377.3 for women. The burden of diabetes attributable to overweight and obesity was higher than that of any other disease in both genders. The disease burden attributable to overweight was 3.2 times higher than that attributable to obesity. In conclusion, the greatest disease burden attributable to a high BMI occurred among those with only moderately elevated BMI scores, such as overweight, and not those with extremely elevated BMI scores, such as obesity. This suggests that population-based, public health intervention is more effective than high-risk group-focused strategies at reducing the burden of disease attributable to overweight and obesity in Korea. List of Abbreviations: AB, attributable burden; AICR, American Institute For Cancer Research; BMI, > body mass index; DALY, disability adjusted life-year; GBD, > global burden of disease; PAFs, population attributable fractions; Pe, prevalence of exposure; PYs, person-years; RR, relative risk; WHO, World Health Organization; YLL, years of life lost due to premature death; YLD, > years lived with disability
1
Introduction
Overweight and obesity are major > risk factors for > cardiovascular diseases, > stroke, diabetes, hypertension, hyperlipidemia, musculoskeletal diseases and cancer, and they are serious public health threats (British Nutrition Foundation Task Force, 1999; National Task Force on the Prevention and Treatment of Obesity, 2000). The increased prevalence of overweight adds a socioeconomic burden to the public health system. In Korea, obesityrelated medical expenses accounted for 0.91% to 1.88% of the total national health expenditure in 1998, and this socioeconomic burden is likely to increase in the future (Jeong et al., 2002). The socioeconomic burden associated with overweight and obesity is higher globally and the 25% ($1.7 billion) of global population has been suffered from obesity and overweight (WHO, 2000). In Korea, the socioeconomic burden related to obesity and overweight also increases from 11.7 billion won in 1998 to 11.8 billion won in 2005. In a 1997 report reflecting the opinions of health care experts from 25 countries, WHO predicted that obesity would emerge as a great health risk, as was the case with smoking, in the twenty-first century, even though it has been largely ignored as a public health concern, despite its effect on a number of health care problems. In Korea, the prevalence of overweight increased from 25.0% in 1998 to 32.2% in 2001 and to 35.1% in 2005 in the male population and from 27.0% in 1998 to 27.9% in 2001 and to 28.0% in 2005 in the female population during the same period (> Figure 63-1). The prevalence of obesity also increased from 1.7% in 1998 to 3.96% in 2001 in the male population and to 3.4% from 3.0% in the female
Burden of Disease Attributable to Obesity and Overweight: Korean Focus
63
. Figure 63-1 Age-standardized prevalence of overweight in Korea (1998–2005). Source: Ministry of Health and Welfare (2007). National Health and Nutrition Survey, Korea
population during the same period (Ministry of Health and Welfare, 2001, 2007). Notably, the prevalence of overweight in the male population increased in all age groups over a 7-year period. In the female population, the prevalence of overweight only increased in the age group older than 60 during the same period (> Figure 63-2). The burden of disease attributable to overweight and obesity is expected to be a significant public health concern in coming years. The burden of disease attributable to obesity can be prevented through public health policy. For this reason, in its 2002 report, WHO introduced the use of the Disability-Adjusted Life Year (DALY) as the unit of measurement for the Global Burden of Disease (GBD) attributable to 26 risk factors, including obesity (Murray and Lopez, 1996; Murray et al., 2002). However, the WHO report only presented the burden of disease attributable to obesity for six sub-regions, and not for specific countries, such as Korea. Furthermore, the report discussed uncertainties in estimation of disease incidence, duration, and disability weighting. Uncertainty in this risk assessment is by far dominated by the absence or limitations of direct studies on exposure, hazard, and background disease burden. The extrapolation of hazard from the limited number of studies in other populations, which was the case in the WHO report, is another source of potential error (Rodgers et al., 2004). Therefore, as the public health impact of obesity increases, accurate assessment of the burden of disease attributable to overweight and obesity is an essential step in establishing public health interventions designed to prevent obesity-induced diseases in Korea. With the above background, this study aimed to estimate the burden of disease attributable to overweight and obesity by gender and age using the DALY in Korea and to provide a basis for developing the public health strategies needed for the prevention of overweight and obesity. Investigators usually use the Disability Adjusted Life-Year (DALY) developed by World Health Organization (WHO) researchers to quantify the burden of disease. The DALY is the sum of the life years lost due to disability and premature death. DALY-based estimation of disease burden is useful for measuring the general state of health in a population, comparing the state of health and public health systems between countries, setting national priorities in the allocation of limited medical care resources and evaluating the cost-effectiveness of public health interventions (Murray and Lopez, 1996).
1121
1122
. Figure 63-2 Age-standardized prevalence of overweight by sex and age group in Korea (1998–2005). Source: Ministry of Health and Welfare (2007). National Health and Nutrition Survey, Korea
63 Burden of Disease Attributable to Obesity and Overweight: Korean Focus
Burden of Disease Attributable to Obesity and Overweight: Korean Focus
2
Measuring Methods
2.1
Disease Selection Related to Overweight and Obesity
63
According to the WHO criteria for overweight and obesity, overweight was defined as a BMI between 25 and 29.9 kg/m2 and obesity was defined as a BMI of 30 kg/m2 or above in the present study (National Institutes of Health Clinical, 1998). Overweight- and obesity-related diseases were selected through a systematic review using MEDLINE, the Korean National Assembly Library, and the Research Information Center for Health to screen for studies with the following key words: obesity, body weight, overweight, fat, adiposity and body mass index (BMI). In addition, a manual search was conducted by referring to the citations identified through the recent reviews of obesity-induced diseases (Bray, 2003a, b; National Institute of Health, 2000; Reilly et al., 2003; Visscher and Seidell 2001). The criteria used for the selection of research evidence were as follows: it had to have (1) been published between 1970 and 2004 (2) suggested obesity-induced diseases, (3) suggested relative risk (RR) values of obesity-induced diseases, (4) used BMI to define obesity, and (5) met the 2b level of evidence in the Oxford Center for Evidence Based Medicine Levels of Evidence (CEBM, 2003) (a cohort study and a systematic review using cohort). We used the research results of the American Institute for Cancer Research (AICR) (1997) to identify obesity-induced cancers (American Institute for Cancer Research, 1997). The AICR study classified the association between the incidence of cancer and overweight or obesity into the following four categories based on systematic reviews and the opinions of investigators: convincing, probable, possible and insufficient. The study further stressed that the impact of overweight and obesity on those captured under the latter two categories was not clear due to a lack of evidence or divided opinions among investigators. The study included research evidence related to endometrial, breast and renal cancers, as these cancers are included under the ‘‘convincing’’ and ‘‘probable’’ categories. Finally, recent studies have shown that Asian populations have a higher risk of metabolic syndromes, such as type 2 diabetes and cardiovascular disease (CVD), than Caucasians (Deurenberg et al., 2000, 2001, 2002, 2003; Dudeja et al., 2001; Norgan, 1994; Tai et al., 1999). Therefore, for diabetes, > ischemic heart disease (IHD), and stroke, we selected studies performed in Asian populations (Jee et al., 2005; Oh et al., 2004; Zhou, 2002). More specifically, studies performed in a Korean population were selected for diabetes and ischemic heart disease (Jee et al., 2005; Oh et al., 2004). For stroke, we selected studies performed in a Chinese population (Zhou, 2002) because we could not find a study that suggested the relative risk values of obesity-induced stroke for a Korean population. A total of 63 studies met the selection criteria, and 14 diseases related to overweight and obesity were identified (> Table 63-1) (Cedergren, 2004; Felson et al., 1997; Furberg and Thune, 2003; He et al., 2001; Jee et al., 2005; Kato et al., 1992; Layde et al., 1982; Marshall, 2002; O’Brien et al., 2003; Oh et al., 2004; Van den Brndt et al., 2000; Weintraub et al., 2002; Zhou, 2002).
3
Relative Risk (RR) Assessment
The study used the RR ratio suggested in previous studies with the highest possible degree of internal and external validity for each of the 14 diseases related to overweight and obesity.
1123
1124
63
Burden of Disease Attributable to Obesity and Overweight: Korean Focus
. Table 63-1 Overweight- and obesity-related diseases and their relative risks (RRs) Disease [NO scalea]
ICD-10
Ischemic heart disease (Jee et al., I20–I25 2005)
Males
Females
2.39 (BMI 25–25.9)
2.39 (BMI 25–25.9)
2.61 (BMI 26–26.9)
2.61 (BMI 26–26.9)
3.26 (BMI 27–27.9)
3.26 (BMI 27–27.9)
3.29 (BMI 28–28.9)
3.29 (BMI 28–28.9)
3.54 (BMI 29–29.9)
3.54 (BMI 29–29.9)
4.37 (BMI 30)
4.37 (BMI 30)
1.061 (BMI per 2-unit difference)
1.061 (BMI per 2-unit difference)
Congestive heart failure (He et al., I50.0 2001)
1.23 (BMI 27.8)
1.34 (BMI 27.8)
Diabetes mellitus (Oh et al., 2004) E10–E14
2.74 (BMI 25– Figure 65-1). More specifically Hotz and Brown (2004) estimated that one-quarter of the Indian population were at-risk of inadequate zinc intakes, putting India in the ‘‘high’’ risk category for zinc deficiency (> Table 65-1). Hence, zinc deficiency in India clearly warrants further analyses, from both the biological and economic perspectives.
1.3
The Need to Measure the Burden of Zinc Deficiency Correctly
Knowing how many people are at risk of zinc deficiency is a first and important step to ascertain the extent of the problem. However, a head-count approach can neither show the depth or severity of the problem – as encountered with head-count measurements in the poverty literature (cf. Sen, 1976) – nor does it allow for comparisons to be made between the severity of ill health from zinc deficiency and other public health problems. Likewise, summing up the magnitude of various health problems to obtain an overall burden of disease is not possible with a head-count approach alone.
In assessing the impact of a public health program – or in continuous monitoring – it may be misleading to simply look at the change in the prevalence of zinc deficiency: the intervention may reduce the severity of the ill-health outcomes of the deficiency although the absolute number of cases may remain the same. In establishing the cost-effectiveness of a public health program, using the effect of zinc deficiency on mortality rates (e.g., to express the cost per death averted) may also be misleading: a disease or deficiency may result in many cases with severe but non-lethal health outcomes, while another may result in fewer but more severe health outcomes or even deaths. In the first case the program may improve the lives of many people considerably (e.g., reduce morbidity) but not save any lives, while in the latter case the program may prevent a few deaths but not affect the severity of the morbidity. Finally, to carry out purely economic analyses, e.g., by assessing the productivity loss due to a deficiency, merely knowing the number of people who are zinc deficient does not say anything about the extent to which their productivity may be affected. Hence, simply looking at the prevalence or > incidence of zinc deficiency or at mortality rates is not very informative. The forgoing has shown that zinc deficiency is a global public health problem, particularly in India. It has explained the need for a better economic indicator that captures both the extent and severity of the problem. The rest of this chapter explains the concept of disability-adjusted life years (DALYs) and puts forward a framework for its application to zinc deficiency in India.
2
Zinc Deficiency and DALYs
2.1
The DALY Methodology
2.1.1
Application and Strengths of the Methodology
As explained above, head-count approaches to measure ill-health are not satisfactory because they do not capture the severity of a condition. One concept that addresses this issue is the
. Figure 65-1 Zinc deficiency worldwide. This figure shows for different regions in the world the percentage of people having insufficient zinc intakes (by US standards); India belongs to the SEAR-D region where this share is highest. Region code: AFR African Region; AMR Region of the Americas; EMR Eastern Mediterranean Region; EUR European Region; SEAR South-East Asia Region; WPR Western Pacific Region. Mortality strata: A, very low child, very low adult; B, low child, low adult; C, low child, high adult; D, high child, high adult; E, high child, very high adult. RDIs recommended dietary intakes. Source: After WHO (2002)
Zinc Deficiency and DALYs in India: Impact Assessment and Economic Analyses
65 1155
1156
65
Zinc Deficiency and DALYs in India: Impact Assessment and Economic Analyses
. Table 65-1 Zinc deficiency in India and its neighbors Country
Total population (millions)
India
At risk of ZnD (%)
At risk of ZnD (millions)
982
25.9
254
1,263
14.1
178
Pakistan
148
11.1
16
Bangladesh
China
125
50.4
63
Myanmar
44
34.6
15
Nepal
23
21.3
5
Sri Lanka
18
44.7
8
2,604
20.8
540
Total
This table shows the total population size of India and its neighbors and the percent of the population of each of these countries that is at risk of zinc deficiency, as well as the absolute number of people in this region who are at risk of zinc deficiency (by the standards of the International Zinc Nutrition Consultative Group). ZnD zinc deficiency. Source: Hotz and Brown (2004)
disability-adjusted life years (DALYs) approach that builds on the idea of quality-adjusted life years (QALYs). The World Bank, in collaboration with the WHO, developed the DALYs methodology and introduced DALYs as a measure for the global burden of disease (GBD) (World Bank, 1993). The use of DALYs was further popularized following the work of Murray and Lopez (1996); nowadays, the approach is also often used by other international organizations (e.g., FAO, 2004; UN-SCN, 2004) and in a diverse set of contexts, e.g., civil war and sanitation infrastructures (Collier and Hoeffler, 2004; Rijsberman, 2004). DALYs are used particularly in global or developing country settings (c.f. Fox-Rushby, 2002). In the context of dietary deficiencies in India, DALYs were used to determine the cost-effectiveness of > biofortification (Stein et al., 2006, 2007, 2008) and the human and economic cost of micronutrient malnutrition (Stein and Qaim, 2007). The strength of the DALY methodology is that it includes both the severity – what economists may refer to as its depth – and the incidence – or width – of a condition. This is done through the use of disability weights that define the extent of the loss of an individual’s function with each adverse health outcome relative to death (100% loss) and complete health (no loss). This relative weighting means that the combined health losses due to morbidity and mortality can be measured in one single index, and the latter is comparable across different conditions. The duration of the adverse health outcomes are counted in life years. With this, the burden of disease, or of a particular dietary deficiency, can be expressed as the aggregate number of DALYs that are lost due to the condition(s) of interest, i.e., as the sum of the years of life lost (YLL) due to the condition-related mortality and the sum of years lived with disability (YLD) that is due to the condition (Murray and Lopez, 1996).
2.1.2
The Calculation of DALYs
Because the severity – or the level of disability – is weighted in the YLD (relative to the YLL), the basic rationale underlying the DALY concept can be represented as: Burdenof disease ¼ DALYslost ¼ YLL þ YLDweighted
Zinc Deficiency and DALYs in India: Impact Assessment and Economic Analyses
65
This burden of disease needs to be calculated and aggregated for each target group (i.e., each relevant age and gender group) that is affected by the condition – as incidence and severity may differ between different groups – and future health losses need to be discounted. Following Zimmerman and Qaim (2004) and Stein et al. (2005), the revised formula can then be represented more formally as: X X X 1 erLj 1 erdij þ T M T I D DALYslost ¼ j j j ij ij j i j r r Where Tj is the total number of people in target group j, Mj is the mortality rate associated with the condition in target group j, Lj is the average remaining life expectancy for target group j (i.e., if persons in that target group live 1 year shorter they lose one DALY), Iij is the incidence rate of condition i in target group j, Dij is the disability weight for condition i, in target group j (ranging from 1 to 0, with 1 representing a complete loss of functioning and 0 perfect health), dij is the duration of the condition i in target group j (for permanent conditions dij equals the average remaining life expectancy Lj), and r is the discount rate for future life years.
2.2
General Criticism of DALYs and Methodological Differences
The DALY approach has not been without criticism (e.g., Allotey et al., 2003; Anand and Hanson, 1998; Arnesen and Nord, 1999; Groce et al., 1999; Lyttkens, 2003; Olsen et al., 2002; Richardson, 1999). In this section the main points of critique are discussed and the modifications that have been made to the original methodology to address some of the issues are explained. A more detailed discussion can be found in Murray and Lopez (1996) and FoxRushby (2002).
2.2.1
Discounting of DALYs and Future Life and Health
> Discounting DALYs (and thus future life and health) is a contentious issue in the literature. The main reproach to discounting DALYs is that it is unfair toward future generations because, with discounting, saving one DALY next year is worth less than saving one DALY today and saving one DALY in 20 years is – at present – worth even less (> Figure 65-2). This issue has been discussed by Murray and Lopez (1996) and Tan-Torres Edejer et al. (2003, p. 67) who explained succinctly: "
The logic for discounting costs is that the value of a unit of consumption to individuals and society decreases over time, for three possible reasons. First, individuals take into account the fact that they might not be alive to benefit from future consumption, and society takes into account the possibility of catastrophe – the possibility that any or all interventions might at some point in the future become valueless due to the technology becoming obsolete, climate change or social chaos, for example. Second, people and society might simply prefer consumption now to consumption in the future – called the pure rate of time preference or, sometimes, myopia. Third, if it is expected that incomes will increase, the marginal welfare gain from an additional unit of consumption will be lower in the future, when people are richer, meaning that any given increase in consumption is more valuable now than in the future. Accordingly it is standard practice to discount future costs to their present values to allow for differences in the value of one extra unit of consumption over time.
1157
1158
65
Zinc Deficiency and DALYs in India: Impact Assessment and Economic Analyses
. Figure 65-2 The impact of discounting DALYs. This figure illustrates the impact of discounting DALYs by showing how the present value of 100 DALYs diminishes the higher the discount rate (r) and the more years the saving of these DALYs lies ahead in the future (t); without discounting (r = 0%) the current equivalent of 100 DALYs that are saved in the future is also 100 DALYs. DALYs: disability-adjusted life years. Source: Adapted from Stein (2006)
It is now usual practice to use a discount rate of 3% in DALY calculations (Murray and Lopez, 1996; Stein et al., 2005; Tan-Torres Edejer et al., 2003; WHO, 2002; World Bank, 1993). Nevertheless, when calculating DALYs it is considered good practice to carry out sensitivity analyses in which the discount rate is varied and includes a 0% rate (Murray and Lopez, 1996; Stein, 2006). Sensitivity analysis can be used both to illustrate the impact of the choice of discount rate on the final results and to facilitate comparisons across different programs and studies.
2.2.2
The Use of Disability Weights
The use of disability weights has also been criticized because they neither value the intrinsic value of the life of a person suffering from a condition, nor question their potential contribution to society or their capacity to realize individual achievements and to have a fulfilled life. Disability weights merely measure the loss of function of individuals, i.e., the degree to which they are unable to achieve their full physical and cognitive potential. The disability weights used in the GBD project were derived at a meeting convened at the WHO, in which a rigorous consultative protocol was followed; the results matched closely the pooled results of similar previous exercises (Arnesen and Nord, 1999; Murray and Lopez,
Zinc Deficiency and DALYs in India: Impact Assessment and Economic Analyses
65
1996). Subsequent studies have found that the approach of putting person trade-off questions to a group of experts is sufficiently robust and can yield comprehensive and coherent disability weights (Baltussen et al., 2002; Stouthard et al., 2000). As most decisions in public health imply a certain trade-off between one treatment or group of patients against another, disability weights contribute to greater transparency and objectivity. Allotey et al. (2003), however, note that it may be necessary to assess their general validity and to adapt disability weights to particular settings. The disability weights reported below for India are the outcome of a workshop – with experts from the Indian subcontinent – in which the disability weights used in the GBD project (Murray and Lopez, 1996) were used as benchmarks to set the appropriate disability weights to use in a developing country context (Stein et al., 2005). One of the more serious reproaches to the use of disability weights is that they discriminate against the disabled because they imply that saving the lives of disabled people saves fewer DALYs than that of fully functional individuals (Anand and Hanson, 1998; Arnesen and Nord, 1999). Yet this problem does not actually arise because when calculating the YLL no disability weights are applied, i.e., 1 year of life lost from premature mortality counts the same for all people irrespective of whether they are disabled. A more detailed discussion of the use of disability rates, including a justification for the use of disability weights is given by Stein (2006), and the concept as such is explained well by Tan-Torres Edejer et al. (2003).
2.2.3
The Basis for Calculating the Average Remaining Life Expectancy
In the GBD project a hypothetical general life expectancy was used to derive the average remaining life expectancies (Lj). This general life expectancy was based on the longest observed life expectancy of 82.5 years for Japanese women – adjusted to 80 years for men because of the biological difference in survival potential (Murray and Lopez, 1996). The use of a standardized approach was necessary because the GBD project undertook to measure the burden of all diseases and injuries at once at a given point in time, i.e., for the average remaining life expectancies a world without diseases and accidents had to be assumed. Moreover, because the GBD was calculated for the entire world, this approach ensured inter-regional equality: no matter where a premature death occurred – be it in a rich country with a long-living population or in a poor country with a low national average life expectancy – general life expectancy counted the same in terms of the number of DALYs lost. However, the gender differentiation has provoked criticism because of the inherent inequality and the potential ethical implications if decisions in the field of health are based on biological – or genetic – factors (Lyttkens, 2003). For analyses of individual conditions at a national level, such as zinc deficiency in India, the issue of which life expectancy to apply (the national one or a theoretical maximum life expectancy, possibly differentiated between men and women) can be circumvented because eradicating the condition probably only has a minor impact on national average life expectancy and, therefore, national life tables can be used (Stein, 2006). More details are given in the section on ‘‘Calculating the loss of DALYs due to zinc deficiency’’.
2.2.4
The Exclusion of Age Weights
A departure from the DALY formula used by Murray (1996) is that no age weights are included in the formula presented here, which is taken from Zimmermann and Qaim (2004). Murray had proposed age-based weights to value the lives of young, productive adults higher than the
1159
1160
65
Zinc Deficiency and DALYs in India: Impact Assessment and Economic Analyses
lives of infants and the elderly, i.e., more DALYs are lost if a disease is suffered by a 33 year old person than if the same disease is suffered by a 11 year old child or a 66 year old individual, and more DALYs are saved if the life of a 33 year old person is extended by 5 years than if the life of either the child or the elder individual is extended by 5 years. Murray justified this approach based on studies on the social and individual willingness to pay for health care, and the contribution of young adults to social welfare and their role as ‘‘care-givers’’ for their children and perhaps also their parents. Age weights have been criticized on various grounds (Anand and Hanson, 1998; Richardson, 1999; Williams, 1999), not least because weighting an individual’s social importance implies an ethical value judgment and could represent an opening of Pandora’s Box (Lyttkens, 2003) that leaves the question where to stop with the weighting (Stein, 2006): are doctors or film stars more important than other people, or orphans less? By not using age weights (or using unity weights for all individuals) this problem can be reduced. Because even if Musgrove (2000) rightly points out that treating all age groups equally is not a value-free judgment either, the ethical ramifications of not using age weights are less far reaching. Given the concerns outlined here, it is also noteworthy that in their sensitivity analyses of the GBD calculations Murray and Lopez (1996) had changed the age-weighting assumptions.
2.2.5
Arguments in Favor of the DALY Approach
As explained above, the DALY formula presented here is a modification of the original one to counter some of the general criticisms of the DALY methodology. While quantifications in the field of public health will inevitably be controversial, it is nevertheless often expedient to do the calculations – and DALYs are a transparent and widely used way of doing so. In particular, DALYs differ from other economic measures of disease because – unlike, for instance, the costof-illness or willingness-to-pay (WTP) approaches that aim at approximating the impact of ill-health through monetary quantification – they are not influenced by the earnings of individuals (or the national product). Hence, DALYs are more equitable. (In the case of cost-of-illness analyses higher incomes of individuals increase the benefit of avoiding or curing a disease that affects them – e.g., against controlling a disease that affects poorer individuals. Hence, one result of such analyses may be that curing the more affluent members of a society (or the citizens in richer countries) should take precedence over saving the lives of the poor. Similar outcomes are possible with WTP approaches, as health typically has a positive income elasticity of demand, i.e., the more individuals earn the more they are willing to pay for health.) While standardization can help to reduce the equity issue of these alternative approaches, it cannot quantify the actual burden of ill-health. DALYs, however, can be used for both: they measure the burden of ill-health directly and, by attaching a standard monetary value to a DALY, they can also be used for monetary approximations of the economic burden of a disease (see ‘‘Using DALYs for impact assessment and economic analyses’’).
3
Calculating the Loss of DALYs Due to Zinc Deficiency
In its World Health Report 2002 the WHO reported the number of DALYs that are lost due to various risk factors. According to this report it was estimated that in 2000 worldwide over 28 million DALYs were lost due to zinc deficiency; 9.6 million of which were in the countries in
Zinc Deficiency and DALYs in India: Impact Assessment and Economic Analyses
65
South-East Asia that have high child and high adult mortality rates, including India. In terms of DALYs lost, zinc deficiency was ranked among the top five risk factors that contribute to the burden of disease in high mortality developing countries; these were underweight (15% of the burden), unsafe sex (10%), unsafe water, sanitation and hygiene (5.5%), indoor smoke from solid fuels (3.7%) and zinc deficiency (3.2%) (> Figure 65-3). In an application of the methodology outlined above and largely based on the data presented below, Stein et al. (2007) calculated that zinc deficiency in India resulted in 2.8 million DALYs lost in 2004. This is less than the WHO estimated loss of 9.6 million DALYs in South-East Asia – where India is the biggest country. It is important to note, however, that the World Health Report reported the burden of zinc deficiency at the regional level and not the country level. Moreover, apart from the methodological differences explained above, and the perhaps more conservative assumptions presented in this section, the World Health Report did not directly calculate the loss of DALYs for zinc deficiency. Instead, the burden of zinc deficiency was based on attributing a share of the DALYs lost due to other, related conditions. (‘‘Worldwide, zinc deficiency is responsible for approximately 16% of lower respiratory tract infections, 18% of malaria and 10% of diarrheal disease.’’ WHO, 2002, p. 55.). This section describes the data required for calculating the number of DALYs lost due to zinc deficiency. While special reference is made to India, the requirements would be essentially the same also in other country contexts. Much of the information presented here draws on Stein et al. (2005) who summarize the outcomes of a series of workshops held in 2003 and 2004 at which nutritionists and economists discussed and substantiated the scientific basis for the recommendations. An overview of the data requirements and data sources is given in > Table 65-2, and an illustration of how DALYs are lost with different health outcomes is given in > Figure 65-4.
3.1
Outcomes and Target Groups
Zinc deficiency increases the risk of young children contracting diarrhea, pneumonia that is a severe acute respiratory infection, and growth faltering manifested as stunting or short stature. The increased susceptibility to diarrhea and pneumonia affects both infants – defined as children below 1 year old – and young children – age 1 through 5 years, while the onset of stunting affects only infants. Thus, the incidence rates of the above three conditions in the two age groups are needed. Furthermore, it is assumed that both diarrhea and pneumonia result in a certain percentage of fatalities.
3.2
Size of the Relevant Target Groups
The size of the two target groups – infants and young children – can be derived from population and demographic statistics of international organizations or official national sources. For India updated national census data is available online (Census, 2007).
3.3
Mortality Rates
The mortality due to zinc deficiency per se is unknown, but both diarrhea and pneumonia result in a certain percentage of fatalities. Jones et al. (2003) estimated that at least 4% of
1161
1162
65
Zinc Deficiency and DALYs in India: Impact Assessment and Economic Analyses
. Figure 65-3 The ten leading risk factors that contribute to the burden of disease in high mortality developing countries. This figure illustrates the importance of zinc deficiency as a public health problem in high mortality developing countries by showing its contribution to the burden of disease in percent, together with the contribution of the other nine leading risk factors (including iron deficiency, another micronutrient deficiency) that contribute to the burden of disease in these countries. Source: After WHO (2002)
mortality among children under-5 years old in low-income countries could be prevented if the intake of zinc was adequate. Hence, mortality due to zinc deficiency in children under 5 years old in India can be approximated by multiplying the overall mortality rate of children under 5 years old by 0.04. The under-five mortality rate for India in 2005 was 74 per 1,000 live births (UNDP, 2007; UNICEF, 2007; WHO, 2007a; World Bank, 2007a). Because mortality rates for children under-5 years old are given as the number of deaths per 1,000 live births, the target group for which the mortality rate needs to be applied is the number of live births and not the number of children under 5 years old.
3.4
Remaining Life Expectancy
Of the three adverse functional outcomes caused by zinc deficiency, only stunting is permanent. This means that the duration of stunting dij equals the average remaining life expectancy Lj.
65
Zinc Deficiency and DALYs in India: Impact Assessment and Economic Analyses
. Table 65-2 Data requirements and data sources to calculate DALYs for zinc deficiency in India Condition i Diarrhea
Target group j
Tj
Infants Directly Observed Therapy Strategy; FAO, Food and Agriculture Organization; FIND, The > Foundation for Innovative New Diagnostics; GAVI, > Global Alliance for Vaccines and Immunization; GDI, > Gender-related Development Index; GDP, Gross Domestic Product; GEM, > Gender Empowerment Measure; GFFATM, > Global Fund to Fight AIDS, Tuberculosis and Malaria; HDI, > Human Development Index; HHVI, The > Human Hookworm Vaccine initiative; HIV, Human Immunodeficiency Virus; HPI, Human Poverty Index; MDG, > Millennium Development Goals; MMV, The > Medicines for Malaria Ventures; ND, Neglected Disease; SARS, Severe Acute Respiratory Syndrome; STD, Sexually Transmitted Disease; TB Alliance, The Global Alliance for Tuberculosis Drug Development; UNAIDS, United Nations AIDS; UNDP, United Nations Development Program; WHO, World Health Organization; YLD, Years Lived with Disability; YLL, Years of Life Lost
1
Introduction
Infectious diseases continue to be the major causes of mortality in Africa. Well known existing, emerging and re-emerging diseases like malaria, tuberculosis, HIV/AIDS, cholera, meningitis, hepatitis, schistosomiasis, lymphatic filariasis, sleeping sickness, Ebola, SARS and others are causing suffering and mortality to a wide population in developing countries in general, and in Africa in particular (WHO, 2003a). Beyond mortality statistics, different methods can be considered to quantify the burden of disease expressed in terms of socio-economic costs such as productivity losses, care and treatment, hospitalization and handicap. In order to overcome the specific problems of each
The Impact of Infectious Diseases on the Development of Africa
66
. Table 66-1 Deaths and burden (DALYs) caused by infectious diseases, in thousands, 2002 (WHO, 2003b) Cause/disease HIV/AIDS
World Deaths
DALYs
Africa Deaths
DALYs
2821
86072
2203 (78%)
66772 (78%)
Malaria
1222
44716
1087 (89%)
39165 (88%)
Respiratory diseases
3845
90252
1071 (28%)
32703 (36%) 22992 (38%)
Diarrhea
1767
61095
695 (39%)
Tuberculosis
1605
35361
303 (19%)
8230 (23%)
11260
317496
53595 (48%)
169862 (48%)
Total
Data worked from the WHO statistics, by kind permission of the World Health Organization Nearly 50% of deaths and DALYs caused by infectious diseases in 2002 occurred in Africa
country, the most used method is the approach that measures the global burden of disease in terms of Disability Adjusted Life Years (DALYs) which is a combination of Years of Life Lost (YLL) through premature death, and Years Lived with Disability (YLD). Thus, DALY is thought of as one lost year of healthy life (Mathers et al., 2007) (> Table 66-1). Worldwide, nations are economically classified into two groups as developed and developing countries according to their level of development. Sometimes, a third group is specified as least developed countries. Using the per capita income, the World Bank has introduced four levels of income, namely: high-income, higher-middle-income, lower-middle-income and lower income (World Bank, 2003). Considering a broader definition of development that supplement income by other components like levels of health and education, the United Nations have developed five main composite indices to measure the average achievements in basic human development (human development index (HDI)), gender-related development index (GDI), > human poverty indices (HPI-1 and HPI-2) and the gender empowerment measure (GEM). HDI is the most used index giving a summary measure of human development, allowing a yearly comparison between countries around the world and indicating the relative ranking evolution in time of each country. HDI is a three dimensional composite index obtained as a mean of three indicators weighed equally: health (life expectancy at birth), standard of living (GDP per capita) and education (literacy and enrolment) (UNDP, 2006). In this chapter, we consider the impact of infectious diseases on the development of African countries, showing that, beyond health issues, these diseases should and must globally be seen as a development concern, affecting education and knowledge acquisition, income and social status, productivity and economic growth and other direct and indirect components of human development such as gender equality and human rights.
2
The Burden of Infectious Diseases in Africa
According to the World Health Organization report (WHO, 2003b), it has been estimated that, in 2002, nearly 60% of the 57 million total reported deaths in the world and approximately 47% of the global burden of disease is attributable to chronic diseases and cardiovascular diseases
1175
1176
66
The Impact of Infectious Diseases on the Development of Africa
. Figure 66-1 Disease burden (DALYs) among adults by broad cause, selected subregions, 2002 (WHO, 2003b). In Africa, the burden of communicable diseases is higher than those of non communicable diseases and injuries, whereas, in other regions, the burden of non communicable diseases is predominant. Reproduced by kind permission of the World Health Organisation
(CVDs) in particular. Deaths caused by > non communicable diseases dominate the mortality statistics in five out of six regions of the World Health Organization. The exception is Africa where deaths caused by HIV/AIDS, malaria, tuberculosis and other > communicable diseases are predominant (> Figure 66-1).
2.1
Major Infectious Diseases
With malnutrition as a common contributor, the five biggest infectious killers in Africa are acute respiratory infections, HIV/AIDS, diarrhea, malaria and tuberculosis, responsible for nearly 80% of the total infectious disease burden and claiming more than 6 million people per year. In five out of six WHO regions, the burden of non communicable disease is greater than that of communicable diseases. The exception is Africa where communicable diseases are predominant (> Table 66-2). Despite the success of vaccination programs for polio and many childhood diseases, malaria, tuberculosis, HIV/AIDS and others are still out of control in the majority of African countries. Children remain at high risk. Indeed, in 2002, of the 57 million deaths reported worldwide, 10.5 million deaths were among children of less than 5 years of age, of which 98% were in developing countries in general and in Africa in particular (WHO, 2003a, b, 2004, 2005) (> Table 66-3). Consequently, while life expectancy at birth reached 78 years for women in developed countries, it fell back to less than 46 years in sub-Saharan Africa.
66
The Impact of Infectious Diseases on the Development of Africa
. Table 66-2 Deaths by causes in WHO regions, estimates for 2002, in thousands (WHO, 2003b)
Cause
Communicable diseases, maternal and perinatal conditions and nutritional deficiencies
Non communicable diseases
Injuries
WHO regions Africa
7779
2252
The Americas
875
4543
540
Eastern Mediterranean
1746
2030
391
567
8112
803
South-East Asia
5730
7423
1467
Western Pacific
1701
9000
1231
18416
33424
5188
Europe
World % of total deaths
32.3%
747
58.6%
9.1%
Reproduced by kind permission of the World Health Organisation This figure shows the predominance of communicable diseases in Africa by opposition to the rest of the world where non communicable diseases are prevalent
. Table 66-3 Leading causes of deaths in children in developing countries in 2002 (WHO, 2003a) Condition
Numbers (in thousands)
% of all deaths
Perinatal conditions
2375
23.1%
Lower Respiratory Infections
1856
18.1%
Diarrheal diseases
1566
15.2%
Malaria
1098
10.7%
Measles
551
05.4%
Congenital anomalies
386
03.8%
HIV/AIDS
370
03.6%
Pertussis
301
02.9%
Tetanos
185
01.8%
Protein-energy malnutrition
138
01.3%
Other causes Total
1437
14.0%
10263
100%
Reproduced by kind permission of the World Health Organisation Lower respiratory infections, malaria, diarrhea, measles, AIDS and other infectious diseases are the leading causes of child mortality in developing countries and in Africa in particular. They affect life expectancy and human development of African countries in general
1177
1178
66 2.1.1
The Impact of Infectious Diseases on the Development of Africa
HIV/AIDS
The number of people affected by HIV/AIDS is exponentially increasing, passing from 10 million cases in 1990 to more than 25 million cases in 1996 and reaching 42 million in 2002 (> Figure 66-2). Around 75% of all cases are taking place in Africa.
. Figure 66-2 HIV/AIDS cases have skyrocketed. In 12 years time, the number of cases affected by HIV/AIDS worldwide has increased more than four-fold. In 2002, about three-quarter of all cases occurred in Sub-Saharan Africa. Reproduced by kind permission of UNDP
In 2005, 38.6 million [33.4–46.0] people were living with HIV worldwide, 4.1 million [3.4– 6.2] became newly infected and 2.8 million [2.4–3.3] lost their lives to AIDS. If the present trend is maintained, by 2020 AIDS will have caused more deaths than any other disease > epidemic in history. Moreover, beyond mortality figures and beside the disease considered as a public health challenge, HIV/AIDS is becoming a real problem of development. In Africa and especially in countries with high > prevalence of HIV/AIDS, devastating consequences are already strikingly apparent not only on health systems and health indicators but also in terms of income and productivity, education and knowledge, human rights and gender equality (UNAIDS, 2006). Nearly 80% of the 3 million global deaths from HIV/AIDS that occurred by the end of 2002 were in sub-Saharan Africa, dramatically cutting life expectancy and leaving a legacy of millions of orphans. Seven countries had a prevalence of HIV over 20% (UN, 2004).
The Impact of Infectious Diseases on the Development of Africa
2.1.2
66
Diarrheal Disease
It is well known that diarrhea is among the leading causes of mortality and morbidity in children in developing countries generally and in Africa particularly. Using different methods and sources of information, several authors have attempted to evaluate the burden of this disease. Despite the differences existing between estimates, they all show a declining trend in mortality figures but a relatively stable morbidity. (Rohde and Northrup, 1976; Snyder and Merson, 1982; NVD, 1986; Bern et al., 1992; Murray et al., 2001; Kosek et al., 2003; WHO, 2003a) (> Table 66-4). . Table 66-4 Disease burden due to diarrhea worldwide Deaths per year
Authors
Rohde and Northrup (1976) 5 million
Period covered and data used in the review Author’s data, 1976
Snyder and Merson (1982)
4.6 million
Review of data published between 1954 and 1979
NVD-US Institute of Medicine (1986)
3.5 million
Based on published data and field experience
Bern et al (1992)
3.3 million
Using data published between 1980 and 1990
Murray et al (2001)
1.4 million
Data analysis in 2000
Kosek et al (2003)
2.5 million
Using data published between 1992 and 2000
WHO (2003a)
1.9 million
Report: Global defence against the infectious disease threat, data 2001
Various estimates provided by different authors who have tempted to evaluate the disease burden caused by diarrhea diseases. They all show a declining trend in mortality but morbidity remains relatively stable
2.1.3
Malaria
The WHO statistics indicate that malaria claims more than 1 million lives a year. Beyond mortality, the disease affects more than 300 million every year with a high handicapping rate. Children, pregnant women and vulnerable people in general are the most exposed. Moreover, malnutrition and other diseases like pulmonary infections constitute favourable environment for the spread of malaria. Countries in tropical Africa bear the brunt of malaria, accounting for more than 90% of all cases occurring worldwide each year (Ruxin et al., 2005).
2.1.4
Tuberculosis
Tuberculosis is among the top ten causes of global mortality (Dye, 1999; Borgdorff et al., 2002). It has been estimated that approximately one-third of the world’s population is infected with tuberculosis bacillus, and that each year 8 million people develop tuberculosis disease and about 2 million die of the disease. Once more, the highest > incidence rates are found in Africa and South-East Asia. The HIV/AIDS epidemic and multi-drug resistance have worsened the tuberculosis situation over the last two decades. Tuberculosis is a leading killer of people with
1179
1180
66
The Impact of Infectious Diseases on the Development of Africa
HIV, and 80% of tuberculosis patients are HIV positive in countries with high prevalence of HIV (Dye, 1999; Ruxin et al., 2005; Laxminarayan et al., 2007) (> Table 66-5). However, despite the importance of weakened immunity due to HIV/AIDS, the global reductions in immunity caused by malnutrition, diabetes, co-infections, drug use, alcoholism, and the stress of poverty an migration could be as great if no greater than those caused by HIV/AIDS (Currey et al., 2007). . Table 66-5 The 22 countries the most affected by tuberculosis (Laxminarayan et al., 2007) Number of countries by region
Total Deaths
HIV + deaths
Excluding HIV + deaths
Africa (9 countries)
609986
444289
201760
Asia (9 countries)
987231
81826
888859
South America (3 countries)
74209
7148
69061
Europe (1 country)
34144
2902
31242
Total (22 countries)
1705612
536162
1205513
African countries are the most affected by the co-infections of HIV and tuberculosis
2.1.5
Respiratory Infections
According to WHO estimates, Respiratory infectious diseases were the first cause of mortality from infectious diseases in 2001, responsible for nearly 4 million deaths. In 2002, lower respiratory infections caused nearly 2 million deaths in children, ranking at the second leading cause. In Africa, this disease caused more than one million deaths and nearly 33 million DALYS in 2002.
2.1.6
Preventable Sexually Transmitted Diseases
Worldwide, in 1999, preventable sexually transmitted diseases (STD) caused about 350 million infections in the population aged 15–49. However, in the era of AIDS and the high level of politicisation and priority given to HIV, other sexually transmissible diseases may receive less attention. For instance, North African countries are known to have high prevalence of STD and relatively low prevalence of HIV/AIDS (Boutayeb, 2006). Although tools of prevention have been available for decades, congenital syphilis is still endemic in many African countries. In 1999, there were some 4 million cases of syphilis among adults in sub-Sahara Africa compared to 4 million cases in south and south-east Asia and 3 million cases in Latin America and the Caribbean (Berman, 2003; Hawkes et al., 2003; Peeling and Ye, 2003; Schmid, 2003).
2.1.7
Neglected Diseases
Beyond mortality figures, the health impact of a number of infectious diseases can be measured by severe and permanent disabilities and deformities affecting approximately 1 billion people in the world and causing millions of DALYs. Africa is particularly concerned
The Impact of Infectious Diseases on the Development of Africa
66
with these diseases. Indeed, lymphatic filariasis, leishmaniasis, schistosomiasis, sleeping sickness, dengue, Chagas disease, Buruli ulcer and others are responsible for impaired childhood growth, mental retardation, blindness, amputation and diverse > disability conditions (Derouich and Boutayeb, 2006; Boutayeb, 2007; See details in the chapter The Burden of > Neglected Diseases in Developing Countries of this book).
3
Impact of Infectious Diseases on the Development
3.1
Sectorial Impact
3.1.1
Impact on Health Indicators
In Africa, respiratory diseases, HIV/AIDS, diarrhea, malaria, tuberculosis and other infectious diseases are directly affecting health and demographic indicators such as mortality rates, life expectancy, and sex and age distributions. In particular, millions of healthy years are amputated from African populations due to infectious diseases (> Figure 66-3). Globally, in . Figure 66-3 Loss of life expectancy due to HIV/AIDS. From 2000 to 2005 Zimbabwe, Botswana, Swaziland and Lesotho have lost respectively 35, 28, 28 and 24 years in life expectancy due to HIV/AIDS. Reproduced by kind permission of UNDP
1181
1182
66
The Impact of Infectious Diseases on the Development of Africa
1995–2000, 38 African countries had a mean life expectancy of 47 years, representing tens of years of loss attributable to infectious diseases. In the most affected countries by HIV/AIDS, life expectancy declined by 12.1 years during the period 1995–2000 and it is expected to decline by 29.4 years by 2010–2015. More generally, deaths and especially infant mortality are ravaging sub-Saharan Africa (> Table 66-6). The impact of infectious diseases on health indicators affects individuals, households, communities and the whole nation (hospitalisation, healthcare, orphan hood). In sub-Saharan Africa, HIV/AIDS and related diseases are mobilizing more than half of all hospital beds. In some countries, 30%–50% of hospital admissions and around 50% of out-patient visits are due to malaria which is also responsible for more than 30% of hospital deaths. Many African countries have lost a large part of their healthcare workforce due to AIDS and other infectious diseases. In other countries, midwives and health workers are affected by infectious diseases (UN, 2004). . Table 66-6 Estimated and projected impact of HIV/AIDS on mortality indicators in the seven most affected countries in Africa 1995–2000
2010–2015
Life expectancy at birth (years) Without AIDS
62.3
67
With AIDS
50.2
37.6
Absolute difference
12.1
29.4
Without AIDS
3
3
With AIDS
5
10
Absolute difference
2
6
Without AIDS
55.4
40.7
With AIDS
66.1
54.6
Absolute difference
10.2
13.9
80.2
56.9
108.8
100.2
28.6
43.3
Number of deaths (millions)
Infant mortality rate (per 1000)
Child mortality rate (per 1000) Without AIDS With AIDS Absolute difference
Reproduced by kind permission of UNDP In Africa in general and in the most affected countries in particular, HIV/AIDS is seriously affecting life expectancy, rates of mortality and child mortality. If the current trend is not reversed or at least stabilised then African countries will be exposed to a catastrophe
3.1.2
Impact on Economic Indicators
Worldwide, about 2 billion people have inadequate or no access to life-saving treatments. More than 80% of these deprived people are living in developing countries where infectious diseases constitute serious impediments to economic development by reducing productivity,
The Impact of Infectious Diseases on the Development of Africa
66
setting aside saving possibilities and slowing economic growth in general. In many African countries, the quasi totality of public health expenditure is a consequence of infectious diseases. For instance, in tropical rural areas with limited access to preventive and curative health facilities, malaria can have devastating consequences on agriculture households by a simple episode that coincides with a plantation or harvest season. It is estimated that malaria costs Africa more than US$12 billion a year, slowing its economic growth by 1.3% annually (Bartram et al., 2005). Contrary to the majority of diseases, HIV/AIDS kills and disables adults in the best productive part of their lives, affecting business, investment, industry and agricultural sustainability. African agriculture labour force is particularly affected (> Figure 66-4). . Figure 66-4 Reduction in African agriculture labour force due to HIV/AIDS, as estimated in 2000 and projected for 2020 (UNAIDS, 2006). If the present trend is maintained, by 2020, HIV/AIDS will have caused more than a 25% reduction in agriculture labour force in some African countries. Reproduced by kind permission of UNAIDS
Sleeping sickness is having devastating consequences on the development in Africa. It causes over 3 million livestock deaths each year and an annual loss of US$4.5 billion in agriculture. Similarly, gains following lymphatic filariasis elimination are expected to approach US$4 billion per year. More globally, infectious diseases are reducing families’ income and slowing economic growth as indicated by the recent report released by the World Bank on tuberculosis control, showing that besides deaths, tuberculosis costs more than US$3.3 billion annually in lost productivity though important economic benefit is at hand (Laxminarayan et al., 2007) (> Table 66-7). At the moment, it is difficult to estimate precisely the real economic impact that infectious diseases are having on the whole African continent. However, many studies have attempted to evaluate the disease burden case by case or country by country. In the case of HIV/AIDS, the cost was estimated to be between 11.7 and 35.1% of the GNP in Africa
1183
1184
66
The Impact of Infectious Diseases on the Development of Africa
. Table 66-7 Costs and Economic Benefit of TB Control Strategies in Sub-Saharan African countries, in billions of dollars (2006–2015) (Laxminarayan et al., 2007) Sustained DOTS (relative to no DOTS)
Costs
Benefits
Africa
Ratio
12.24
129.44 [112.81–146.07]
11 [9–12]
Africa High HIV+
9.45
97.59 [85.83–109.35]
10 [9–12]
Africa Low HIV+
2.79
31.85 [26.89–36.80]
11 [10–13]
High Burden African countries
7.70
81.06 [71.34–90.77]
11 [9–12]
African countries can make important economic benefits by controlling TB through a sustained DOTS strategy. The ratio between benefit and cost can be greater than 10
(Macroeconomics, 2001). Other figures were given by different organisms like the World Health Organization (WHO, 2003b), The European Parliament (2005), The World Bank (World Bank, 2003), and UNAIDS (UNAIDS, 2006).
3.1.3
Impact on Education
As stressed in the Millennium Development Goals, education is essential for human development and needs to be enhanced especially in sub-Saharan African countries. Unfortunately, malaria, tuberculosis, HIV/AIDS and infectious diseases in general, are reversing the trend towards the achievement of universal primary education in most African countries. In Africa, less than 65% of children are enrolled in primary school and thousands of enrolled children will prematurely leave school under the pressure of infectious diseases, including orphans, disabled, impoverished and those who withdraw to look after ill members of their families (UNICEF, 2005; UNAIDS, 2006; UNESCO, 2007) (> Table 66-8). During the period 1999–2004, orphaned children represented 12.3% of all children under age 18 in sub-Saharan Africa and the percentage of child labour reached 41% in West and Central Africa (UNICEF, 2005). More globally, these diseases are seen to have a four-fold impact on education. They affect the cognitive ability of children, the capacity of teachers, the upbringing of families and the efficiency of staff and managers. For instance, HIV prevalence among South African . Table 66-8 Impact of orphanhood on school attendance among 10–14 years-olds (%) (UNAIDS, 2006) Percentage in school Non-orphan
West: 9 countries
Central: 6 countries
Eastern: 9 countries
Southern: 10 countries
All: 34 countries
67
75
70
88
74 69
Orphan
58
69
54
84
Double orphan
57
58
49
80
64
Ratio double versus non orphan
.86
.94
.72
.90
.87
Reproduced by kind permission of UNAIDS In African countries, orphan children are more likely to miss school than others
The Impact of Infectious Diseases on the Development of Africa
66
teachers reaches 21% among those aged 25–34 and 13% among those aged 35–44, whereas in the Zambian school system, over 60% of teacher absence is due to HIV/AIDS (illness, care for ill family members, family funeral, etc..) (UNAIDS, 2006). Malaria is the leading cause of mortality in the under-five African children. Consequently, it affects the education capacity of African countries. Similarly and on top of high mortality figures, with a median of 3.2 episodes per child-year, diarrhea is highly affecting school attendance.
3.2
Global Impact on Economic and Human Development
As indicated in the previous sections, infectious diseases are globally affecting economic and human development of African countries. Malaria is responsible for 10% of Africa’s disease burden and the current GDP is thought to be 32% lower than it would have been without malaria. Consequently, in Africa, malaria is seen both as a disease of poverty and a cause of poverty. The devastating impact of HIV/AIDS is recognised in the majority of national human development reports as stressed in the South Africa human development report 2003 (South Africa, 2003), the Zimbabwe human development report 2003 (Zimbabwe, 2003) and the Malawi human development report 2005 (Malawi, 2005). These reports give details of the impact of HIV and AIDS on household welfare, orphaned children, the extended family, educational and health sectors, agricultural production, business and public service delivery. They also stress that HIV/AIDS and its far-reaching consequences mean that the disease is no longer a crisis only for the healthcare sector, but presents a challenge to all sectors. It has the potential to reverse those gains made in human development in the last few years. Economically, 34 African countries belong to the lower-income group, 12 are lowermiddle-income countries, only eight countries are classified as higher-middle-income and no African country has reached the high-income level. Sadly and more interestingly, it can be seen from the UNDP report 2006 (UNDP, 2006) that infectious diseases are affecting all components of human development (HD) in African countries. Indeed, no African country belongs to the ‘‘High HD’’ group, almost all of the ‘‘Low HD’’ countries at the bottom of the index table are in sub-Saharan Africa, and the remaining African countries occupy uncomfortable places among the ‘‘Medium HD’’ group. Moreover, in 15 years (1990–2006), the most affected African countries by infectious diseases in general and by HIV/AIDS in particular have lost tens of places in the human development ranking. In 2001, the Commission on Macroeconomics and Health provided empirical evidence on how investing in health can achieve economic development and poverty reduction. It was estimated that eight million lives per year could be saved by essential interventions against infectious diseases and nutritional deficiencies by 2010, resulting in economic benefits adding up to more than US$360 per year by 2015. In the same spirit, the United Nations (UN) Millennium Summit adopted in 2001 the Millennium Development Goals (MDGs) by fixing eight goals to be reached in 2015. Preventing the spread of HIV/AIDS, tuberculosis, malaria, and other infectious diseases is one of the goals. However, mid-way, most African countries, have made little (if any) headway in preventing and controlling infectious diseases like HIV/ AIDS, malaria, tuberculosis, diarrhea, respiratory disease and a multitude of the so-called neglected diseases. Opposite to that, wars and conflicts are financed at the expenses of disease control as it can be sadly noticed from the amount of US$7 billion that African governments dedicated to military spending in 1999 (Mashelkar, 2005).
1185
1186
66 4
The Impact of Infectious Diseases on the Development of Africa
Conclusions
In African nations, millions of people live with less than 1$ a day and on fragile and often remote rural ecosystems, most of them lack access to basic health services and safe drinking water. Moreover, many countries, being heavily indebted, affect less than 1% of the national global budget to health, and few governments are putting science, technology and innovation at the centre of their strategies. Considering the treatment cost of communicable and chronic diseases, and the level of poverty, the most affected countries are unable to cope with the burden of disease. For health strategies to be successful, international solidarity and public-private partnerships are needed to tackle the problems of shortage and lack of treatments, resistance and the need for new drugs, vaccines and diagnostic procedures. In this direction, several programmes have already been launched such as the Global Fund to Fight AIDS, Tuberculosis and Malaria (GFFATM), Global Alliance for Vaccines and Immunization (GAVI), The medicines for Malaria Ventures (MMV), the Global Alliance for TB Drug Development (TB Alliance), The Human Hookworm Vaccine initiative (HHVI), the Foundation for Innovative New Diagnostics (FIND), the Drug for Neglected Diseases Initiative (DNDi), Stop TB, Roll back Malaria, etc. . .. However, this global strategy is insufficient without the national and local implication. Health decision makers, non governmental organizations (NGOs), research institutions, community groups and individuals must join their efforts in order to attenuate the incidences of specific diseases, control the spread of epidemics and development of complications, and optimise the health management of human and material resources. Stressing that treatment has been the most neglected element in developing countries where almost 6 million people will die from AIDS in the near future if they do not receive treatment, the WHO report 2004 calls for a comprehensive HIV/AIDS strategy that links prevention, treatment, care and long-term support. The objective fixed by WHO and its partners to provide 3 million people in developing countries with antiretroviral therapy by the end of 2005 was far to be reached, suggesting that more effort and efficient measures are needed. It is also hoped that combined efforts between donor nations and affected countries will provide the US$2–3 billion per year required to scale up the response against malaria in endemic areas. Similarly, although 16 million patients have been treated so far with the WHOrecommended directly observed therapy strategy (DOTS), yet more than half of those affected by tuberculosis still do not have access to this treatment, especially those living in the Nine African countries the most affected by tuberculosis. However, an indisputable economic benefit is predicted from a sustained DOTS and global plan strategy as indicated by the World Bank report (Laxminarayan et al., 2007). A recent cost-benefit analysis by WHO showed that achieving the global MDG target in water and sanitation would bring substantial economic gain in both health and other benefits (consequences of reduction in diarrheal episodes): each $1 invested would yield an economic return between $3 and $34 depending on region. The health-related costs avoided would reach $7.3 billion per year, and the annual value of adult working days gained as a result of less illness would be almost $750 million. Last but not least, tens of thousands of deaths can be avoided in Africa, and billions of dollars saved by reducing the impact of respiratory diseases and the so-called neglected diseases such as schistosomiasis, intestinal helminth infections, trachoma, dengue and others. Meanwhile, by the dawn of the third millennium, a Japanese woman can expect to receive, on average, care and medications worth about US$550 per year and much more if needed. Whereas, a woman in the least developed African countries can expect, on average, medicines
The Impact of Infectious Diseases on the Development of Africa
66
worth about US$3 per year. Consequently, a woman in sub-Saharan Africa is 100 times more likely to die in pregnancy or childbirth than is a woman in Western Europe.
Summary Points Infectious diseases are causing about 15 million deaths annually with more than 80% in Africa
Low-income economies is the group of countries where the annual per capita income is less than 600 dollars
Tuberculosis is the second leading infectious disease, causing about 2 million deaths every year
Malaria is responsible for one million deaths annually almost exclusively happening in Africa
HIV/AIDS, appeared about 25 years ago, this disease has become a real development
problem, causing about 3 million deaths annually, responsible for more than 7% of the global disease burden (DALYs) Diarrheal disease is the third leading cause of child mortality, killing about 1.5 million children every year Respiratory Diseases constitute the second killer of children with nearly two million deaths annually Neglected Diseases is the name given to a multitude of diseases, including lymphatic filariasis, leishmaniasis, schistosomiasis, sleeping sickness, dengue, Chagas disease, Buruli, ulcer, trachoma and others Infectious diseases are reducing life expectancy in many African countries. For instance, the reduction due HIV/AIDS is between 25 and 35 years in Botswana, Swaziland and Zimbabwe The economic cost of HIV/AIDS is estimated to be between 11.7 and 35.1% of the GNP in Africa Sleeping sickness is having devastating consequences on the development in Africa. It causes over 3 million livestock deaths each year and an annual loss of US$4.5 billion in agriculture Malaria costs Africa more than US$12 billion annually
References Bartram J, Lewis K, Lenton R, Wright A. (2005). Lancet. 365: 810–812. Berman SM. (2003). Bull World Health Organ. 82: 433–438. Bern C, Martines J, de Zoysa I, Glass RI. (1992). Bull World Health Organ. 70: 705–714. Borgdorff MW, Floyd K, Broekmans JF. (2002). Bull World Health Organ. 80: 217–227. Boutayeb A. (2006). Trans R Soc Trop Med Hyg. 100: 191–199.
Boutayeb A. (2007). Developing countries and neglected diseases: challenges and perspectives. Int J Eq Health. doi: 10.1186/1475-9276-6-20. Currey B, Quamruzzaman Q, Rahman M. (2007). Lancet. 371: 1401–1403. Derouich M, Boutayeb A. (2006). Appl Math Comput. 177: 528–544. Dye C. (1999). Int J Tub Lung Disease. 4: 146–152. European Parliament (EP). (2005). Report on major and neglected diseases in developing countries. Available
1187
1188
66
The Impact of Infectious Diseases on the Development of Africa
at http://www.europarl.europa.eu (accessed at 8 December 2007). Hawkes S, Miller S, Rechenbach L, Nayyar A, Buse K. (2003). Bull World Health Organ. 82: 417–423. Kosek M, Bern C, Guerrant RL. (2003). Bull World Health Organ. 81: 197–204. Laxminarayan R, Klein E, Dye C, Floyd K, Dareley S, Adeyi O. (2007). Economic Benefit of Tuberculosis Control. The World Bank, Washington, DC. Malawi HDR. (2005). Reversing HIV/AIDS in Malawi. United Nations Development Programme, Washington, DC. Mashelkar RA. (2005). Innov Strat Today. 1: 16–32. Mathers CD, Ezzati M, Lopez AD. (2007). Measuring the burden of neglected tropical diseases: the global burden of disease framework. PLOS Negl Trop Dis. doi: 10.1371/journal.pntd.0000114. Murray CJL, Lopez AD, Mathers CD, Stein C. (2001). The Global Burden of Disease 2002 Project. Word Health Organization, Geneva. NVD. (1986). Establishing Priorities volume 2. Institute of Medicine, Washington, DC. Peeling RW, Ye H. (2003). Bull World Health Organ. 82: 439–446. Rohde JR, Northrup RS. (1976). Ciba Foundation Symposium Acute Diarrhoea in Children. Elsevier, Amsterdam, pp. 339–365. Ruxin J, Paluzzi JE, Wilson PA, Tozan Y, Kruk M, Teklehaimanot A. (2005). Lancet. 365: 618–621. Schmid G. (2003). Bull World Health Organ. 82: 402–409. Snyder JD, Merson MH. (1982). Bull World Health Organ. 60: 605–613.
South Africa. (2003). Human development report. Available at http://www.undp.org.za (accessed at 8 December 2007). UNESCO. (2007). Education for all in least developed countries. Available at shttp://unesdoc.unesco.org (accessed at 8 Dec 2007). UNICEF. (2005). The state of the world’s children. The United Nations Children’s Fund, New York. United Nations. (2004). World population prospects. United Nations, Geneva. UNAIDS. (2006). The impact of AIDS on people and societies. Available at http://data.unaids.org (accessed at 20 January 2008). UNDP. (2006). Human development indicators. Available at http://hdr.undp.org (accessed at 8 December 2007). World Bank. (2003). Sustainable development in a dynamic world. Available at http://www.worldbank. org (accessed at 20 January 2008). WHO. (2003a). Global defence against the infectious diseases threat. World Health Organization, Geneva. WHO. (2003b). World Health Organization annual report. Available at http://www.who.int/whr/2003/en (accessed at 20 January 2008). WHO. (2004). World Health Organization annual report. Available at http://www.who.int/whr/2004/ en (accessed at 20 January 2008). WHO. (2005). World Health Organization annual report. Available at http://www.who.int/whr/2005/ en (accessed at 20 January 2008). Zimbabwe HDR. (2003). Redirecting our responses to HIV and AIDS. United Nations Development Programme, Washington, DC.
67 Measuring the Global Burden of Tuberculosis I. Onozaki . N. Ishikawa . D. A. Enarson 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1190
2
Global TB Burden and its Trend . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1192
3 3.1 3.2 3.3 3.4 3.5 3.6
How the Burden is Estimated . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1192 Incidence as an Indicator of Burden . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1192 Estimation of Incidence Using Different Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1195 Estimation of Incidence from Annual Risk of TB Infection . . . . . . . . . . . . . . . . . . . . . . . 1198 Estimation of Incidence from Disease Prevalence Surveys . . . . . . . . . . . . . . . . . . . . . . . . . 1198 Estimation of Incidence from Vital Registrations of Deaths . . . . . . . . . . . . . . . . . . . . . . . 1199 Estimation of Trends in Incidence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1200
4
4.3 4.4
Prevalence as an Independent Indicator for Millennium Development Goals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1200 How to Detect TB Cases in the Community . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1201 Culture Examinations are Recommended to Measure TB Burden and to Confirm TB . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1204 High Risk Groups and Their Role in Resurgence of TB . . . . . . . . . . . . . . . . . . . . . . . . . . . 1205 Prevalence Survey Brings Other Benefits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1205
5
Delay Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1205
6
New Technology for Screening and Diagnosis to Measure the Burden . . . . . . . . . . 1205
7
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1206
8
Case Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1206
4.1 4.2
Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1209
#
Springer Science+Business Media LLC 2010 (USA)
1190
67
Measuring the Global Burden of Tuberculosis
Abstract: There are several ways to measure the burden of TB. Estimated TB incidence and its rate have been utilized as the most popular indicators of TB burden since 1997. According to the WHO’s estimation of the global TB burden, there were 9.2 million new TB cases in 2006 with 1.7 million deaths due to TB. Although the estimated global incidence rate stabilized or began to fall slowly, the number of new cases was still increasing because population growth had a greater effect on the number of cases than did the declining rate of TB. However, for each individual country, the uncertainty of the incidence estimation became a concern for monitoring and evaluating the progress of the TB situation at the country level. Methods for estimating current burden were reviewed. An important limitation is that the current estimation of the burden, incidence, is mainly based on the notification data from countries and assumption of case detection rate. Very few countries have scientific data to estimate the epidemiological situation and trend. It is essential to strengthen routine recording and reporting systems, and surveillance, to improve the estimates of burden and trend. However, as the current surveillance systems are not always reliable in many countries where TB is common, it is recommended that a prevalence survey be conducted every 5–10 years. Prevalence itself is one of the MDG’s indicators, and survey findings can be utilized to evaluate trends derived from routine surveillance. Combinations of prevalence surveys and notification data might be the best method to measure disease burden. Although a prevalence survey is costly and labour intensive, it may be a great help to a country in improving TB control policy. Surveys made in some countries might help neighbouring countries with similar situations to understand their epidemiological status of TB through interpretation of routine surveillance data. By adding delay analysis, much insight can be gained into the quality and use of TB services. List of Abbreviations: BCG, Anti- Tuberculosis vaccine containing the Bacille de Calmette et Gue´rin used by the expanded immunization programme worldwide; CDR, Case Detection Rate; DALYs, Disabled Adjusted Life-Years; DOTS, The brand name of a comprehensive TB control program recommended by the WHO and Stop TB Partnership originated from Directly Observed Treatment, Short Course; HBC, High-burden country where the absolute number of TB incident cases is high. 22 HBCs comprise 80% of the Global TB burden; HIV, Human Immuno-deficiency Virus; IGRAs, Interferon gamma release assays; MDGs, > Millennium Development Goals; MDR TB, Multidrug-Resistant Tuberculosis. TB that is resistant, at a minimum, to the two major anti-TB drugs Rifampicin and Isoniazid; NTP, National TB Control Program; TB, Tuberculosis; WHO, World Health Organization; XDR TB, Extensively Drug Resistant Tuberculosis, which is MDR but also resistant to any fluoroquinolone and any second-line anti-TB injectables, namely Amikacin, Kanamycin or Capreomycin
1
Introduction
The Burden of Tuberculosis (TB) usually refers to the epidemiological burden and is measured by disease incidence, prevalence and/or mortality, and infection prevalence, and incidence (annual risk). It may also be measured by economic indicators such as medical expenditure on TB and economic loss due to TB. Until the early 1990s, short-course chemotherapy, a rifampicin-containing regimen had been believed to be too expensive for developing countries. However, the cost effectiveness of TB control was demonstrated by estimating the number DALYs saved by its activities, and the World Bank started to invest in TB control (Murray et al., 1991; World Bank, 1993). This marked the turning of the tide in terms of international assistance to TB control with the adoption of DOTS as the international standard for TB services (WHO, 1994; WHO Tuberculosis programme, 1994).
Measuring the Global Burden of Tuberculosis
67
Since 1997, the WHO publishes the Global TB report every year with estimates of disease incidence, prevalence and mortality at global, regional and national levels (WHO, 2008). The indicators are shown in absolute numbers and rates per 100,000 population. High burden countries (HBCs) for TB have been designated based on the absolute number of the estimated new TB cases per year, or incidence, in a country. Since then, the WHO’s estimation of TB incidence has been widely used to reflect the burden. According to the Global TB report of 2008, there were an estimated 9.2 million new cases in the world in 2006 (139 per 100,000 population) including 4.1 million smear positive cases. Twenty-two HBCs account for 80% of the global TB burden in terms of incidence (WHO, 2008). For each country, smear positive incidence per 100,000 population has been the most popular indicator to express a country’s TB burden for a decade. Because one of the two pillars of international outcome targets was to detect 70% or more of existing smear positive cases (a case detection rate of 70% or more), along with a treatment success rate of 85% or more (WHO Tuberculosis programme, 1994), countries have become more interested in incidence. It is estimated that detecting 70% of incident cases and curing 85% of them will cause the TB burden to be halved in 8–12 years (in the absence of HIV) (Styblo and Bumgarner, 1991). The CDR target became a vital force to push countries to improve their case finding and their surveillance systems. It is also a visible indicator at the global level to demonstrate and monitor progress. In 2006, the global estimation of the case detection rate of smear positive TB was 61% and 77 countries met the 70% target (WHO, 2008). However, as DOTS, and better TB services expanded in countries, several questions were raised concerning the extent to which these estimations are reliable especially at the country level. We observed several discrepancies even within the 22 HBCs: Myanmar achieved 70% detection in 2003, and efforts on further expansion of the service obtained 100% or more CDR consecutively in 2005, 2006 and 2007; Vietnam has been achieving both CDR and treatment success targets since 1997. However, it experienced no decline of case notification; despite extensive efforts, most African countries such as Nigeria, Ethiopia, Tanzania and Uganda fell far short of the 70% CDR target. Is the disease burden in Asian countries underestimated? Is it overestimated in African countries? How do countries reliably demonstrate impact of intervention in terms of decline of incidence? The uncertainty over the current approach to estimation of TB burden has created much discussion, and several countries are now showing interest in surveys to understand the epidemiological burden more accurately. Current global targets and indicators for TB control have been developed within the framework of the MDGs as well as by the Stop TB Partnership and the WHO’s World Health Assembly (Stop TB Partnership, 2006). The impact targets are to halt and reverse TB incidence by 2015 and to halve the prevalence and death rates by 2015 compared with a baseline of 1990. Although it is questionable how accurately we can estimate a 1990 baseline, it is notable that a measurable indicator derived from community surveys, namely disease prevalence, became one of the indicators of the MDGs. Disease prevalence surveys began to be promoted to more accurately determine the TB situation in countries (Dye et al., 2008; WHO, Western Pacific Region, 2007), even as serious limitations of tuberculin surveys began to be recognized (Dye, 2008; Reider, 1995). In this chapter, we first summarize the current estimation of the global TB burden by the WHO. Second, we try to provide an overview of current methods to measure burden, and, third, we discuss what might be the best way for a country to measure its TB burden especially in resource poor high burden settings.
1191
1192
67 2
Measuring the Global Burden of Tuberculosis
Global TB Burden and its Trend
The WHO estimated that there were 9.2 million new cases of TB in 2006 (139 per 100,000 population) including 4.1 million smear positive cases (WHO, 2008). Although HIV is often thought to be the cause of the resurgence of TB (Corbett et al., 2003), 0.7 million HIV-positive cases make up only 8% of the total TB cases. However, the African Region has the highest incidence per capita (363 per 100,000 population), while the Western Pacific and the South East Asian Regions accounted for more than half of the incidence, 55%. Twenty-two countries designated as HBC accounted for 80% of the incidence, and 50% of those incident cases in the 22 HBC, 3.8 million, were in the top three countries: India, China and Indonesia (WHO, 2008; > Table 67-1). The magnitude of the TB burden in each country can also be expressed as the incidence rate (WHO, 2008; > Figure 67-1). Higher incidence rates were observed mostly in countries in Africa, South East Asia and the Former Soviet Union. HIV co-infection, poverty and recent experiences of social instability seem to be major factors contributing to the high burden of TB in these countries. There were 29 countries that were estimated to have the highest incidence rate of all TB (300 or more per 100,000 population). All but three (Cambodia, East Timor and Kiribati) were countries in the African continent. The high incidence rates in African countries are partially explained by a high rate of HIV co-infection (Corbett, 2003). This may also explain the higher mortality in the African Region; while the African Region accounted for 31% of the incidence, it accounted for 39% of TB deaths. According to WHO estimates of global incidence of TB per capita, the incidence rate peaked around 2003 and appears to have stabilized or begun to decline slowly, while the mortality and disease prevalence rate began to decline earlier because these factors respond more quickly to the expansion of case detection and treatment (WHO, 2008; > Figure 67-2). Among 134 countries that have a reliable series of case notification reports for the decade 1997–2006, data from 93 countries indicate that the incidence rate was falling. However, globally, the slow decline of the incidence rate was overwhelmed by population growth in actually increasing the numbers of cases (the estimated incidence in 2005 was 9.1 million, while that of 2006 was 9.2 million). Decline of HIV prevalence in the general population in high HIV prevalence countries in Africa was mainly contributing to the declining incidence rate of TB in the region as well as that in the world. Recently, the threat of MDR and XDR TB has started to attract significant attention. While most drug sensitive TB can be cured with a 6-month chemotherapy regimen costing 20$ per course, MDR and XDR TB are often fatal and require 2 years for treatment with a cost that is 100 times more expensive than that of drug sensitive TB. Therefore, MDR and XDR TB can be proposed as a significant part of the global burden; it is estimated that 490,000 MDR TB cases emerged in 2006, and 40,000 of those were XDR TB (The WHO/IUATLD Global Project on Anti-Tuberculosis Drug Resistance Surveillance, 2008).
3
How the Burden is Estimated
3.1
Incidence as an Indicator of Burden
As an indicator to decide HBC, the incidence of all TB is usually calculated from the estimation of smear positive incidence in most resource-limited countries, as diagnosis of
60,644
39,459
29,899
UR Tanzania
Uganda
20,971
63,444
Mozambique
Thailand
189,323
36,553
Kenya
Brazil
86,206
Vietnam
143,221
DR Congo
Russian Federation
86,264
160,943
Pakistan
Philippines
81,021
155,991
Ethiopia
144,720
Bangladesh
48,282
Nigeria
South Africa
228,864
1,320,864
China
Indonesia
1,151,751
India
1000s
90
93
94
106
123
141
149
153
237
248
292
306
351
450
454
534
1,311
1933
1000s
Population Number
142
443
50
355
312
384
173
107
392
287
181
378
225
311
940
234
99
168
40
39
59
46
53
56
66
68
105
111
131
136
158
198
184
240
590
867
1000s
Number
62
186
31
154
135
153
77
48
173
129
82
168
101
137
382
105
45
75
100,000 pop
Per
Smear positive
Incidence
100,000 pop
Per
All forms
. Table 67-1 Epidemiological burden of TB, 2006 estimated by WHO
125
131
104
168
181
122
194
179
391
373
423
520
610
890
482
578
2,658
3,445
1000s
Number
197
624
55
561
459
334
225
125
645
432
263
641
391
615
998
253
201
299
100,000 pop
Per
All forms
Prevalence
13
24
7.6
25
26
26
20
24
51
39
55
68
70
117
105
88
201
325
1000s
Number
20
117
4.0
84
66
72
23
17
84
45
34
83
45
81
218
38
15
28
100,000 pop
Per
All forms
Mortality
11
30
12
16
18
52
5.0
3.8
9.2
0.1
0.3
6.3
0.0
9.6
44
0.6
0.3
1.2
%
Incident TB cases
HIV prev.* in
Measuring the Global Burden of Tuberculosis
67 1193
Afghanistan
6,590,088
9,157
1,915
3,100
433
570
331
2,808
7,334
42
71
74
83
1000s
Per
139
109
180
49
105
37
363
177
161
500
557
171
100,000 pop
4,068
860
1,391
194
256
165
1,203
3,265
19
31
30
37
1000s
Number
62
49
81
22
47
18
155
79
73
220
227
76
100,000 pop
Per
Smear positive
*HIV This table shows epidemiological burden of TB in 2006 by incidence, prevalence, and mortality
1,764,231
Global
887,455
European Region
Western Pacific Region
544,173
Eastern Mediterranean Region
1,721,049
899,388
American Region
South East Asian Region
773,792
African Region
4,150,313
26,088
Cambodia
High burden countries
13,228
14,197
Zimbabwe
48,379
1000s
Population Number
All forms
Incidence
1,424
3,513
4,975
478
826
398
4,234
11,899
60
94
79
82
1000s
Number
219
199
289
54
152
44
547
286
231
665
597
169
100,000 pop
Per
All forms
Prevalence
1,656
291
515
62
108
41
639
1,330
8.3
13
17
6.1
1000s
Number
25
17
30
7.0
20
4.5
83
32
32
92
131
13
100,000 pop
Per
All forms
Mortality
7.7
1.2
1.3
3.0
1.1
6.4
22
11
0.0
9.6
43
2.6
%
Incident TB cases
HIV prev.* in
67
Myanmar
. Table 67-1 (continued)
1194 Measuring the Global Burden of Tuberculosis
Measuring the Global Burden of Tuberculosis
67
smear positive TB is mostly standardized and available in most countries. Smear positive TB is said to account for 45% of all TB, even though it has been shown that this proportion varies widely in different population groups. High HIV prevalent countries are given a lower proportion with the assumption that the smear positive accounts for 35% of TB incidence among HIV positives (Corbett et al., 2003; Dye et al., 1999; WHO, 2008). Although TB incidence is theoretically measurable by a prospective study, it is practically impossible as it is necessary to have a very large population to follow up because cases of TB disease occur only in hundreds per 100,000 a year in the general population even where TB is most frequent. Moreover, because there is no suitable biomarker to detect all TB cases accurately especially in resource poor settings. Thus it is no possible to measure TB incidence directly in the general population. Some may argue that it is easy to study the incidence of smear positive tuberculosis. However, because some cases may convert to negative without treatment prior to regular examination, and because other serious cases may die very quickly, it might be difficult to catch all smear positive incident cases as smear positive TB patients (Toman, 2004). Moreover, to accurately determine the incidence of smear positive tuberculosis must take account of that fact that some patients with smear negative TB might progress to smear positive without treatment, but carrying out such as study would be unethical. It is not feasible to estimate the incidence by direct measurement. We need alternative methods to measure burden and to estimate incidence. The current estimations of the TB burden by the WHO are based principally on four different methods and their combinations: notification data with an assumption concerning the case detection rate (1); annual risk of TB infection using the Styblo ratio (2); prevalence from surveys with the assumption of an average disease duration (3); and TB mortality from vital registration data used together with case fatality rate (4). The method(s) selected for making estimates for a specific country depend(s) upon the availability of study results and quality of data. They use the following parameters: 1. 2. 3. 4.
Incidence = case notification/estimated proportion of cases detected. Incidence (smear positive) = > annual risk of infection (ARI) (%) 50/100,000. Incidence = prevalence/average duration of condition. Incidence = TB deaths/proportion of incident cases that die (case fatality rate).
We will discuss each method in the following four sections.
3.2
Estimation of Incidence Using Different Methods
Disease prevalence and tuberculin surveys have been conducted in a limited number of countries (WHO, Western Pacific Region, 2007), and most developing countries do not have reliable mortality data from vital registration that covers the whole population. Therefore, incidence for the majority of countries is estimated from notification data, routine surveillance, with assumptions of the case detection rate. Although the WHO’s estimates are the best guess of the country burden with careful review of all available information, it may cause confusion in countries as they calculate one of their key target indicators of CDR by dividing their notification by the incidence estimation that was estimated using an assumption of CDR for the country. In countries where access to medical services is good for most of the population and where functional disease surveillance exists, notification is considered to be a reliable indicator of
1195
1196
. Figure 67-1 Estimated TB incidence rate, 2006. Source: Figure 1.3 in page 20 in Global Tuberculosis Control 2008 (World Health organization 2008). This figure shows the estimated incidence rate over the world. Higher incidence rates were observed mostly in countries in Africa, South East Asia and the Former Soviet Union.
67 Measuring the Global Burden of Tuberculosis
Measuring the Global Burden of Tuberculosis
67
. Figure 67-2 Estimated global prevalence, mortality and incidence rates, 1990–2006. Figure 1.20 in Global Tuberculosis Control 2008 (World Health Organization 2008). This figure shows estimated global prevalence, mortality and incidence rates. WHO estimates of global incidence of TB per capita, the incidence rate peaked around 2003 and appears to have stabilized or begun to decline slowly, while the mortality and disease prevalence rate began to decline earlier
disease incidence. The private sector and laboratories are also included in the surveillance system in many of such countries in order to catch as many TB cases as possible. The burden of disease can be estimated correctly through the routine surveillance system in these countries, and this is ideal. However, it is obvious that this method cannot be applied in most TB endemic developing countries, where the surveillance system is weak. Not only does the TB service fail to reach every segment of the population, such as migrants, the very remote, and the urban poor, but the surveillance system often does not even cover public medical facilities outside the disease control services (such as teaching hospitals or central hospitals) nor does it cover the growing private sector. In these settings, when data from other methods is not available, it is necessary to make a best assumption of the case detection rate. A baseline case detection rate was established in a series of consultation workshops in each WHO region in 1997, country by country (Dye et al., 1999). If a country has some model or pilot areas of DOTS with improved access and quality of care, the case notification rate of those areas might be considered as a standard, assuming their case detection rate is high. However, such projects might be more likely to be implemented in very remote or hot spots where high incidence is expected or may attract clients from outside the area. Whether the results can be generalized to the entire country should be carefully discussed. While continuous efforts are essential to expand services to reach those vulnerable patients without access, it is important to make efforts to improve the current disease surveillance
1197
1198
67
Measuring the Global Burden of Tuberculosis
system. However, since it may take years to establish it in most developing countries, it is necessary to have alternative ways to estimate the burden.
3.3
Estimation of Incidence from Annual Risk of TB Infection
Although the annual risk of TB infection (ARI) available from tuberculin surveys used to be utilized as a key indicator of TB epidemiology (Cauthen et al., 2002), using ARI to estimate TB incidence is no longer recommended (Dye, 2008; Dye et al., 2008). ARI available from a tuberculin survey has been utilized to estimate TB incidence through applying Styblo’s rule: a 1% ARI was considered to be equivalent to a smear positive TB incidence of 50/100,000 (Styblo, 1985). However, the results of recent tuberculin surveys are hardly interpretable without any anti-mode to distinguish infected from non-infected both in BCG vaccinated and non-vaccinated individuals. High BCG coverage also hampers the interpretation of studies. It has been very difficult to estimate ARI from most recently conducted tuberculin surveys (Dye et al., 2008; National Center for Tuberculosis and Leprosy Control, 2005; Reider, 1995). Moreover, it has been shown that the Styblo model does not fit the epidemiological situation in many countries where surveys were recently conducted (van Leth et al., 2008; National Center for Tuberculosis and Leprosy Control, 2005). One smear positive patient may not infect as many as was estimated; circulation of the infection may be occurring in populations far from children such as in factories in urban areas; schoolchildren might be better protected from TB infection than the general population. Disease incidence previously estimated by using ARI from tuberculin surveys seems to be underestimated. The WHO and its technical task force no longer recommend using a tuberculin survey to estimate TB disease incidence. However, although the tuberculin survey is not recommended to measure the burden, it might help to estimate epidemiological trends in a country if a series of surveys can be conducted in comparable ways. > Interferon-gamma release assays (IGRAs) introduce a new technology to diagnose latent infection of TB, and results are less affected by BCG vaccination and infections with mycobacteria other than tuberculosis (Lalvani, 2007; Pai et al., 2004). They may be expected to be utilized for surveys to determine the TB infection rate in the population. However, the technical stability of handling IGRAs with a large number of samples from a survey is still a concern in resource poor settings. Observations of conversion and reversion of results of IGRAs suggest that what is being measured by tuberculin test and by IGRAs may be different (Menzies et al., 2007). Moreover, there marked limitations because of the necessity of taking blood by veno-puncture where the majority of children (those who are negative) cannot obtain any benefit from the survey examination. Therefore, infection surveys with IGRAs among children are not recommended. Development of less invasive, simpler examinations to determine infection is essential.
3.4
Estimation of Incidence from Disease Prevalence Surveys
Like incidence, prevalence is a direct measure of illness in a population. Although measuring incidence is extremely difficult, prevalence can be measured in a single population study, and a series of studies can indicate a trend. Although disease prevalence surveys were conducted
Measuring the Global Burden of Tuberculosis
67
widely until 1970, the national level data was available only from Asian countries until recently (World Health Organization, Western Pacific Region, 2007). Since active case detection has not been encouraged for decades because of questions of cost and effectiveness, prevalence surveys that use the same methodology were also suspended. However, many countries are showing interest in disease prevalence surveys to identify the TB burden, incidence, and to determine the real CDR there. However, countries should be aware that to estimate the incidence even with a quality prevalence survey is extremely difficult and may be unreliable. When smear positive disease prevalence from a survey was used to estimate incidence, prevalence was traditionally divided by two with the assumption that the average duration of a smear positive condition was 2 years. However, successful treatment greatly reduces average duration, drug resistant disease and inadequate treatment may prolong it and HIV infection may also change the average duration of disease, by rapid progression and death in advanced cases. It is necessary to provide country specific disease duration while also considering various factors such as service coverage, HIV prevalence and MDR TB prevalence. For most HBCs, it is necessary to consider six different scenarios: three kinds of patients with different treatment; patients who received DOTS, patients who received non-DOTS treatment, and those who received no treatment; and each of them with and without HIV. This process must always be applied when making such estimates as prevalence is a product of incidence and disease duration. The proportion of patients who received non-DOTS or treatment in the private sector without notification might be available from a population-based prevalence survey. However, it is difficult to estimate treatment results, or average time to smear conversion. When treatment through non-DOTS accounts for a large proportion of the patients being treated in a country, it may affect the estimation of incidence. For patients who cannot afford or do not have access to any modern treatment, a historical observation of 2 years is often given to the non-HIV positive. The WHO estimation gives 6 months to HIV positives, considering the rapid progress of disease without treatment (WHO, 2008). However, this assumption among the HIV negative patients is no longer reliable in the chemotherapy era. Characteristics of non-detected may be different from those detected; among those who do not have access to chemotherapy, some are too ill to visit the TB service and die earlier, while some stay in the community without taking action longer because they are not seriously ill. For HIV positive patients, disease duration might be longer than 6 months. Results of intensified case detection through screening cannot explain such a short duration of illness of 6 months (Cheng et al., 2008; Kimerling et al., 2002; Wood et al., 2007). Therefore, even if a very high quality prevalence study is conducted in a country, the estimated incidence for a country is not easily calculated. For the 2006 estimates, disease duration (= prevalence/incidence ratios of smear positive cases) provided to countries by the WHO varies from 0.8 to 2.4 years, with an average of 1.46 years.
3.5
Estimation of Incidence from Vital Registrations of Deaths
Mortality itself has been a popular indicator of the TB burden, and it is one of the MDGs indicators. When we know the case fatality rate among TB patients in a specific population or in a whole country through a study, the case fatality rate can be used to estimate TB incidence from TB mortality.
1199
1200
67
Measuring the Global Burden of Tuberculosis
The numbers of deaths during treatment are regularly reported to national TB programs and to the WHO. Increase of deaths during TB treatment alarmed TB programs in the 1990s so that they paid attention to HIV associated TB, before they were able to provide HIV testing to TB patients. MDR and X-DR TB also threaten TB patients with increasing treatment failure and deaths. However, in the era of short-course chemotherapy, the majority of deaths are not from TB/HIV or MDR/XDR TB; they are from those with delayed or no access to TB services, diagnosis and treatment. In countries with a working vital registration for deaths, any death should be reported with the cause(s) of death documented by a doctor’s certificate; and the police or local government even has the obligation to conduct an autopsy when the cause of death is uncertain. However, reliable vital registrations for deaths to cover sufficient populations are not available in most TB endemic countries, so the number of reported TB deaths is hardly available not only to measure the TB burden, but also to the estimate incidence. The verbal autopsy is a method that can be used to clarify the proportion of TB deaths among all deaths. It is conducted to review registered deaths to improve the accuracy of the cause of death statistics (Gajalakshmi et al., 2002; Jha et al., 2006). Some proportion of deaths by ‘‘unknown cause’’ may be reclassified as TB deaths. However, sensitivity and specificity of verbal autopsy have not been accurately assessed, and a very large sample size is necessary to conduct a study for a rare event such as TB death (Yang et al., 2006). Incidence estimation from mortality is applied only in a few countries.
3.6
Estimation of Trends in Incidence
Once the estimation of incidence has been established with the consensus of experts, estimation of change gives incidence for a given year. Notification data, data from routine country surveillance, and regional trends (trends of other countries in the surrounding area) are usually used to estimate change in a country. When there are supportive studies such as prevalence surveys, tuberculin surveys and HIV prevalence studies among the general population as well as among TB patients, those are taken into account to calibrate surveillance data.
4
Prevalence as an Independent Indicator for Millennium Development Goals
Many countries are now showing interest in conducting TB disease prevalence surveys to determine the country’s epidemiological situation more accurately in order to combat disease, while funding agencies and donors also want to determine the impact of their efforts and investment on the burden of the disease. Countries’ primary interest seems to be to estimate incidence from a prevalence study, since a key target of CDR is based on incidence. However, although a series of disease prevalence surveys may provide evidence, countries should understand that the estimation of incidence from prevalence is difficult as we discussed in previous sections in this chapter. MDGs impact targets are to halt and reverse TB incidence by 2015 and to halve prevalence and death rates by 2015 compared with the baseline of 1990 (STOP TB Partnership, 2006). Although, once again, it is questionable how accurately we can estimate a 1990 baseline, it is an important advance that a measurable indicator obtained from a community study, prevalence, has been adopted as one of the indicators of progress toward the MDGs. Prevalence, measurable
Measuring the Global Burden of Tuberculosis
67
by a prevalence survey, can be an independent indicator of TB burden, while other indicators are hardly measurable in most TB endemic countries due to the currently weak surveillance system. It also has the advantage that prevalence changes quickly in response to the expansion and/or improvement of the service that detects and treats patients more frequently and earlier; shortterm impact can be observed as well. Moreover, it is the prevalence and not the incidence that determines the probability of transmission in a community. Countries with an estimated TB prevalence of more than 100 per 100,000 are encouraged to conduct a TB prevalence survey or a series of such surveys, as these are likely to be beneficial in assessing disease burden and trends, and in optimising planning for TB control (China Tuberculosis Control Collaboration, 2004; World Health Organization, Western Pacific Region, 2007). However, it is essential to standardize survey methods to have accurate estimation and to make international comparison possible. Two major challenges here are screening strategy for bacteriological examination and the availability of culture examinations. There are also major technical limitations to diagnose extra-pulmonary TB and TB in children through community surveys. Therefore, the TB prevalence survey is recommended to target pulmonary TB among adults aged 15 or more. A guidebook on TB prevalence surveys was published by the Western Pacific Regional Office of the WHO, and it is available through the Internet (World Health Organization, Western Pacific Region, 2007). A prevalence survey is usually recommended only in countries with high prevalence, because a very large number of samples is necessary in a country in those with lower prevalence, and because countries with a high burden of TB often don’t have a notification or surveillance system with proven accuracy and completeness. Although the necessary sample size should be calculated according to established epidemiological methods, a rough idea of the necessary number of study participants for a single survey is available by answering the question: ‘‘How many participants are necessary to catch 100 target cases?’’: If a country expects a smear positive prevalence of 200/100,000, we need 50,000 participants in a survey to detect 100. However, if we would like to measure the impact of control efforts, decline of TB prevalence, by performing two surveys within some interval, at least 50–60% more participants are necessary in each survey to show a 30% decline with statistical significance.
4.1
How to Detect TB Cases in the Community
If we could provide smear, culture and chest radiography for all, theoretically, all pulmonary cases could be identified. Moreover, old TB cases with sequelae such as destroyed lung can be detected by chest radiography as a disease burden. However, it is frequently not feasible to conduct all examinations for all in a large-scale survey, considering limited resources. A typical survey in a high burden setting may require 100,000 culture examinations. High cost and the significant demand for laboratory capacity could be major constraints. Alternative approaches should be sought with current technologies depending on the availability of resources. Since a TB disease prevalence survey usually targets bacteriologically confirmed pulmonary TB cases in the community, different screening options could be considered (> Table 67-2):
Screening method I: Sputum collections from all participants for smear. Screening method II: Symptom screening by interview and sputum collection from those with suspected TB symptoms.
1201
1202
67
Measuring the Global Burden of Tuberculosis
Screening method III: Screening by chest radiography and sputum collections from those with abnormal shadows.
Screening method IV: Screening by tuberculin test, and further examinations including sputum examinations from those with positive reaction. The method of no screening (to take sputum from every participant for smear) is aimed at not missing any sputum positive patients in the community without screening (Method I) (Sebhatu et al., 2007). It might be the most economical and technically feasible way in a resource-limited setting to detect all smear positive cases in the community. Two or three sputum samples should be taken from every participant. Only smear positive cases can be detected by this method. Although the workload of the laboratory could be a limitation of this method, sputum samples can be screened by fluorescent microscope to cope with a large number of samples. Other limitations of this method are uncertainty of the quality of sputum sample collection from real TB patients and possible lower positive predictive value due to the low prevalence of bacteriologically positives. Additional examinations such as culture and/or chest radiography may be required to confirm a case. Further study is necessary prior to conducting a national prevalence survey only by Method I. Sputum collection from study participants with symptoms compatible with TB such as chronic cough for 2 or 3 weeks or more identified by interviews (Method II) had been considered a reasonable approach, because it is a standard method of screening in routine
. Table 67-2 How to detect TB cases in the community
Screening method
Smear positive
Smear negative Culture positive
TB suggestive by X-ray
Weakness
No screening (sputum from all)
N.A.
Theoretically all
Not feasible due to high lab workload
N.A.
Possible low positive predictive value
Suspected TB symptoms
Interview
30–70%
Less than 50%
N.A.
Low sensitivity
Chest X ray Chest Most abnormality radiography
Most
Possible
Cost Capacity of X ray reading, Potential for overdiagnosis
TB infection Tuberculin test
Probably most
Most, but limited in high HIV settings
N.A.
Very high positive rate in most HBCs in adults, 3 days after injection
Most
Most
N.A.
Cost, Time, Blood taking from vein
Gamma interferon assay
This table shows the different screening options to detect TB cases in the community and the weakness of each option
Measuring the Global Burden of Tuberculosis
67
case detection practice by NTP in resource-limited countries. And it has been used as a methodology of active case detection in the community. However, past prevalence surveys showed that the sensitivity of symptomatic screening of ‘‘TB suspects’’ by interview is very limited; only 30–70% of smear positive prevalent cases in the community can be identified by interviews (den Boon et al., 2006; Gopi et al., 2006; National Center for Tuberculosis and Leprosy Control, 2005). Characteristics of patients who show up at a clinic to seek a treatment and those of patients staying in the community are different. A survey with a screening methodology only by interviews is not recommended even in resource-limited settings because it underestimates the prevalence. Chest radiography is usually used as a screening method for bacteriological examinations in prevalence surveys (Method III). Sputum samples are collected from participants with a chest radiography abnormality. Chest radiography is a very sensitive examination to screen TB, but the appearance of TB in chest radiography is not specific (Koppaka and Bock, 2004). Inter-reader discordance is not negligible. Sputum examinations should be performed from those with any abnormal findings, since a significant number of bacteriologically positive patients are detected from ‘‘TB healed’’ and ‘‘disease other than TB’’ categories. Therefore, sputum collection only from ‘‘TB suspects’’ by radiographic findings underestimates the prevalence. The limitation of chest radiography screening is highlighted especially in high HIV prevalent settings; according to the experiences in an intensified TB case detection program among the HIV positive, bacteriologically positive TB patients are detected among those without having any abnormality in a chest radiograph (Bakari et al., 2008; Day et al., 2006). Method IV may be applied when a survey in children is conducted. As TB disease develops only in those infected, it is not necessary to examine those not infected. However, as already discussed, there are clear limitations to the tuberculin test both in sensitivity and specificity. Tuberculin negative often does not mean non-infected. Tuberculin negative TB is especially common among those infected with HIV (Cobelens et al., 2006). It is difficult to apply this screening method even for a survey among children in high TB burden settings. A waiting time of 3 days to read the result after tuberculin injection is also a constraint in a community survey. Recently, there has been a proposal to use new technology for screening instead of the tuberculin test; IGRAs have the potential to detect latent TB infection and TB disease with high sensitivity. However, the high cost and technical requirements of this method are constraints for a large-scale survey. Moreover, as it requires veno-puncture to collect blood, it is an invasive examination and it cannot be justified to introduce it in a prevalence survey from an ethical point of view. Considering the advantages and disadvantages of the screening methods discussed above, currently, a combination of Methods II and III is recommended for the survey in TB endemic countries. The combination of interview to ask about symptoms and chest radiography is recommended as a screening strategy; sputum samples should be collected from those with any abnormality in chest radiography and those with symptoms that make them suspects for TB regardless of chest radiographic findings (WHO, Western Pacific Region, 2007). TB patients who are ill but without clear chest radiographic abnormality can be detected by this combined strategy. Having this combined screening strategy is also expected to contribute to quality sputum sample collection. Theoretically, there may be asymptomatic TB patients without any chest radiographic abnormality that cannot be identified by this strategy. But these are expected to be very limited in proportion, and there is no way to diagnose these cases by the medical service.
1203
1204
67 4.2
Measuring the Global Burden of Tuberculosis
Culture Examinations are Recommended to Measure TB Burden and to Confirm TB
There is a debate over whether smear positive TB is enough to measure the TB burden or not. Experience shows that the proportion of smear positive cases among all those bacteriologically positive in the community detected by prevalence surveys has been declining (Hong, 1995; National Center for Tuberculosis and Leprosy Control, 2005; Tupasi et al., 1999; > Table 67-3). Smear negative and culture positive cases are more than smear positive
. Table 67-3 Whether is smear positive TB enough to measure the TB burden or not? Year
X-ray active
Bac(+)
Smear(+)
S(+)/B(+) (%)
Cambodia
2002
1,928
901
269
30
China
1990
523
177
134
76
2000
367
160
122
76
1983
2,900
860
660
77
1997
4,200
810
310
38
1965
5,100
940
690
73
1975
3,300
760
480
63
1985
2,200
480
240
50
1,000
220
90
41
Philippines
Republic of Korea
1995
(Prevalence /100,000 population) This table shows that the proportion of smear positive cases among all those bacteriologically positive in the community detected by prevalence surveys has been declining
in most surveys where culture examinations are systemically conducted. This was observed not only in recent surveys in Asia, but also in national surveys in Japan in the 1950s–1960s (Omura et al., 1962; Yamaguchi, 1955) and in the Kolin Study in Czechoslovakia where only 29% of new bacteriologically positive cases were smear positive (Styblo et al., 1967). Quality TB service, DOTS, might have removed smear positive TB cases from the community efficiently enough to decrease smear positive TB prevalence. However, if we do not measure those underlying smear negative/culture positives, we may not be able to determine the real TB burden in the community. Especially under the Stop TB strategy, since all TB cases are targeted in detection and treatment, it is recommended that culture examination should be adopted in prevalence surveys as much as possible. There are other advantages of culture examination: it can identify mycobacterium other than TB that is frequently observed in immuno-compromised individuals and those with a history of TB treatment; the drug susceptibility pattern of patients in the community is available through culture, and it may help the NTP if it does not have national surveillance data.
Measuring the Global Burden of Tuberculosis
4.3
67
High Risk Groups and Their Role in Resurgence of TB
Countries may have a concentrated epidemic of TB in some high-risk populations such as migrants, prisoners and intravenous drug users (IVDUs). However, a prevalence survey in a country often targets only residents or registered populations. When TB among those highrisk populations is considered to account for a significant proportion of the country’s prevalence, it is necessary to review the survey design if those populations are properly included in a survey. Another feasible option might be to estimate the TB prevalence and number of those high-risk populations independently by other sources and studies, and add those when we estimate country prevalence from the prevalence survey.
4.4
Prevalence Survey Brings Other Benefits
As a community survey, prevalence surveys can be used to provide more information than simply of the burden of disease. For example, some risk factors such as tobacco use and socioeconomic status, and patients’ behaviour and the utilization of the health system (for example, patient’s tendency to first seek TB treatment in the private sector) are often investigated. In addition, the sex ratio between notification and prevalence is also of interest. And, delay analysis (described immediately below) can also be conducted as part of a prevalence survey. However, since it costs 1 million dollars to conduct a typical single survey to estimate country burden, and because a prevalence survey is very labour intensive, it should be discussed whether a country can afford to increase the sample size 50–60% to detect the impact in a country, or whether we should aim for this kind of impact analysis only at subregional, regional and global levels.
5
Delay Analysis
Delay analysis of individual patients can be conducted comparatively easily and provides the information on the service quality and its use. It measures the duration from onset of the disease to starting treatment against TB and analyzes the process of the health seeking behavior of TB patients and diagnosis. Longer duration (delay) before diagnosis and treatment shows the delayed health seeking behavior of the patients (= patient delay) or the poor quality of services (= health service delay). This leads to more severe TB disease in individual patients and increased transmission in the community. As a result, delay analysis can be a useful tool to measure the TB burden in terms of time duration (= delays) of the process before diagnosis and treatment, and also changes of the delays through various interventions, though it has a limitation due to recall bias (uncertain memory about the illness by patients). The Eastern Mediterranean Region of WHO recently conducted the comparative study among seven countries, using this method (WHO/EMRO, 2006).
6
New Technology for Screening and Diagnosis to Measure the Burden
In the near future, the development of molecular bacteriology may bring a new approach to screen study participants and to diagnose TB (STOP TB Partnership Retooling Task Force,
1205
1206
67
Measuring the Global Burden of Tuberculosis
2007). Highly sensitive and non-invasive screening examination in combination with a new culture system such as liquid culture may have the potential to detect all targeted cases, smear positive, culture positive, culture negative, pulmonary, extra-pulmonary and childhood TB.
7
Conclusion
According to the WHO’s estimation of the global TB burden, there were 9.2 million new TB cases in 2006 with 1.7 million deaths due to TB. Although the estimated global incidence rate stabilized or began to fall slowly, the number of new cases was still increasing because population growth exceeded the declining rate of this disease. However, for each individual country, the uncertainty of the incidence estimation became a concern to monitor and evaluate the progress of the TB situation at country level. There are several ways to measure the burden of TB. It is essential to strengthen the routine recording and reporting system and surveillance, in order to estimate the burden and its trend. However, as the current surveillance system is not always reliable in most TB endemic countries, conducting a prevalence survey is recommended every 5–10 years. Prevalence itself is one of the MDGs indicators, and survey findings can be utilized to calibrate routine surveillance. Combinations with prevalence survey and notification data might be the best method to measure the disease burden right now. Although a prevalence survey costs a lot and it is labor intensive, it may help a country considerably to improve TB control policy. Although a prevalence study cannot be implemented in every country, studies conducted in some countries will help neighbouring countries with similar situations to understand their epidemiological status of TB through the interpretation of routine surveillance data.
8
Case Study
One Country’s efforts to have more accurate measurement of TB burden – CAMBODIA. In the National TB survey in Cambodia, 2002, BCG scar and Tuberculin surveys, presence of BCG scar and tuberculin test results were analyzed in children aged between 1 year and 15 years. 5,835 children without BCG scar and 5,886 with BCG scar were examined. The proportion of children with BCG scar is higher in the younger age group. There was no obvious anti-mode to separate populations of infected and non-infected children in any age category (> Figure 67-3), while tuberculin among TB patients provided a clear histogram (> Figure 67-4). ARI was calculated by two different methods: One was with the conventional cut off point of 10 mm; the other was with a mirror image with 16 mm obtained by a survey of tuberculin reactions among TB patients. There were significant differences between the results using these two methods. It is consequently very difficult to determine estimated ARI (> Table 67-4). If we applied Styblo’s rule to estimate the smear positive TB incidence rate with a point estimate for ARI of 2.8% obtained, the incidence rate for Cambodia would be estimated at 140 (2.8 50), and that was almost equivalent to the case notification rate in Cambodia, 141/100,000 in 2002, while the WHO’s estimated incidence at that time was 256/100,000. However, no one assumed because of this that Cambodia had achieved 100% CDR in 2002; the estimation of the TB burden by tuberculin must not have been valid. For the disease prevalence survey, 22,160 participants aged 10 or more underwent an interview and chest
Measuring the Global Burden of Tuberculosis
67
. Figure 67-3 Tuberculin reaction among children, age 5–9
. Table 67-4 Annual risk of TB infection by different methods Cut-off 10 mm Age group
16 mm mirror
Point estimate
95% Cl
Point estimate
95% Cl
1–4
0.76%
0.44–1.33
0.25%
0.08–0.85
5–9
2.34%
1.85–2.95
1.15%
0.80–1.68
10–14
3.03%
2.51–3.61
1.37%
1.02–1.84
Total
2.80%
2.37–3.28
1.32%
1.02–1.70
Source: National Center for Tuberculosis and Leprosy Control, 2005
radiograph, and those with cough for 3 weeks or more and/or any abnormality in X-ray were asked to submit two sputum specimens for smear and culture. Eighty-one smear positive and 190 smear negative/culture positive cases were detected. The prevalence of smear positive TB was 269/100,000 population, only half of the WHO’s estimation, 548/100,000, while the prevalence of bacteriologically positive TB was as high as 898/100,000. These results suggest that DOTS was working to decrease the smear positive prevalence in the community. Not only the number and rate, but also the pattern of age distribution of patients suggested that the program was working. Since 1997, the WHO had been attributing to Cambodia an increasing trend of TB incidence mostly due to the spread of HIV infection in the 1990s. However, according to the survey results, the trend was corrected from ‘‘increasing’’ to ‘‘declining,’’ following the Western Pacific Regional trend, although the estimation was still as high as 220/100,000 in 2006. The NTP learned several unexpected things from the survey: as smear positive TB made up only 30% of bacteriologically positive cases, the prevalence of all TB seemed to be higher than the WHO’s estimation; Only 62% of smear positive cases and 39% of bacteriologically positive cases complained of cough for 3 weeks or more; although females accounted for 50% of case notifications in Cambodia, the prevalence
1207
1208
67
Measuring the Global Burden of Tuberculosis
. Figure 67-4 Tuberculin reaction of TB Patients Kg Chlanang and Kg Thom Provinces, Cambodia, 2003
. Figure 67-5 Prevalence of bacteriologically positive TB by age and sex
Measuring the Global Burden of Tuberculosis
67
of smear positive in males was 2.6 times higher than that in females (> Figure 67-5). In addition, the prevalence in older persons was extremely high. These findings helped the NTP and partners to develop further plans to expand and improve TB services (National Center for Tuberculosis and Leprosy Control, 2005).
Summary Points The burden of TB can be described in various ways, such as epidemiological burden and financial burden.
Estimated disease incidence and its rate have been the most popular indicators of TB
burden since 1997 when the WHO began to publish the estimates every year. There were an estimated 9.2 million new TB cases (139/100,000 population) and 1.7 million deaths globally in 2006, according the most recent estimation by WHO. The estimated global incidence rate was reported to have stabilized or begun to fall slowly after a peak in 2003. However, as population growth exceeds the decline in incidence, the number of new cases is still increasing. For an individual country, it is often difficult to have an accurate estimation of the incidence due to limited reliability and coverage of case notification through the routine disease surveillance system. As services expand, discrepancies have been noted between estimates of incidence and notification rates in several countries. Tuberculin surveys are no longer recommended to measure TB burden, while disease prevalence surveys are promoted for countries with expected prevalence of smear positive TB of 100 or more per 100,000, in order to have more accurate figures with which to measure the TB situation. Prevalence is not only an indicator that is measurable by a community survey, but also one of the indicators of the MDGs. A series of prevalence surveys can provide a measure of trend and may show the impact of control efforts. The results of successive surveys can also be utilized to evaluate the data on trend derived from routine surveillance. It is essential for countries to make continuous efforts to improve their disease surveillance system, since it is cost efficient and an ideal way to measure the disease burden of TB as well as other diseases.
References Bakari M, Arbeit RD, Mtei L, Lyimo J, Waddell R, Matee M, Cole BF, Tvaroha S, Horsburgh CR, Soini H, Pallangyo K, von Reyn CF. (2008). BMC Infect Dis. 8: 32. Cain KP, Kanara N, Laserson KF Vannarith C, Sameourn K, Samnang K, Qualls ML, Wells CD, Varma JK. (2007). Int J Tuberc Lung Dis. 11: 1008–1013. Cauthen GM, Pio A, ten Dam HG. (2002). Bull World Health Organ. 80: 503–511. Chheng P, Tamhane A, Natpratan C, Tan V, Lay V, Sar B, Kimerling ME. (2008). Int J Tuberc Lung Dis. 12 (Suppl 1): 54–62.
China Tuberculosis Control Collaboration. (2004). Lancet. 364: 417–422. Cobelens F, Egwaga SM, van Ginkel T, Muwinge H, Matee MI, Borgdorff MW. (2006). Clin Infect Dis. 43: 634–639. Corbett EL, Watt CJ, Walker N, Maher D, Williams BG, Raviglione MC, Dye C. (2003). Arc Intern Med. 163: 1009–1021. Day JH, Charalambous S, Fielding KL, Hayes RJ, Churchyard GJ, Grant AD. (2006). Int J Tuberc Lung Dis. 10: 523–529. den Boon S, White NW, van Lill SW, Borgdorff MW, Verver S, Lombard CJ, Bateman ED, Irusen E,
1209
1210
67
Measuring the Global Burden of Tuberculosis
Enarson DA, Beyers N. (2006). Int J Tuberc Lung Dis. 10: 876–882. Dye C, Scheele S, Dolin P, Pathania V, Raviglione MC. (1999). JAMA. 282: 677–686. Dye C. (2008). Bull World Health Organ. 86: 4. Dye C, Bassil A, Bierrenbach AL, Broekmans JF, Chadha VK, Glaziou P, Gopi PG, Hosseini M, Kim SJ, Manissero D, Onozaki I, Rieder HL, Scheele S, van Leth F, van der Werf M, Williams BG. (2008). Lancet Infect Dis. 8: 233–243. Gajalakshmi V, Peto R, Kanaka S. Balasubramanian S. (2002). BMC Public Health 2: 7. Gopi PG, Subramani R, Sadacharam K, Narayanan PR. (2006). Int J Tuberc Lung Dis. 10: 343–345. Hong YP, Kim SJ, Lew WJ, Lee EK, Han YC. (1995). Int J Tuberc Lung Dis. 2: 27–36. Jha P, Gajalakshmi V, Gupta PC, Kumar R, Mony P, Dhingra N, Peto R; RGI-CGHR Prospective Study Collaborators. (2006). PLoS Med. 3(2): e18. Kimerling ME, Schuchter J, Chanthol E, Kunthy T, Stuer F, Glaziou P, Ee O. (2002). Int J Tuberc Lung Dis. 6: 988–994. Koppaka R, Bock N. (2004). In: Toman’s Tuberculosis Case Detection, Treatment and Monitoring: Questions and Answers, 2nd ed. World Health Organization, Geneva, pp. 51–60. Lalvani A. (2007). Chest. 131: 1898–1906. Menzies D, Pai M, Comstock G. (2007). Ann Intern Med. 146: 340–354. Murray CJ, DeJonghe E, Chum HJ, Nyangulu DS, Salomao A, Styblo K. (1991). Lancet. 338(8778): 1305–1308. National Center for Tuberculosis and Leprosy Control. (2005). National tuberculosis prevalence survey, 2002, Cambodia. Phnom Penh: Royal Government of Cambodia, Ministry of Health, Phnom Penh. Omura T, Oka H, Kumabe H, Kobayashi A. (1962). Bull World Health Organ. 26: 19–45. Pai M, Riley LW, Colford JM. Jr. (2004). Lancet Infect Dis. 4: 761–776. Reider HL. (1995). Tubercle Lung Dis. 76: 114–121. Sebhatu M, Kiflom B, Seyoum M, Kassim N, Negash T, Tesfazion A, Borgdorff MW, van der Werf MJ. (2007). Bull World Health Organ. 85: 593–599. STOP TB Partnership Retooling Task Force. (2007). New Technology for Tuberculosis Control: A Framework for their Adoption, Introduction and Implementation. World Health Organization, Geneva. STOP TB Partnership. (2006). The Global Plan to Stop TB, 2006–15. World Health Organization, Geneva. Styblo K, Dankova D, Drapela J, Galliova J, Jezek Z, Krivanek J, Kubik A, Langerova M, Radkovsky J. (1967). Bull World Health Organ. 37: 819–874.
Styblo K, Bumgarner JR. (1991). Tuberculosis Surveillance Research Unit Progress Report: 2: pp. 60–72. Styblo K. (1991). Epidemiology of Tuberculosis, 2nd ed. KNCV Tuberculosis Foundation, Hague, The Netherlands. Styblo K. (1985). Bull Int Union Tuberc Lung Dis. 60: 117–119. The WHO/IUATLD Global Project on Anti-Tuberculosis Drug Resistance Surveillance. (2008). Anti-Tuberculosis drug resistance in the world, Report No.4. World Health Organization, Geneva. Toman K. (2004). In: Toman’s Tuberculosis Case Detection, Treatment and Monitoring: Questions and Answers, 2nd ed. World Health Organization, Geneva, pp. 66–71. Tupasi TE, Radhakrishna S, Rivera AB, Pascual ML, Quelapio MI, Co VM, Villa ML, Beltran G, Legaspi JD, Mangubat NV, Sarol JN Jr., Reyes AC, Sarmiento A, Solon M, Solon FS, Mantala MJ. (1999). Int J Tuberc Lung Dis. 3: 471–477. van Leth F, Vander Werf MJ, Borbodorff MW. (2008). Bull World Health Organ. 86: 20–24. WHO Tuberculosis Programme. (1994). Framework for Effective Tuberculosis C control. World Health Organization, Geneva. WHO, EMRO. (2006). Diagnostic and treatment delay in tuberculosis, an in-depth analysis of the healthseeking behaviour of patients and health system response in seven countries of the Eastern Mediterranean Region. WHO, EMRO, Cairo. Wood R, Middelkoop K, Myer L, Grant A, Whitelaw AD, Lawn SD, Kaplan G, Huebner R, McIntyre J, Bekker LG. (2007). Am J Respir Crit Care Med. 175: 87–93. World Bank. (1993). World Development Report 1993: Investing in Health, Oxford University Press, New York, pp. 25–29. World Health Organization. (2008). WHO Report 2008, Global Tuberculosis Control: Surveillance, Planning, Financing. Geneva. World Health Organization, Western Pacific Region. (2007). Assessing Tuberculosis Prevalence Through Population-based Surveys. Manila. http://www.wpro.who.int/NR/rdonlyres/F49273CB-4CAB4C38-B1E3–500108BA4A97/0/AssessingTBprevalence. pdf. World Health Organization. (1994). TB: A Global Emergency, WHO Report on the TB Epidemic. World Health Organization, Geneva. Yamaguchi M. (1955). Bull World Health Organ. 13: 1041–1073. Yang G, Rao C, Ma J, Wang L, Wan X, Dubrovsky G, Lopez AD. (2006). Int J Epidemiol. 35: 741–748.
68 Burden of Tuberculosis: Serbian Perspectives Z. Gledovic . H. Vlajinac . T. Pekmezovic . S. Grujicic-Sipetic . A. Grgurevic . D. Pesut 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1212
2 The Burden of Tuberculosis in Serbia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1213 2.1 Incidence and Incidence Trends . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1213 2.2 Mortality and Mortality Trends . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1215 3
Disability Adjusted Life Years (DALY) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1217
4
Serbian Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1218 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1219
#
Springer Science+Business Media LLC 2010 (USA)
1212
68
Burden of Tuberculosis: Serbian Perspectives
Abstract: > Tuberculosis still remains the largest single infectious cause of death among young persons and adults, and continues to be a major public health problem in many parts of the world. TB > incidence in Serbia, very high in the 1950s (about 350/100,000 population), significantly decreased in subsequent years and fell below 50/100,000 in the last decade of the twentieth century. As a result of deteriorating social and economic conditions during the 1990s, TB incidence leveled off at an annual average of 34.6/100,000 in the period 1992–2002. During the same period childhood tuberculosis incidence rates in Serbia tended to fall, which could be explained by the decreasing trend in overall tuberculosis incidence and the good control program including chemoprophylaxis of young household contacts. TB mortality in Serbia showed a significant decreasing trend from the 1950s, even in the period 1992–2002 in which the average annual mortality rate was 4.1/100,000. The decrease in > DALY caused by TB in Serbia, from 1992 to 2002, was the result of a lower death rate. This means that, in addition to the improvement of primary prevention, which is very important, the improvement of secondary prevention, consisting of early diagnosis and adequate treatment, is needed to prevent premature death. The good organization and efficient work of anti-tuberculosis dispensaries in Serbia, the low frequency of HIV/AIDS, the low frequency of Mycobacterium tuberculosis resistant strains and the implementation of the > DOTS strategy explain how in the 1990s, during a period of social and economic crisis, TB incidence did not increase and TB mortality continued to drop. Thanks to the intensification of health care activities regarding TB control from 1998, the epidemiological situation of TB in Serbia is not expected to deteriorate in the future despite the anticipated inevitable increase of HIV/AIDS. List of Abbreviations: AIDS, acquired immunodeficiency syndrome; BCG, bacillus calmettegue´rin; DALY, disability adjusted life years; DOTS, directly observed therapy; GBD, global burden of disease study; GFTAM, global fund of tuberculosis, AIDS & malaria; HIV, human immunodeficiency virus; MDR-TB, multidrug resistant tuberculosis; PCR, > polymerase chain reaction; TB, tuberculosis; > YLD, years lived with disability; > YLL, years of life lost; WHO, World Health Organization
1
Introduction
Today, when the human race has reached a state of intellectual and technological sophistication, the war on the most primitive species on earth – microbes – continues to be waged. Tuberculosis is one such microbial disease that has been known since the beginning of recorded history. Robert Koch’s discovery of Mycobacterium tuberculosis in 1882 was one of the landmarks in the history of microbiology, which enabled the introduction of powerful anti-TB drugs to control this disease. Over 30 years ago, the prospect of eradicating tuberculosis seemed quite possible. Although the rates had fallen to very low levels in the developed world, the optimism of that time was short-lived. Tuberculosis remains a global emergency because of a lack of understanding of the details of its pathogenesis and the difficulty in identifying the nature of protective immune mechanisms. Our failure to control tuberculosis results from our lack of understanding of the mechanisms of latency and persistence, and the reasons for the failure of the vaccine currently available. Estimates of the protective efficacy of this vaccine vary from 0 to 80%. The apparent failure of antimicrobial chemotherapy to control the TB epidemic increases our need to understand the epidemiology of this disease. In recent years, traditional methods in epidemiology have been augmented by molecular typing methods. Genetic sequence analysis using PCR offers the potential for nonculture identification (Fletcher et al., 2001).
Burden of Tuberculosis: Serbian Perspectives
68
The natural history of tuberculosis is complex. Primary infection occurs in people without specific immunity, generally children and young adults. The initial infection can occur at any time during childhood, but adolescence is the peak time of risk. Any condition affecting the integrity of the immune system will increase the chance of infection following initial contact. The very young, patients with HIV infection, diabetic patients, those on corticisteroids, and those undergoing chemotherapy for cancer fall into this category. Poor nutrition and alcohol abuse can be added to the list. Prevalence studies from most parts of the world suggests that tuberculosis is more a disease of males than of females. The male to female ratio of TB cases reported to public health authorities worldwide is approximately 2:1(Thorson and Diwan, 2001). It is not known whether this is a genuine difference in incidence, the result of underreporting of female cases, or a combination of both. Tuberculosis has long been regarded as the disease of the poor, the homeless and the marginalized. Recent studies indicate that the relationship between poverty and disease is not straightforward. The incidence of the disease is increasing in many industrially-developed countries, particularly among the poor, ethnic minorities, prisoners and other institutionalized persons, and the socially and geographically isolated. In developing countries, millions of people are disadvantaged by poverty, and recent health sector reforms are not entirely in their interest. Ninety-five percent of all TB cases and 98% of all deaths caused by TB occur in the developing countries, especially those in Sub-Saharan Africa and parts of Asia (India, China, Indonesia). Eighty percent of all incident TB cases were found in 22 countries, with more than half the cases occurring in five South-East Asian countries. Nine of the ten countries with the highest incidence rates were in Africa. Prevalence of TB/HIV co-infection worldwide is 0.2% (Dye et al., 1999). The global burden of tuberculosis remains enormous, mainly because of poor control in South-East Asia, Sub-Saharan Africa, and Eastern Europe, and because of high rates of M. tuberculosis and HIV coinfection in some African countries (Corbett et al., 2003). The regular decline in TB notification rates has leveled off or reversed in recent years in both the USA and in Europe where relevant differences between Central/Eastern and Western countries have become more and more apparent. The main consequences of sub-optimal TB control in Eastern Europe (the Baltic states, Romania, the Russian Federation) are the export of TB and MDRTB through immigration (Migliori and Centis, 2002). Tuberculosis still remains the largest single infectious cause of death among young persons and adults and continues to be a major public health problem in many parts of the world. The World Health Organization (WHO) in 1993 declared TB a global emergency. At the beginning of the twenty first century WHO estimates that M. tuberculosis has infected about 2 billion people around the world, that is one in every three, with 8 million new cases and 2.9 million deaths annually. At least 1 million cases occur annually in children, with 450,000 tuberculosis-associated deaths (Dye et al., 1999).
2
The Burden of Tuberculosis in Serbia
2.1
Incidence and Incidence Trends
The analysis of incidence data for tuberculosis in Serbia over the period 1956–2002 was done for Central Serbia excluding the provinces of Kosovo and Vojvodina. The sources of morbidity data were the annual reports of the Institute of Lung Diseases and Tuberculosis in Belgrade,
1213
1214
68
Burden of Tuberculosis: Serbian Perspectives
which is the reference centre for diagnosis and treatment of tuberculosis in Serbia. Incidence includes active new or relapsed cases of pulmonary tuberculosis. We have assumed that these data were a good approximation of all active cases in Serbia. Population denominator data for the period 1956–2002 were obtained according to the official estimate carried out on the basis of national censuses (1953, 1961, 1971, 1981, 1991 and 2002) with extrapolation. The average population of Central Serbia in the period 1956–2002 was 5.9 million. Refugees who came to Serbia during 1990s from other parts of Former Yugoslavia as a result of the civil war were not counted as part of the population when incidence rates were calculated. Due to the great differences in incidence rates at the beginning (324.0 per 100,000) and the end (34.8 per 100,000) of the period 1956–2002 it was inappropriate to calculate average values of incidence rates (> Figure 68‐1). The increase of incidence in the first 2 years of the period studied was the result of the difficulties at the beginning of the notification process. After 1958 there is a steady decline in the incidence rates. (y = 335.964e 0.0592, F = 526.59, p = 0.000). Exponential fit is used to express the speed of decline. Bearing in mind the worsening social and economic conditions in Serbia during the last decade of the twentieth century caused by the war in Former Yugoslavia, the economic sanctions imposed on Serbia by the United Nations, and population migration, we observed the epidemiological situation of TB in the period 1992–2002 separately (Gledovic et al., 2006b). In that period the average annual tuberculosis incidence rate was 34.6 per 100,000 persons (95% CI: 33.2–35.8) for both sexes. The rate for men was 43.9 per 100,000 (95% CI: 42.1–45.7) and for women, 25.3 per 100,000 (95% CI: 24.3–24.6). These rates were near or even below the European average. Compared to other Eastern European countries, they were similar to the incidence rates in Bulgaria, two times lower than the rates in Russia, and three times lower than those in Romania (WHO, 1998). TB incidence varied in 25 geographical regions in Serbia from 7.2 per 100,000 to 74.6 per 100,000. The incidence was highest in regions with the greatest number of refugees from Bosnia and Herzegovina (near its border) where the highest TB rates were registered in Former Yugoslavia, and in the two regions adjacent to Kosovo, where, according to past data, TB incidence was higher than in other parts of Serbia (Jovanovic et al., 2007). In Serbia, TB incidence increased with age. In both sexes, the highest incidence rates were registered in those over 65 years of age (Gledovic et al., 2006b). Tuberculosis in Serbia was more frequent in males than in females, and the male to female ratio was almost 2:1, similar to the male/female ratio worldwide (Murray and Lopez, 2000). In the period 1992–2002, TB incidence rates in Serbia leveled off (y = 28.8588 + 3.3975x 0.07292x2 + 0.0449x3, F = 3.44, p = 0.081), after a long period of falling off during the previous decades and this trend has continued in the years since 2002 (Gledovic et al., 2000; Jovanovic et al., 2007). This change in the incidence trend could be expected as a result of the above-mentioned deteriorating social and economic conditions in Serbia during the 1990s (Vlajinac et al., 1997). The prevalence of HIV infection, even if we take into account that it was most probably underestimated, was not high enough to have a significant impact on TB incidence (Wong, 2002). In the period 1985–2002, there were 886 AIDS cases in Serbia, and 1,378 HIV-infected people without symptoms were reported (Gledovic et al., 2006b). In the European Community the incidence rates of tuberculosis have steadily fallen, but in several countries during the 1990s an increase in incidence was registered in foreign-born persons, and the country of birth of the TB patients is the most influential factor explaining the variability of TB incidence rates (Decludt, 2002). These cases were responsible for 51% TB
Burden of Tuberculosis: Serbian Perspectives
68
. Figure 68‐1 Tuberculosis incidence rates in Serbia in the period 1956–2002
cases in Switzerland, 41% in the Netherlands and Sweden and 38% in Denmark (Raviglione et al., 1995). In the USA, from the middle of the twentieth century, incidence rates decreased about 6% a year, but from 1985 an upward trend in incidence rates of tuberculosis due to the AIDS epidemic was observed (WHO, 1995). As tuberculosis in children reflects the prevalence of the disease in adults as well as current transmission rate, special attention was focused on TB in children from 0 to 14 years old. In the period 1992–2002, 280 children (0–14 years) in Serbia were reported as having been diagnosed with tuberculosis. The average annual incidence rate was 1.79 per 100,000 (95% CI: 0.92–3.13). Childhood tuberculosis accounts for about 5% of the reported cases of tuberculosis in Serbia (Gledovic et al., 2006a). A similar percentage was observed in France (Gaudelus, 2002) but the percentage in the United States was lower (Starke, 2004). The gender ratio of childhood TB in Serbia of 0.8:1 (129 boys and 151 girls) is similar to this ratio in other countries (Datta and Swaminathan, 2001). In Serbia, tuberculosis was more frequent in children from 5 to 14 years old (77.5%) than in those from 0 to 4 years old, which is the opposite of findings in other countries (Datta and Swaminathan, 2001). In the period 1992–2002, childhood tuberculosis incidence rates in Serbia tended to drop (y = 2.412–0.088x; P = 0.010) (> Figure 68‐2), which may be explained by the decreasing trend in overall tuberculosis incidence at the beginning of the period when notification began, and the good control program including chemoprophylaxis of young household contacts.
2.2
Mortality and Mortality Trends
TB mortality rates were calculated using the official data of the Republican Institute for Statistics. Rates are based on all tuberculosis deaths. Population denominator data for the period 1956–2002 were obtained from the official estimate carried out on the basis on national censuses (1953, 1961, 1971, 1981, 1991 and 2002) with extrapolation. In the period 1956–2002 mortality rates decreased from 76.4 per 100,000 (88.0 per 100,000 in males, 65.3 per 100,000 in females) in 1956 to lowest value of 0.9 per
1215
1216
68
Burden of Tuberculosis: Serbian Perspectives
. Figure 68‐2 Incidence trend of childhood tuberculosis in Serbia in the period 1992–2002
100,000 in 1982 (1.3 per 100,000 in males and 0.6 per 100,000 in females). After that mortality rates varied and in 1996 they were 6.4 per 100,000 in males and 2.7 per 100,000 in females. Over the entire period studied mortality rates decreased (y = 50.3745e 0.0732, F = 126.78, p = 0.000). Deviation from the steady rate of decline in 1969 may be explained by the epidemic of influenza (type A2, Hong Kong) in Serbia within the pandemic wave of influenza in the world. The slight increase of TB mortality in 1961 and 1981 was at a lower level and there is no adequate explanation for this (> Figure 68‐3). Comparing exponential fit coefficients it can be concluded that mortality rates declined faster than incidence rates in the period observed. In the period 1992–2002 the average annual mortality rate was 4.1 per 100,000 (95% CI 3.7– 4.5), 6.0 per 100,000 (95% CI; 5.4–6.6) in males and 2.3 per 100,000 (1.9–2.6) in females. The average TB mortality rate in Serbia was below the European average (Dye et al., 1999). In Serbia, mortality rates increased with age in both men and women. The highest rates were in those over 65 years of age. The male to female ratio in TB mortality rates was 3:1 (Gledovic et al., 2006b). Although it could be expected that the deteriorating performance of the economy during 1990s would have an adverse effect on TB mortality, TB mortality rates in Serbia significantly declined in the period 1992–2002 (y = 4.2920 + 0.6015x–0.1596 x2 + 0.090x3, F = 68.74, p = 0.000). The fall in the death rate from TB is continuing (Jovanovic et al., 2007). This is a . Figure 68‐3 Tuberculosis mortality rates in Serbia in the period 1956–2002
Burden of Tuberculosis: Serbian Perspectives
68
pattern similar to that already seen in many other parts of the world. TB incidence and mortality are only weakly correlated across Europe. In low-incidence countries in Western Europe, a high proportion of cases occur in elderly people who die from other causes while undergoing treatment for TB. In Eastern European countries, mortality is markedly reduced by drug therapy. The decreasing mortality trend in Serbia can be ascribed to better therapy thanks to the good organization and efficient work of anti-tuberculosis dispensaries. DOT was applied for the first time in Serbia in 1992 and its use has steadily expanded since that time. In the period observed, more than 80% of cases were successfully cured. Also the frequency of Mycobacterium tuberculosis resistant strains was low. Initial and acquired resistance to izoniazid or izoniazid and rifampicin and other drugs was 6.3% (range, 2.6–11.5%) and 5.3% (range, 2.7–8.3%) respectively (Atanaskovic-Markovic et al., 2003). In the period 1992–2002, tuberculosis was reported in 11 children, 5 boys and 6 girls. The average annual childhood tuberculosis mortality rate was 0.1 per 100,000 (95CI; 0.01–0.56). The mortality rates for both boys and girls were higher in the age group 0–4 than in the age group 5–14 (Gledovic et al., 2006a). Childhood tuberculosis mortality rates fell, although that decrease was not statistically significant. One of the reasons for the low mortality rates in children might be the low incidence rates in the youngest (0–4 years old), who are at greatest risk to develop severe and most fatal forms of disease. In the period 1992–2002 the percentage of BCG–vaccinated children was 94% on average (BCG vaccination is mandatory for infants in Serbia) (The Republic Health Institute, 2003). There is some evidence that BCG vaccination is effective against disseminated forms of the disease in childhood (Cooper et al., 2003; Datta and Swaminathan, 2001). Properly conducted short-course multi-drug treatment in Serbia is accepted as a treatment of choice for TB cases in adults as well as in children (Shingadia and Noveli, 2003).
3
Disability Adjusted Life Years (DALY)
DALY is the aggregation of years of life lost (YLL), and a year lived with disability (YLD) at the population level and thus reflects the ‘‘burden of disease’’ among the population. The method used in this study for estimating DALY was largely based on that developed for the Global Burden of Disease (GBD) study (Murray and Lopez, 1996, 2000). The average duration for TB in the region of the former socialist economies (FSE) region was assumed to be 6 months in the GBD study, and we used the 6 months duration of TB disability for the YLD calculation (the average duration of treatment was 6 months). The TB burden for the period 1992–2002, expressed as YLD, YLL, and DALY, was estimated for Central Serbia and the province of Vojvodina, which together comprised 7.7 million inhabitants on average. DALY rates per 1,000 population were greater for males than for females, and increased with ageing. The average DALY rate for males was highest in the age group from 55 to 64 years and, for females, in those aged 65 and over (> Table 68‐1). During the period observed, DALY rates significantly decreased in both males (y = 0.668– 0.021x, P = .0090 and females (y = 0.273–0.012x, P = .008), (> Figure 68‐4). The burden of tuberculosis in Serbia according to YLD, YLL and DALY was three times higher for males than for females. The total TB burden (DALY) in Serbia in the period 1992– 2002 has been decreasing thanks to decreasing TB mortality. This means that, in addition to the improvement of primary prevention, which is very important, the improvement of secondary prevention, consisting of early diagnosis and adequate treatment, is needed to prevent premature death (> Table 68‐2).
1217
1218
68
Burden of Tuberculosis: Serbian Perspectives
. Table 68‐1 YLD, YLL, and DALY per 1,000 population from TB in Serbia: average values, 1992–2002 Sex
No. cases
No. deaths
YLD
YLL
DALY
Males
1658.4
221.0
0.057
0.638
0.702
Females
1003.0
86.3
0.037
0.226
0.263
YLD years lived with disability; YLL years of life lost; DALY disability adjusted life years
. Figure 68‐4 DALY rates from TB in Serbia in the period 1992–2002, rhomb-male; quadrant-female
. Table 68‐2 Incidence, YLD, DALY and YLD/DALY ratio for tuberculosis in Serbia excluding Kosovo and Metohia, and in EURO regions Region
Incidence per 1,000
YLD per 1,000
DALY per 1,000
YLD/DALY %
Serbia
0.34
0.05
0.50
11%
EURO
0.56
0.21
1.83
11%
EUROA
0.17
0.05
0.15
33%
EURO B
0.62
0.31
2.03
15%
EURO C
1.16
0.55
4.51
12%
EURO A – Andorra, Austria, Croatia, Czech Republic, Denmark, Finland, France, Germany, Greece, Iceland, Ireland, Israel, Italy, Luxemburg, Malta, Monaco, The Netherlands, Norway, Portugal, San Marino, Slovenia, Spain, Sweden, Switzerland, the United Kingdom. EURO B (B1) – Albania, Bosnia and Herzegovina, Bulgaria, Georgia, Poland, Romania, Slovakia, The Former Yugoslav Republic of Macedonia, Turkey, Serbia and Montenegro, (B2) – Armenia, Azerbaijan, Kyrgyzstan, Tajikistan, Turkmenistan, Uzbekistan. EURO C – Belarus, Estonia, Hungary, Kazakhstan, Latvia, Lithuania, the Republic of Moldavia, the Russian Federation, Ukraine
4
Serbian Perspective
It is now widely recognized that understanding why tuberculosis remains a major killer among infectious diseases worldwide requires links between advances in biomedical knowledge and the wider social and economic dynamics of disease epidemiology.
Burden of Tuberculosis: Serbian Perspectives
68
In recent times, the emphasis has moved away from didactic principles of tuberculosis ‘‘control’’ to community- and patient-centered health services, based on analysis of local factors affecting case finding and adherence to therapy. Although WHO is urging governments to adopt its five-point DOTS strategy, there is increasing evidence that numerous social factors, differing in various communities, must receive serious consideration for the effective control of this preventable and treatable disease. In order to cope more successfully with tuberculosis in Serbia, the National TB Control Program was created in 1998. Since 2004 all activities for TB control at national level have been achieved within the framework of the project ‘‘TB control in Serbia by implementation of the DOTS strategy and outreach services for vulnerable populations,’’ which was financed by the Global Fund of Tuberculosis, AIDS & Malaria (GFATM). The main goal of the study was to reduce TB incidence in the country to 25 per 100,000 population by the end of 2009, by strengthening health care capacities for TB control, by implementation of the DOTS strategy throughout the country, by improvement of TB control in population groups at high risk for TB, and by prevention of MDR TB. Thanks to these activities, TB incidence in Central Serbia and Vojvodina fell from 37 per 100,000 in 2003 to 29 per 100,000 in 2006, and the planned implementation of the DOTS strategy was achieved. In 2007, the requisite conditions were set up for the improvement of diagnosis, treatment and follow-up of MDR TB. Thanks to these health care activities, it can be expected that the epidemiological picture of TB in Serbia will not deteriorate in the future despite the anticipated inevitable increase of HIV/AIDS.
Summary Points
After the steady decline of TB incidence rates in the period 1992–2002 leveled off. The average TB incidence rate in Serbia is below the European average. Childhood tuberculosis accounts for 5% of TB cases in Serbia. TB mortality in Serbia is decreasing significantly. The total TB burden (DALY) in Serbia in the period 1992–2002 has decreased thanks to declining mortality.
References Atanaskovic-Markovic Z, Bjegovic V, Jankovic S, Kocev N, Laaser U, Marinkovic J, et al. (2003). Burden of Disease and Injury in Serbia. Belgrade: Ministry of Health of the Republic of Serbia: 2003. Available from URL: www.sbds.sr.gov.yu. Cooper WO, Boyce TG, Wright PF, Griffin MR. (2003). Bull World Health Org. 81: 821–826. Corbett E, Watt C, Walker N, Maher D, Williams BG, Raviglione MS, Dye C. (2003). Arch Intern Med. 163: 1009–1021. Datta M, Swaminathan S. (2001). Pediatr Respir Rev. 2: 91–96.
Decludt B. (2002). Rev Prat. 52: 2106–2010. Dye C, Scheele S, Dolin P, Pathania V, Raviglione MC. (1999). JAMA. 282: 677–686. Fletcher H. (2001). Curr Opin Pulm Med. 7: 154–159. Gaudelus J. (2002). Rev Prat. 52: 2133–2138. Gledovic Z, Grgurevic A, Pekmezovic T. (2006a). Pediatr Infect Dis J. 25: 269–270. Gledovic Z, Jovanovic M, Pekmezovic T. (2000). Int J Tuberc Lung Dis. 4: 32–35. Gledovic Z,Vlajinac H, Pekmezovic T, Grujicic-Sipetic S, Grgurevic A, Pesut D. (2006b). Am J Infect Control. 34: 676–679.
1219
1220
68
Burden of Tuberculosis: Serbian Perspectives
Jovanovic D, Skodric-Trifunovic V, Markovic-Denic l, Stevic R, Vlajinac H. (2007). Int J Tuberc Lung Dis. 11: 647–651. Migliori GB, Centis R. (2002). Monaldi Arch Chest Dis. 57: 285–290. Murray CJL, Lopez AD (eds.). (1996). The Global Burden of Disease: a Comprehensive Assessment of Mortality and Disability from Diseases, Injuries and Risk Factors in 1990 and Projected to 2020. Global Burden of Diseases and Injury Series, Vol. 1, Harvard University Press, Cambridge, pp. 125–133. Murray CJL, Lopez AD. (2000). Health Econ. 9: 69–82. Raviglione MC, Snider DE, Kochi A. (1995). JAMA. 273: 220–226. Shingadia D, Noveli V. (2003). Lancet Infect Dis. 3: 624–632.
Starke JR. (2004). Semin Respir Crit Care Med. 25: 353–364. The Republic Health Institute. (2003). Belgrade Annual Report. 1992–2002. Thorson A, Diwan VK. (2001). Curr Opin Pulm Med. 7: 165–169. Vlajinac H, Marinkovic J, Kocev N, Adanja B, Pekmezovic T, Sipetic S, Jovanovic D. (1997). J Epidemiol Commun Health. 51: 106–109. Wong T. (2002). Rapid Assessment of Serbia HIV/AIDS/ STI Surveillance System. Field Investigation Report. CPHA, Belgrade. World Health Organization. (1998). Global Tuberculosis Control. WHO, Geneva, 273p. World Health Organization. (1995). Wkly Epidemiol Rec. 70: 224–236.
69 DALYs and Diarrhea R. Oria . R. Pinkerton . AAM Lima . R. L. Guerrant 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1222 2 Importance of Disability in Addition to Mortality: Definitions of Terms: DALYs, QALYs, YPLL, YLD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1223 3 Potential Importance of Long-Term Effects of Diarrhea . . . . . . . . . . . . . . . . . . . . . . . . . . . 1224 4 Relevance to Cost Effective Analyses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1227 5 Potential Evolutionary Relevance of DALYs to Human Genetics . . . . . . . . . . . . . . . . . . 1229 6 Importance to Solving one of the Greatest Health Problems Stemming from Poverty: Enteric Infections from Inadequate Water and Sanitation . . . . . . . . . 1230 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1231
#
Springer Science+Business Media LLC 2010 (USA)
1222
69
DALYs and Diarrhea
Abstract: Burden of disease assessments must incorporate both years of life lost and years lost to disability. This is especially important for diseases that have long-term effects in addition to early life mortality. Such is increasingly recognized to be the case for early childhood diarrhea, especially in impoverished areas. Although dramatic improvements have been made in reducing childhood diarrhea mortality, the morbidity (illness) rates persist, and have likely lasting effects on children’s physical and even cognitive development. Hence, even though data are limited, existing evidence suggests that these long-term effects of early childhood diarrhea on child development and thus cause potentially lifelong “disability” may be substantial. Although several approaches have been used to assess these disability and mortality impacts, including “Quality adjusted life years” and “Disability adjusted life years,” the latter (DALYs) have gained widespread use. We review this evidence and provide a range of estimates from published information on the “DALY impact” of early childhood diarrhea. List of Abbreviations: ApoE4, apolipoprotein E4; CAT-1, cationic arginine transporter 1; DALY, disability-adjusted life years; EAEC, enteroaggregative Escherichia coli; ECD, early childhood diarrhea; HAART, highly active antiretroviral therapy; HAZ, height-for-age z scores; HIV, human immunodeficiency virus; HST, Harvard step test; IL-8, interleukin-8; IQ, intelligence quotient; ORNT, oral rehydration and nutrition therapy; ORT, oral rehydration therapy; QALY, quality-adjusted life years; SD, standard deviation; TB, tuberculosis; TONI III, test of nonverbal intelligence, version III; WISC, Wechsler intelligence scale for children; YLD, years lost to disability; YPLL, years of potential life lost
1
Introduction
Assessing the importance of diseases or conditions that alter the health and well being of people necessitates a broad understanding of health as well as the ways it can be altered. Attempts at quantifying the impact of health altering conditions (and hence the “value” or worth of interventions that avoid these impacts or optimize “health”) require not only counting the causes of death, but also the ages affected (hence the “years of potential life lost,” YPLL). However, morbidity (i.e., illnesses or impaired quality of life without death) must also be counted. This requires a full understanding of “good health” that includes a sense of well being that may even include such issues as “stress” or a sense of control in one’s life, as well as freedom from overt “disease” (Bosma et al., 1997). It must also incorporate a sense of ability to achieve one’s full human potential, both physically and cognitively, and a sense of meaning, fulfillment and richness in life. This “achieving of one’s full human potential” becomes important in assessing the long-term impact of early childhood conditions such as enteric infections and their consequent symptoms of overt diarrhea or covert impairment of longterm physical and cognitive development, the latter potentially lasting for a lifetime. As premature deaths are reduced, morbidity gains much greater importance relative to mortality. Though quality of life is increasingly appreciated by those who age and lose functions and quality, the potential life-long impact of conditions that affect an entire lifetime is far greater. Thus, the “years lost to disability” (YLD) become increasingly important when assessing the impact of conditions that affect human development from early childhood. Such is the importance of early childhood intestinal infections from inadequate water and sanitation in impoverished conditions, relevant to 1.2–2.4 billion people (i.e., 17–40% of the world’s population), respectively, who lack these basic necessities (Mara, 2003).
DALYs and Diarrhea
2
69
Importance of Disability in Addition to Mortality: Definitions of Terms: DALYs, QALYs, YPLL, YLD (> Table 69-1)
An understanding of the true (or nearest approximation of) global burden of diarrheal diseases and enteric infections in pediatric populations in the developing world (especially in children under five years of age) is critical to effective planning of public health policies to prevent or diminish their effects on children’s development. Although mortality from early childhood diarrhea (ECD) has been dramatically reduced by oral rehydration therapy (ORT), the persisting poverty and lack of water and sanitation amidst the fastest growing populations (Guerrant and Blackwood, 1999) leave morbidity (illness) rates as high or higher than ever (Kosek et al., 2003). Consequently, as we understand the potentially lifelong impairment of cognitive as well as physical development, the calculation of the global burden of disease from diarrhea and intestinal infections may be far greater than previously estimated (Guerrant et al., 2002). . Table 69-1 Means of assessing burdens of disease QALYs
Quality adjusted life years
DALYs (= YPLL + YLD)
Disability adjusted life years
YPLL
Years of potential life lost
YLD
Years lost to disability
The construction of standard health measurements (and predictive models) to quantify the burden of disease (and estimate changes over time) has provided a new rationale to accomplish a more cost-effective and equitable use of health resources (allocation of health professionals, medications, vaccines, infrastructure and equipment) and for prioritizing the economic investments toward the most important public health needs and health impact. The establishment of measurable health and economic indicators, such as QALYs and DALYs, to estimate the global burden of disease, has also provided an important epidemiological tool to health officials and policy makers to design new road maps and to define time windows to optimize the most effective prevention and control measures (such as immunizations or improved water and sanitation) and treatment (such as ORT, oral rehydration and nutrition therapy (ORNT), micronutrients, macronutrients, and antimicrobials) strategies based on endemic environments and risk groups. QALYs (Quality-Adjusted Life Years) have been designed to assess individual preference to different non-fatal health outcomes that might result from a specific intervention, therefore helping to identify types of disability, impairment and handicap (Morrow and Bryant, 1995). However, their utility has been more restricted to high income countries; since more refined measurements of disability diminish in importance when juxtaposed against attempts to the account for preventable burdens of premature deaths and long-term disabilities in developing countries. In addition, the usually high cost of these quality-based studies and subjective assignments of this index reduce its feasibility to estimate quality of life in poor settings. Thus QALYs not only suffer problems of subjective value assignments that vary substantially with
1223
1224
69
DALYs and Diarrhea
who makes the choices, they also do not capture the broader benefits (externalities) that may accrue to society, family or friends (Guerrant et al., 2004). In 1995–1996, the Global Burden of Disease Programme called for DALY (DisabilityAdjusted Life Years) estimates for many diseases, which was followed subsequently by WHO, the World Bank and other agencies to rank the burden (King et al., 2005). DALYs incorporate not only age-specific mortality (as years of potential life lost (YPLL) to fatality) but also includes the reduction of the quality of life due to disabilities and handicap (e.g., years lost to disability (YLD)). In calculating DALYs, a healthy “non-disabled” state is assign as 0 and death is defined as 1; in between the different degrees of disability are found. This measurement greatly simplifies the calculation of disease burden and ensures comparability between populations. The losses from disability and premature death are combined to define the final DALY value, which allows the evaluation of short and long-term consequences on the quality of life. This kind of measurement is therefore appropriate to understand the impact of chronic and debilitating diseases that do not kill (with low mortality rates) but are either chronic or cause long-term morbidity or disability. Neuropsychiatric illnesses (such as depression, dementia, and schizophrenia), rheumatic diseases (such as arthritis and neuropathic conditions) and neurovascular diseases (such as atherosclerosis and diabetes) may have a more chronic course and thus have a greater YLD impact than that due to YPLL. These debilitating chronic conditions are far more costly over the long term due to the staggering toll of disability, hospitalization and life-long medications. The improvement of health worldwide has reduced the mortality rates, which in turn lead to increase in longevity. Aging populations, especially in the developed world, are therefore more subjected to chronic diseases or greater disease risk such as with Clostridium difficile antibioticassociated colitis, which has a distinct increased risk with increasing age and likely impairs quality of life more than currently appreciated.
3
Potential Importance of Long-Term Effects of Diarrhea
According to Murray and Lopez, (Lopez et al., 2006) diarrhea ranks second only to respiratory illnesses among the top three causes of DALYs worldwide. The global DALY for diarrheal diseases is estimated to be 99.6 million. If iron-deficiency and protein-energy malnutrition are added to this number (both are commonly seen with enteric infections), this number approximates 145 million total DALYs. For instance, malnutrition was found to be the major risk factor responsible for the greatest loss of DALYs (15.9%), together with poor water and sanitation (6.8%). Pruss et al. (2002) estimated DALYs due to diarrheal illnesses (related to poor hygiene and sanitation) to be 82,196,000 DALYs per year, with the total amount of 2,213,000 deaths worldwide (especially in sub-Sahara Africa). The lack of potable water and adequate distribution systems in the developing world clearly pose the greatest risk of water-borne enteric illnesses and diarrhea (Lee and Schwab, 2005). Although trends of declining mortality due to diarrheal illnesses have been reported in children under five years old, morbidity rates have not declined (Kosek et al., 2003; Parashar et al., 2003). However, there is considerable uncertainty in all-cause mortality in low-income countries in the sub-Saharan Africa (Lopez et al., 2006). In contrast to aging populations with chronic illnesses, who are admitted to hospitals and health centers and have medical records which contribute to population statistics, children who silently survive following repeated or persistent intestinal infections and diarrhea early in life but are never hospitalized or otherwise
DALYs and Diarrhea
69
brought to medical attention have physical and cognitive impairment or chronic disabilities that remain obscured or uncounted if not completely undiscovered. Mortality rates due to diarrhea and malnutrition in children under five years old, thus likely represent only a tip of the iceberg (Guerrant et al., 2005). Children’s developing their full potential as productive adults might be forever lost if a critical time window (the first two years of life) for brain development is compromised by debilitating and malnourishing diarrhea. Diarrhea and enteric infections might disrupt intestinal barrier function and therefore reduce the supply of critical nutrients (such us zinc, fatty acids or iron) to the developing brain, thus jeopardizing the development of normal brain plasticity. For example, zinc deficiency in Latin America, Africa and Asia, was projected to be responsible for over 453,000 deaths (4.4% of childhood deaths) and 1.2% of the burden of disease (3.8% among children between 6 mo and 5 yo) in 2004, mostly in Africa (Fischer Walker et al., 2008). The global burden of other micronutrient deficiencies and their interactions with each other are not well established. In addition, diarrhea and enteric infections may increase the risk for or worsen the effects of micronutrient deficiencies or environmental toxins such as low-dose chronic exposure to lead; for example, zinc deficiency may contribute to increased lead absorption (Levander, 1979; Oria et al., 2007). Our studies in a Brazilian cohort of shantytown children under active surveillance for diarrheal illnesses since birth suggest that heavy burdens of enteric illnesses early life (depicted by higher number and duration of diarrheal episodes in the first two years of life) are associated with long-term physical (Guerrant et al., 1999a; Moore et al., 2001) and cognitive deficits into schooling years (Guerrant et al., 1999a; Lorntz et al., 2006; Moore et al., 2000, 2001; Niehaus et al., 2002). These prolonged cognitive and growth deficits remain significantly associated with early childhood diarrhea (ECD) even after controlling for several confounding factors (such as maternal education, breast feeding and household income). Shown in > Table 69-2 are the published evidence that early childhood diarrhea and enteric parasitic infections in just the first two years of life have profound, lasting effects on child development. > Table 69-3 summarizes the magnitude of these estimates just from our studies in Northeastern Brazil, which, given the documented secular improvements in this population over time (Moore et al., 2000) probably represent a “best case” scenario. These correlate with just the average burden of diarrhea in the first two years of life alone and include an estimated 8.2 cm growth shortfall by age 7 years old, an approximate 2–10 point decrement in IQ, an impaired fitness score that equates to a 17% decrement in work productivity, and an approximate 12 month delay in schooling (Guerrant et al., 1999, 2002; Lorntz et al., 2006; Moore et al., 2001; Niehaus et al., 2002). In addition, early childhood diarrhea was the best single predictor of age at starting school and age-for-grade as well as cognitive impairment; the next best “surrogate” predictor (as it reflects ECD) of schooling was height-for-age Z scores at 2 years old (Lorntz et al., 2006). Checkley and colleagues found significant negative correlations between diarrhea episodes in the first two years of life and height-for-age z scores (HAZ) in Peruvian children living in urban slums. Children afflicted with diarrhea 10% of the time (during the first two years) were 1.5 cm shorter than children without diarrhea history (Checkley et al., 1998, 2003). Moore et al. found a mean 3.6 cm decrement in height at the age of 7 years old in children with a previous history of diarrhea (mean = 9.1 episodes) in the first 24 months in life (Moore et al., 2001). Chronic malnutrition (defined by stunting: below -2 SD under reference values) early in life may also independently affect cognitive performance during late childhood (Pinkerton et al., 2008). Although malnutrition-induced cognitive deficits during the schooling years may be ameliorated with early interventions (Tarleton et al., 2006), malnutrition combined with
1225
1226
69
DALYs and Diarrhea
. Table 69-2 Evidence for lasting disability effects from early childhood diarrhea Disease
Outcome
References
Growth shortfalls Cryptosporidial infections and persistent diarrhea
Increased diarrhea morbidity and nutritional shortfalls for up to 18 months
Agnew et al. (1998) Lima et al. (2000) Newman et al. (1999)
Cryptosporidium hominis is more common Bushen et al. than C. parvum in favela children and is (2007) associated with heavier infections and greater growth shortfalls, even in the absence of symptoms Cryptosporidial infections and 0.95–1.05 cm growth deficits at 1 year later at Figure 69-1.
4
Relevance to Cost Effective Analyses
The now growing evidence that the lasting disability effects of diarrheal infections and malnutrition early in life point necessitate a reassessment of the global DALYs attributed to these often “overlooked and neglected” diseases. The vicious cycle of malnutrition and
1227
1228
69
DALYs and Diarrhea
. Table 69-3 Estimated magnitude of the lasting disability effects from early childhood diarrhea (ECD) in just the first 2 years of life on growth, fitness, cognition and schooling 4–7 years later Lasting disability
Magnitude of disability
Reference
Growth shortfalls
Most prominent on HAZ at 2yo; but persisting to 8.2 cm difference at 7yo
Moore et al. (2001)
Fitness impairment
Approximating 17% work productivity
Guerrant et al. (1999a) Ndambe et al. (1993)
Cognitive impairment
Approximating 2–10 IQ points
Guerrant et al. (1999a) Niehaus et al. (2002)
School Approximately 12 months behind in age-for-grade by third Lorntz et al. (2006) performance grade; Also delayed age at starting school
. Figure 69-1 Vicious cycle of enteric infection and developmental shortfalls
diarrhea interact to compound the burden of both afflictions, where each aggravates or worsens the other. Recently, new calculations of the cost of chronic helmintic infections, often with anemia, diarrhea and undernutrition, suggest the necessity of changing the DALY for schistosomiasis from 0.5 to 2–15% disability (King et al., 2005). Current DALY estimates for diarrheal disease approximate 100 million, with more than 95% derived just from mortality. If only 5% of children who experienced an average of 4–8 episodes of diarrhea in their first two years of life in the lowest disability scenario, this would increase the global DALY for diarrhea by 100 million (Guerrant et al., 2002). Based on others’ and our data, we suggest herein that the disability burden of enteric illnesses associated with malnutrition likely outweighs even the still troubling mortality due to diarrhea, especially if we include growth and cognitive deficits. Furthermore, cognitive and growth deficits can be seen with intestinal infections (with protozoa and bacteria as well as with helminths) without overt symptomatic diarrhea (Checkley et al., 1998; Steiner et al., 1998). These effects likely relate to intestinal malabsorption, which has been documented with subclinical intestinal infections (Guerrant et al., 1999b, 2008). Such impaired absorptive function may also compromise nutrient and drug bioavailability, leading to increased resistance to HIV, tuberculosis,
DALYs and Diarrhea
69
and malaria therapy, which rely on long-term regimens to eradicate or control the infection (Bushen et al., 2004).
5
Potential Evolutionary Relevance of DALYs to Human Genetics
> Table
69-4 summarizes two examples of genetic alleles that appear to determine the impact of enteric infections. We have recently described the remarkable protective role played by the human genetic allele ApoE4 (that predisposes to Alzheimer’s Disease as well as to cardiovascular risk). This potentially devastating gene (in later life) may actually be in our human gene pool because it is protective against heavy diarrhea burdens in early childhood and their consequences of lasting cognitive impairment (Oria et al., 2005, 2006, 2007). Of even greater interest, understanding the cellular effects of this gene (such as the upregulation of an arginine transporter Colton et al., 2001, 2002; Czapiga and Colton, 2003) may provide clues to novel approaches to nutrient therapy. Furthermore, perinatal stressors, such as malnutrition and diarrhea may increase the risk of high cholesterol levels later in life, thus putting children with heavy burdens of malnourishing diarrhea at greater risk for cardiovascular diseases later in life (Oria et al., 2007; Victora et al., 2008). The global burden of cardiovascular diseases has been growing especially in developing countries (emergent countries such as Brazil and India), (Lawes et al., 2008) where malnourished children are more prone to obesity when life style changes (Prentice, 2006). Hence the full impact of early childhood diarrheal illnesses likely extends far beyond DALY calculations to include an evolutionary impact of the human gene pool itself. The second example is that of the proinflammatory allele at the 251 position in the human IL-8 promoter. This allele is not only associated with overt diarrheal symptoms in travelers infected with enteroaggregative E. coli (Jiang, 2003) (of potential relevance to EAEC causing inflammation and malnutrition as well Steiner et al., 1998), but also with C. difficile colitis (Jiang, 2006) and even with ulcer and carcinoma with H. pylori infections (Garza-Gonzalez, 2007). Hence, we are just beginning to appreciate the tremendous interactions of enteric infections with long-term outcomes in humans, but even with the evolution of our human genome itself.
. Table 69-4 Examples of genetic determinants of impact of enteric infections Apo E4: Associated with increased risk for Alzheimer’s and cardiovascular disease Protective against diarrhea and its cognitive impairment (Oria et al., 2005, 2007) Upregulates arginine-selective uptake via the CAT-1 (cationic aminoacid) transporter (Czapiga and Colton, 2003; Colton et al., 2001) Intestinal inflammation and the proinflammatory allele at
251 in the IL-8 promoter:
Associated with enteroaggregative E. coli diarrhea (Jiang et al., 2003) Associated with C. difficile colitis (Jiang et al., 2006) Associated with H. pylori, ulcers and gastric carcinoma (Garza-Gonzalez et al., 2007)
1229
1230
69 6
DALYs and Diarrhea
Importance to Solving one of the Greatest Health Problems Stemming from Poverty: Enteric Infections from Inadequate Water and Sanitation
In conclusion, a full and accurate assessment of the true, potentially lifelong burden of early childhood diarrhea and intestinal infections is imperative in order to place the appropriate priority on interventions that can so effectively ameliorate or avoid these burdens. Such is the importance of the DALY calculations that strive to account for years lost to disability as well as for years of life lost. Despite the greatly reduced mortality from childhood diarrhea, the illness rates continue unabated, with a potentially even greater morbidity impact when combined with food scarcity that now threatens the poorest millions of children around the world. Thus the societal (and even familial and personal) impact of impaired development for most of the world’s poorest children approaches or even exceeds that of the declining mortality from diarrheal diseases, certainly for global DALY recalculations but perhaps even for the costly familial and personal suffering that may accompany significant impairment of a child’s normal development. Shown in > Table 69-5 are the ranges of revised DALY calculations for diarrhea, using the standard formulas with age-weighting and discounting at 3%, with all disability falling into the lowest “class” (of 0.096). These calculations were published in Guerrant et al. (2002) and held mortality and morbidity in >4-year olds fixed. In the 5 scenarios shown, the first shows the original assumptions of just 1 week of “disability” for diarrheal illness 3.6 attacks per child per year. The second through fifth scenarios show how even only 5% (to 25%) of children affected by a disability “class” of 0.096 (the only one available for these formulas) for durations of 25–81.25 years (i.e., a lifetime) results in a two to six-fold increase in the global DALY calculation. The basis for this range of estimates derives from the very limited data suggesting a linear decrement in cognitive performance at 6–9 years of age from increasing diarrhea burdens in the first two years of life, the average of which approximates 2–10 IQ points with the average diarrhea burden in children living in an urban shantytown in northeast Brazil (Guerrant et al., 1999, 2002; Lorntz et al., 2006; Niehaus et al., 2002). Hence, we must see the far greater importance of adequate water and sanitation, poverty relief, and any effective interventions that can ameliorate or prevent these devastating early childhood infections and illnesses. Only with accurate assessments can such critical policy and investment decisions be made that will be so profoundly important to enabling children in greatest need to meet their full human potential.
. Table 69-5 Revised DALY calculationsa Attack rate/year at 0–4yo
Proportion disabled
Duration of durability
DALYs ( 106)
3.6
1
0.02
100.9
1
0.17
81.25
451.3
1
0.25
81.25
616.8
1
0.10
25.00
217.7
1
0.05
81.25
207.2
i.e., A long-term impact of only 1–2% more than doubles the global diarrhea DALY a
Guerrant et al. (2002)
DALYs and Diarrhea
69
Summary Points Data are increasingly showing long-term consequences of early childhood diarrhea for physical and cognitive development.
A recent multi-country analysis documents an adjusted increased odds of stunting by 1.13
(i.e., 13%) for every 5 episodes of diarrhea over the first 2 years of life (i.e., every 2.5 illnesses per child-year) (Checkley et al., 2008). Stunting at age 2 years old (like diarrhea in the first 2 years of life) predicts later cognitive impairment (Mendez and Adair 1999) and HAZ-2 may be the best predictor of human capital (Victora et al., 2008). Mounting evidence shows that intestinal infections lead to malnutrition and that malnutrition worsens intestinal infections thus documenting a “vicious cycle” for child development. (i.e., bad water collides with bad diets to impair child development and human potential) (Guerrant et al., 2008). DALY calculations that have here-to-fore focused on mortality and only acute diarrheal symptoms thus will need substantial corrections as data on long-term impact become available. Understanding the terms to define disease burden is relevant to police makers and health officials to assign short and long-term societal cost-effects of child morbidity. The reciprocal cycle of diarrhea and malnutrition amplifies the burden of diarrheal disease to a level not ever considered before. Novel scientific evidence requires the revisit of global DALYs calculation for diarrheal diseases. The understanding of evolutionary traits for human survival may highlight gene candidates for diarrheal disease susceptibility and to design preventive measures. Solving the greatest health problems requires multi-focused attention to water, sanitation, behavior, microbes, genetics and intestinal repair micronutrients and nutrients in order to reduce the diseases of poverty.
References Agnew DG, Lima AA, Newman RD, et al. (1998). J Infect Dis. 177: 754–760. Bosma H, Marmot MG, Hemingway H, Nicholson AC, Brunner E, Stansfeld SA. (1997). BMJ. 314: 558–565. Bushen OY, Davenport JA, Lima AB, et al. (2004). Clin Infect Dis. 38: 1764–1770. Bushen OY, Kohli A, Pinkerton RC, et al. (2007). Trans R Soc Trop Med Hyg. 101: 378–384. Checkley W, Buckley G, Gilman RH, Assis AMO, Guerrant RL, Morris SS, Molbak K, ValentinerBranth P, Lanata CF, Black RE and The Childhood Malnutrition and Infection Network. (2008). Int J Epidemiol. 37 (4): 816–830. Colton CA, Brown CM, Cook D, et al. (2002). Neurobiol Aging. 23: 777–785. Colton CA, Czapiga M, Snell-Callanan J, Chernyshev ON, Vitek MP. (2001). Biochim Biophys Acta. 1535: 134–144.
Checkley W, Epstein LD, Gilman RH, Black RE, Cabrera L, Sterling CR. (1998). Am J Epidemiol. 148: 497–506. Checkley W, Epstein LD, Gilman RH, Cabrera L, Black RE. (2003). Am J Epidemiol. 157: 166–175. Czapiga M, Colton CA. (2003). J Neuroimmunol. 134: 44–51. Fischer Walker CL, Ezzati M, Black RE. (2008). Feb 13. Eur J Clin Nutr. doi:10.1038/ejcn.2008. 9: 1–7 PMID:18270521. Garza-Gonzalez E, et al. (2007). BMC Cancer. 7: 70. Guerrant DI, Moore SR, Lima AA, Patrick PD, Schorling JB, Guerrant RL. (1999a). Am J Trop Med Hyg. 61: 707–713. Guerrant RL, Blackwood BL. (1999). Clin Infect Dis. 28: 966–986. Guerrant RL, Kosek M, Lima AA, Lorntz B, Guyatt HL. (2002). Trends Parasitol. 18: 191–193.
1231
1232
69
DALYs and Diarrhea
Guerrant RL, Kosek M, Moore S, Lorntz B, Brantley R, Lima AA. (2002). Arch Med Res. 33: 351–355. Guerrant RL, Lima AAM, Barboza M, et al. (1999b). Mechanisms and impact of enteric infections. Proceedings of the 2nd International Rushmore Conference on Mechanisms in the Pathogenesis of Enteric Diseases. MT. RUSHMORE, South Dakota. Guerrant RL, Lima AAM, Moore SR, Lorntz B, Patrick PD. (2004). Potential Long-Term Consequences of Early Childhood Enteric and Parasitic Infections. The Infectious Etilogy of Chronic Diseases, The National Academies Forum on Microbial Threats, Washington DC, pp. 83–92. Guerrant RL, Oria R, Bushen OY, Patrick PD, Houpt E, Lima AA. (2005). Clin Infect Dis. 41(Suppl. 8): S524–S530. Guerrant RL, Oria´ RB, Moore SR, Oria´ MO, Lima AA. (2008). Nutr Rev. 66(9): 487–505. Jiang ZD, et al. (2003). J. Infect. Dis. 188: 506–511. Jiang ZD. et al. (2006). Am. J. Gastroenterol. 101: 1112–1116. King CH, Dickman K, Tisch DJ. (2005). Lancet. 365: 1561–1569. Kosek M, Bern C, Guerrant RL. (2003). Bull World Health Organ. 81: 197–204. Lawes CM, Vander HS, Rodgers A. (2008). Lancet. 371: 1513–1518. Lee EJ, Schwab KJ. (2005). J Water Health. 3: 109–127. Levander OA. (1979). Environ Health Perspect. 29: 115–125. Lima AA, Moore SR, Barboza MS, Jr., et al. (2000). J Infect Dis. 181: 1643–1651. Lopez AD, Mathers CD, Ezzati M, Jamison DT, Murray CJ. (2006). Lancet. 367: 1747–1757. Lorntz B, Soares AM, Moore SR, et al. (2006). Pediatr Infect Dis J. 25: 513–520. Mara DD. (2003). Public Health. 117: 452–456. Moore SR, Lima AA, Conaway MR, Schorling JB, Soares AM, Guerrant RL. (2001). Int J Epidemiol. 30: 1457–1464.
Moore SR, Lima AA, Schorling JB, Barboza MS, Jr., Soares AM, Guerrant RL. (2000). Int J Infect Dis. 4: 179–186. Morrow RH, Bryant JH. (1995). Am J Public Health. 85: 1356–1360. Mendez MA, Adair LS. (1999). J Nutr. 129(8): 1555–1562. Ndamba J, Makaza N, Munjoma M, Gomo E, Kaondera KC. (1993). Ann Trop Med Parasitol. 87: 553–561. Newman RD, Sears CL, Moore SR, et al. (1999). J Infect Dis. 180: 167–175. Niehaus MD, Moore SR, Patrick PD, et al. (2002). Am J Trop Med Hyg. 66: 590–593. Oria RB, Patrick PD, Blackman JA, Lima AA, Guerrant RL. (2007). Med Hypotheses. 68: 1099–1107. Oria RB, Patrick PD, Zhang H, et al. (2005). Pediatr Res. 57: 310–316. Oria RB, Vieira CMG, Pinkerton RC, et al. (2006). Nutr Res. 26: 427–435. Patrick PD, Oria RB, Madhavan V, et al. (2005). Child Neuropsychol. 11: 233–244. Parashar UD, Bresee JS, Glass RI. (2003). Bull World Health Organ. 81: 236. Pinkerton R, Lima AAM, Moore SR, Niehaus M, Oria RB, Guerrant RL. (2008). International Research in Infectious Disease. Presented May 29, NIH, Bethesda, MD. Prentice AM. (2006). Int J Epidemiol. 35: 93–99. Pruss A, Kay D, Fewtrell L, Bartram J. (2002). Environ Health Perspect. 110: 537–542. Steiner TS, Lima AA, Nataro JP, Guerrant RL. (1998). J Infect Dis. 177: 88–96. Stephenson LS, Latham MC, Adams EJ, Kinoti SN, Pertet A. (1993). J Nutr. 123: 1036–1046. Tarleton JL, Haque R, Mondal D, Shu J, Farr BM, Petri WA. Jr. (2006). Am J Trop Med Hyg. 74: 475–481. Victora CG, Adair L, Fall C, et al. (2008). Lancet. 371: 340–357.
70 The Burden of Rotavirus Acute Gastroenteritis in Europe J. Bilcke . P. Van Damme . P. Beutels 1
Introduction: Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1234
2 2.1 2.2 2.3 2.4 2.5 2.6
Rotavirus Disease Burden in Europe: the Available Data . . . . . . . . . . . . . . . . . . . . . . . 1235 Symptomatic Cases Not Requiring Professional Medical Care . . . . . . . . . . . . . . . . . . . . . 1235 Ambulatory Care . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1235 Emergency Department Visits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1236 Hospitalizations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1236 Nosocomial Infections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1236 Deaths . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1237
3
Considerations Regarding Interpretation and Comparison of Reported Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1237 3.1 Seasonality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1239 3.2 Age . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1239 3.3 Case Definitions and Study Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1239 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1240
#
Springer Science+Business Media LLC 2010 (USA)
1234
70
The Burden of Rotavirus Acute Gastroenteritis in Europe
Abstract: > Rotavirus (RV) is the leading cause of severe > gastroenteritis in children under 5 years of age. In this chapter we will document the burden of > rotavirus gastroenteritis in the 27 countries of the European Union. Data on rotavirus disease burden is lacking for about half of the countries of the European Union. The available data for the other half show differences that do not necessarily reflect country-specific differences, but that may also be due to differences in study characteristics and quality. Based on the available literature, the annual > incidence of RV disease in the European Union ranges from 0.06–4.15% of the children aged ambulatory care, 0–2.65% of the children aged International Classification of Disease, ninth edition; NRV, nosocomial rotavirus infection; O, outpatient visit; P, pediatrician; PROTECT, Pediatric ROTavirus European CommiTee; RV, rotavirus; RVGE, rotavirus gastroenteritis
1
Introduction: Background
Rotavirus (RV, > Table 70-1) is the leading cause of severe gastroenteritis in children under 5 years of age (Parashar et al., 2003). Although it is widely accepted that by the age of five years, every child is infected at least once by RV (e.g., Dennehy, 2007; Glass et al., 2006; Parashar et al., 2003; Van Damme et al., 2006), not all of these infections are symptomatic (e.g., Fischer et al., 2002; Mata et al., 1983; Simhon et al., 1985; Velazquez et al., 1996). Only a proportion of children infected with RV suffer from diarrhea and/or vomiting, which ranges from mild episodes for which no medical care is sought, to very severe episodes which may lead, if untreated, to dehydration and death. In developing countries, a substantial number of children die due to RV (de Zoysa and Feachem, 1985; Parashar et al., 2006). In developed countries, RV related deaths are uncommon, but a substantial disease burden still exists (de Zoysa and Feachem, 1985; Parashar et al., 2006). As in temperate areas (e.g., Europe), rotavirus
. Table 70-1 Key facts of Rotaviruses Virus that causes almost half of the cases of gastroenteritis (symptoms of diarrhea and/or vomiting) in young children throughout the world After first infection, subsequent infections generally less severe Treatment: rehydration Highly contagious (oral transmission), so that improving hygiene often not sufficient to reduce spread of the disease Now, oral rotavirus vaccines are on the market, preventing disease (not infection) Some key facts about rotavirus, rotavirus disease, treatment and prevention
The Burden of Rotavirus Acute Gastroenteritis in Europe
70
gastroenteritis (RVGE) occurs in winter epidemics together with influenza and respiratory syncytial virus bronchiolitis, it often creates temporary capacity problems in the health care system (Van Damme et al., 2006).
2
Rotavirus Disease Burden in Europe: the Available Data
In this chapter we will document the burden of RVGE in the 27 countries of the European Union. More detailed information about clinical features, transmission and pathogenesis of rotavirus can be found in another chapter in this book, entitled The economic burden of rotavirus diarrhea in Taiwan. A recent study (PROTECT, 2006) documented rotavirus burden in the European Union (EU), based on a literature search for the period 1994–2005. We updated this search for the years 2006 and 2007, by doing a search with the same strings (‘‘rotavirus’’ ‘‘country name’’ and ‘‘epidemiology’’) in Pubmed. Only papers in English, French, Spanish and German were considered. The results are described below. Note that, unlike the paper from the PROTECT group (PROTECT, 2006), we do not consider clinical trials, because they often use strictly controlled study design and therefore may not be representative for what happens in the general population. Also, we restrict ourselves to original studies estimating incidences directly, and thus exclude incidences inferred indirectly from other countries (as for instance reported by Parashar et al., 2003; Soriano-Gabarro et al., 2006). The disease burden of rotavirus infection is presented hereafter in different categories, according to the type of health care sought when infected (no professional medical care, ambulatory care, emergency department visit or hospitalization). Nosocomial RV infections and RV related deaths are addressed separately.
2.1
Symptomatic Cases Not Requiring Professional Medical Care
This encompasses children who are infected by rotavirus and may show symptoms (diarrhea and/or vomiting), but for whom no professional medical care is sought. By our knowledge, there is no information on the size of this group for (any country of) the European Union. Although it is often stated that by the age of 5 every child is infected at least once by RV, no study in Europe exists about how many of them experienced a > symptomatic infection. Prospective community based studies of symptomatic RV infection performed in countries outside the EU report annual estimates in children aged Table 70-5 and multiply it by the 25 million children Table 70-6). Based on the available literature, the annual incidence of RV disease in the European Union ranges from:
0.06–4.15% of the children aged < 5 years seeking ambulatory care. 0–2.65% of the children aged < 5 years visiting emergency departments. . Table 70-6 Rotavirus disease in Europe: current data Rotavirus disease in Europe results each year in: 0.06–4.15% of children aged < 5 years seeking ambulatory care 0–2.65% of the children aged < 5 years visiting emergency departments 0.03–1.19% of the children aged < 4/5 years being hospitalized for rotavirus and 0.5–14 children per million children aged < 3/5 years that die due to RV infection Information on children experiencing disease but not seeking professional medical care is lacking Information on nosocomial infections is difficult to interpret and compare Summary of the available data on rotavirus disease in the 27 countries of Europe. Note that for about half of the 27 countries no information was available
The Burden of Rotavirus Acute Gastroenteritis in Europe
70
0.03–1.19% of the children aged < 4/5 years being hospitalized for rotavirus. 0.5–14 children per million children aged < 3/5 years that die due to RV infection. Information on children experiencing disease but not seeking professional medical care is lacking for the EU.
Studies on nosocomial RV infections are numerous, but difficult (if not impossible) to interpret and compare, because of very different methodologies and contexts.
References Bilcke J, Van Damme P, Beutels P. (2008). Med Decis Making. doi:10.1177/0272989X08324955 Bilcke J, Van Damme P, De Smet F, Hanquet G, Van Ranst M, Beutels P. (2008). Eur J Ped 167(12): 1409–1419. Bilcke J, Beutels P, De Smet F, Hanquet G, Van Ranst M, Van Damme P. (2007). KCE reports vol.54C. accessed November 6th 2008 Chandran A, Heinzen RR, Santosham M, Siberry GK. (2006). J Pediatr. 149: 441–447. Cilla G, Perez-Trallero E, Lopez-Lopategui MC, Gilsetas A, Gomariz M. (2000). Epidemiol Infect. 125: 677–683. Crowley DS, Ryan MJ, Wall PG. (1997). Commun Dis Rep CDR Rev 7: R82–R86. de Wit MA, Koopmans MP, Blij van der JF, van Duynhoven YT. (2000). Clin Infect Dis. 31: 698–704. de Wit MA, Koopmans MP, Kortbeek LM, van Leeuwen NJ, Bartelds AI, van Duynhoven YT. (2001a). Emerg Infect Dis. 7: 82–91. de Wit MA, Koopmans MP, Kortbeek LM, van Leeuwen NJ, Vinje J, van Duynhoven YT. (2001b). Clin Infect Dis. 33: 280–288. de Zoysa I, Feachem RG. (1985). Bull World Health Organ 63: 569–583. Dennehy PH. (2007). Vaccine 25: 3137–3141. Ehlken B, Laubereau B, Karmaus W, Petersen G, Rohwedder A, Forster J. (2002). Acta Paediatr. 91: 769–775. Fischer TK. (2001). Acta Paediatr. 90: 1073–1075. Fischer TK, Valentiner-Branth P, Steinsland H, Perch M, Santos G, Aaby P, Molbak K, Sommerfelt H. (2002). J Infect Dis. 186: 593–597. Fischer TK, Nielsen NM, Wohlfahrt J, Paerregaard A. (2007). Emerg Infect Dis. 13: 855–859. Floret D, Lina B, Pinchinat S, Billaud G, Ait-Belghiti F, Largeron N, Bellemin B, Trang CN, Fau C, Gaspard C, Mamoux V, Marcelon L. (2006). Eur J Pediatr. 165: 905–906. Foppa IM, Karmaus W, Ehlken B, Fruhwirth M, Heininger U, Plenge-Bonig A, Forster J. (2006). Infect Control Hosp Epidemiol. 27: 633–635.
Fourquet F, Desenclos JC, Maurage C, Baron S. (2003). Arch Pediatr. 10: 861–868. Fruhwirth M, Heininger U, Ehlken B, Petersen G, Laubereau B, Moll-Schuler I, Mutz I, Forster J. (2001a). Pediatr Infect Dis. J 20: 784–791. Fruhwirth M, Karmaus W, Moll-Schuler I, Brosl S, Mutz I. (2001b). Arch Dis Child. 84: 393–397. Gil A, Carrasco P, Jimenez R, San-Martin M, Oyaguez I, Gonzalez A. (2004). Vaccine 22: 2221–2225. Glass RI, Parashar UD, Bresee JS, Turcios R, Fischer TK, Widdowson MA, Jiang B, Gentsch JR. (2006). Lancet. 368: 323–332. Gleizes O, Desselberger U, Tatochenko V, Rodrigo C, Salman N, Mezner Z, Giaquinto C, Grimprel E. (2006). Pediatr Infect Dis J. 25: S12–S21. Hsu VP, Staat MA, Roberts N, Thieman C, Bernstein DI, Bresee J, Glass RI, Parashar UD. (2005). Pediatrics. 115: 78–82. Huet F, Largeron N, Trichard M, Miadi-Fargier H, Jasso-Mosqueda G. (2007). Vaccine. 25: 6348–6358. Jit M, Edmunds WJ. (2007). Vaccine. 25: 3971–3979. Jit M, Pebody R, Chen M, Andrews N, Edmunds WJ. (2007). Hum Vaccin. 3: 23–26. Johansen K, Bennet R, Bondesson K, Eriksson M, Hedlund KO, De Verdier Klingenberg K, Uhnoo I, Svensson L. (1999). Acta Paediatr. 88: 20–23. Koopmans M, Brown D. (1999). Acta Paediatr. 88: 14–19. Lynch M, O’Halloran F, Whyte D, Fanning S, Cryan B, Glass RI. (2001). Pediatr Infect Dis J. 20: 693–698. Mata L, Simhon A, Urrutia JJ, Kronmal RA, Fernandez R, Garcia B. (1983). J Infect Dis. 148: 452–461. Mrukowicz JZ, Krobicka B, Duplaga M, KowalskaDuplaga K, Domanski J, Szajewska H, Kantecki M, Iwanczak F, Pytrus T. (1999). Acta Paediatr. 88: 53–60. Parashar UD, Hummelman EG, Bresee JS, Miller MA, Glass RI. (2003). Emerg Infect Dis. 9: 565–572. Parashar UD, Gibson CJ, Bresse JS, Glass RI. (2006). Emerg Infect Dis. 12: 304–306. Piednoir E, Bessaci K, Bureau-Chalot F, Sabouraud P, Brodard V, Andreoletti L, Bajolet O. (2003). J Hosp Infect. 55: 190–195.
1241
1242
70
The Burden of Rotavirus Acute Gastroenteritis in Europe
PROTECT. (2006). Epidemiol Infect. 134: 908–916. Rendi-Wagner P, Kundi M, Mikolasek A, Mutz I, Zwiauer K, Wiedermann U, Vecsei A, Kollaritsch H. (2006). Wien Klin Wochenschr. 118: 280–285. Riordan FA, Quigley T. (2004). J Infect. 49: 13–16. Ryan MJ, Ramsay M, Brown D, Gay NJ, Farrington CP, Wall PG. (1996). J Infect Dis. 174 Suppl 1: S12–S18. Salo H, Ollgren J, Linna M, Sintonen H, Kilpi T. (2007). Economic evaluation of rotavirus vaccination in Finland. poster, EUPHA, Helsinki. Simhon A, Mata L, Vives M, Rivera L, Vargas S, Ramirez G, Lizano L, Catarinella G, Azofeifa J. (1985). J Infect Dis. 152: 1134–1142. Soriano-Gabarro M, Mrukowicz J, Vesikari T, Verstraeten T. (2006). Pediatr Infect Dis J. 25: S7–S11. Szucs G, Uj M, Mihaly I, Deak J. (1999). Acta Paediatr. 88: 61–65. Tran MN. (2006). Survey of the deaths due to acute gastro-enteritis avoidable with the rotavirus vaccine, France, poster, EPIET Scientific Seminar, Menorca. France, Nov 2004–Mar 2005.
Van Damme P, Wielen Van der M, Ansaldi F, Desgrandchamps D, Domingo JD, Sanchez FG, Gray J, Haditsch M, Johansen K, Lorgelly P, Lorrot M, Parez N, Reschke V, Rose M. (2006). Lancet Infect Dis. 6: 805–812. Van Damme P, Giaquinto C, Huet F, Gothefors L, Maxwell M, Wielen Van der M. (2007). J Infect Dis. 195 Suppl 1: S4–S16. Velazquez FR, Matson DO, Calva JJ, Guerrero L, Morrow AL, Carter-Campbell S, Glass RI, Estes MK, Pickering LK, Ruiz-Palacios GM. (1996). N Engl J Med. 335: 1022–1028. Vesikari T, Rautanen T, Von Bonsdorff CH. (1999). Acta Paediatr. 88: 24–30. Visser LE, Cano Portero R, Gay NJ, Martinez Navarro JF. (1999). Acta Paediatr. 88: 72–76. Zheng BJ, Lo SK, Tam JS, Lo M, Yeung CY, Ng MH. (1989). J Clin Microbiol. 27: 2083–2090.
71 The Economic Burden of Rotavirus Diarrhea: Taiwan Perspectives Kow-Tong Chen 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1244
2
Classification and Characterization of Rotavirus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1245
3
Epidemiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1245
4
Clinical Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1246
5
Transmission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1246
6
Pathogenesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1247
7
Immunity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1247
8
Treatment and Prevention . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1248
9
Cost of Illness Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1248
10 10.1 10.2 10.2.1 10.3 10.3.1 10.3.2 10.4
Rotavirus Disease and Socio-Economic Burden in Taiwan . . . . . . . . . . . . . . . . . . . Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Outpatient . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Burden of Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Inpatient . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Disease Burden . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Economic Burden . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Economic Benefits of Rotavirus Vaccination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1251 1251 1251 1252 1252 1253 1254 1255
Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1259 Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1260 Costs Calculated Among Inpatient . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1260
#
Springer Science+Business Media LLC 2010 (USA)
1244
71
The Economic Burden of Rotavirus Diarrhea: Taiwan Perspectives
Abstract: Worldwide, rotavirus is the principal cause of diarrheal illness in young children. Rotavirus > infection causes significant morbidity and mortality, often inflicting emotional distress and financial burden on the family and the society. Disease surveillance has revealed that the incidence of rotavirus infections in infants and young children is similar in both developing and developed countries. Despite so, the burden of mortality is most pronounced in the poorest countries with limited healthcare access, while morbidity is of greater significance in the developed countries. Asia’s diverse populations and economies therefore present a full spectrum of disease burden. The main drivers of rotavirus burden are direct costs (e.g., hospitalizations) and indirect costs (e.g., loss of productivity). Indirect costs are substantial in industrialized counties. The estimated annual social and medical costs due to rotavirus infections in the United States were over US $1 billion and US $264 million, respectively, demonstrating a clear emphasis of indirect over direct cost. In contrast, the estimated annual social and hospital costs for rotavirus-associated admission in Taiwan were US $13.3 million and US $10.4 million, respectively, showing the equivalence in direct and indirect cost burden. Although few countries have evaluated its economic impact or considered the potential > cost effectiveness (CE) of rotavirus > universal mass vaccination (UMV), the magnitude of rotavirus disease burden is recognized. Rotavirus vaccine could markedly alleviate the ensuing disease toll on both society and the economy but to maximize its benefits, it will be important to increase its coverage. List of Abbreviations: AGE, > acute gastroenteritis; ARSN, > Asian Rotavirus Surveillance Network; BNHI, Bureau of National Health Insurance; CE, cost effectiveness; COI, cost-of-illness; GAVI, > global alliance for vaccine and immunization; ICD-9 CM codes, International Classification of Disease, ninth edition, Clinical Modification Code (ICD-9 CM Code); LOS, > length of stay; OPD, out-patient department; ORT, > oral rehydration therapy; RT-PCR, reverse transcription-polymerase chain reaction; SH, > sentinel hospital; UMV, universal mass vaccination; WHO, World Health Organization
1
Introduction
Rotavirus is the most common cause of severe gastroenteritis and dehydration in young children in both industrialized and developing countries. A World Health Organization sponsored review of rotavirus studies found that 20–70% of all hospitalizations and 20% of deaths from diarrhea were attributable to rotavirus (de Zoysa and Feachem, 1985). Rotavirus infects all children in early life and although most first infections cause mild diarrhea, 15–20% need treatment at a clinic, and 1–3% lead to dehydration needing hospitalization (Cook et al., 1990). The virus can be identified in 15–35% of children younger than 5 years treated in outpatient settings for diarrhea and in 25–55% of those hospitalized (Jain et al., 2001). About 600,000 children die every year from rotavirus, mainly in developing countries, and this figure represents about 5% of all deaths in children younger than 5 years (Parashar et al., 2006). Mortality is great in south Asia and sub-Saharan Africa, about 1 in 200 children borne in these regions will die of rotavirus (Jain et al., 2001). Although most studies of rotavirus deaths have extrapolated numbers from global rates of diarrheal deaths, several investigators have specifically assessed risk factors, and pathological findings of the deceased clearly established rotavirus to be an important causative agent (Lynch et al., 2003).
The Economic Burden of Rotavirus Diarrhea: Taiwan Perspectives
71
Cost-of illness studies are a subset of studies measuring disease burden, years of life lost and money cost (Bloom et al., 2001). Patient, and family, providers, payers, and society perspectives have been used to estimate the economic burden of a disease. Cost estimated have been defined for national population or selected samples, and have estimated patient direct (e.g., medical cost) and indirect (e.g., loss of productivity) costs. In this chapter, we used the data on the economics of rotavirus gastroenteritis in Taiwan to illustrate the direct cost of health care services for rotavirus diarrhea, as well as indirect costs to the individual patient.
2
Classification and Characterization of Rotavirus
Rotaviruses are members of the family Reoviridae. These viruses causing acute gastroenteritis were recognized as a collective pathogen affecting humans and animals world-wide (Flewett and Woode, 1978). In 1974, Thomas Henry Flewett suggested the name rotavirus after observing that, viewed through an electron microscope, a rotavirus particle looks like a wheel (rota in Latin) (Flewett and Woode, 1978). Rotaviruses occur in five groups (serogroups A-E) but only groups A-C infect human and of multiple serotypes within each group. Group A is by far the most common with sporadic episodes due to group C, group B is limited largely to China. Rotavirus is characterized by the presence of 11 segments of double-stranded RNA surrounded by three shells (an outer capsid, inner capsid, and core) (Estes, 2001). Strains are routinely typed by their two outer capsid protein that elicit neutralizing antibodies and determine serotype, the VP7, a glycoprotein or G protein and VP4, a protease-cleaved protein (P-protein). The G- and P-protein form the basis for a dual typing system (e.g., P[8], G3), with G serotypes often determined by enzyme immunoassay and the P genotype by reverse transcription-polymerase chain reaction (RT-PCR). Only five strains are commonly detected (P[8], G1; P[4], G2; P[8], G3; P[8], G4; and P[8], G9) (Gentsch et al., 2005). Reassortment is a process by which cells infected with two viruses can yield mixed progeny with gene segments derived from each parent strain – has been used in the laboratory to develop new > vaccine (Greenberg et al., 1981). Animal strains from monkeys (rhesus RRV) (Kapikian et al., 1996), cows (Clark et al., 1996; Clements-Mann et al., 1999) and lambs (Kapikian et al., 2001) that are naturally attenuated for human beings have been cultured with human strains to yields reassortants. The selected progeny viruses contain ten genes from the animal strain, which maintain the property of attenuation, and one gene encoding the outer capsid proteins, representing the common human serotypes.
3
Epidemiology
Rotavirus infection is a single most important cause of diarrheal illness in young children in both industrialized and developing countries. Rotavirus infects all children in early life and although most first infections cause mild diarrhea, 15–20% need treatment at a clinic, and 1–3% lead to dehydration needing hospitalization (Cook et al., 1990). The virus can be identified in 15–35% of children younger than 5 years treated in outpatient settings for diarrhea and in 25–55% of those hospitalized (Kosek et al., 2003). A previous report estimated globally that about 110 million episodes of gastroenteritis require home care, 25 million require clinic visits, 2 million require hospitalizations, and 352,000–592,000 (median 440,000)
1245
1246
71
The Economic Burden of Rotavirus Diarrhea: Taiwan Perspectives
deaths of children Cost-of-illness (COI) studies are a type of economic study common in the medical literature, particular in specialist clinical journal. The aim of a COI is to evaluate the economic burden illness impose on society as a whole in terms of the consumption of health care resources and production losses. The implicit assumption was that the economic costs of illness represented the economic benefits of a health care intervention had it eradicated the illness (Tarricone, 2006). COI studies have been widely debated and its usefulness as a decision-making tool has been questioned by many health economists. Nevertheless, COIs are among the commonest economic studies in healthcare, and it is widely believed that estimating the total social cost of an illness is useful aid to policy decision making (Bloom et al., 2001). All methods of economic evaluation value both input and consequences following the same three-stop road: identify input and consequences; measure input and consequences using appropriate physical units; valuate input and consequences. Problems are encountered in all three phases. Some items are difficult to identify as some healthcare interventions have hidden or unknown costs and consequences. Not all costs and consequences can be measured in appropriate physical units as some interventions have intangible consequences such as reduction of pain. Other programs use inputs which are equally difficult to quantify such as hi-tech Know-how. COI studies can be described according to the epidemiological data used (prevalence vs. incidence approach), methods chosen to estimate the economic costs (top-down vs. bottom-up), the temporal relation between the initiation of the study and data collection (retrospective vs. prospective studies) (Koopmanschap, 1998; Rice, 1994; Tarricone, 2006). Prevalence versus incidence approach: it is required the clear description of a prevalenceor incidence-based approach to measuring resource use, cost and time frame. If the results are to be used for cost control, prevalence-based costing is appropriate; this method identifies the main components of current expenditures and forgone resource and identifies possible targets for economy. If the analysis is aimed at making decisions about which treatment or research strategy to implement, the incidence-based approach is more appropriate because it provides the basis for predictions about the likely saving from programs that reduce incidence or outcomes. Top-down versus bottom-up method: the method for capturing resource use should be described. The bottom-up strategy is evaluated on the basis of the populations used and the samples and sources of specific services used for estimates. In the bottom-up approach, the estimation of costs can be to estimate the quantity of health inputs used and to estimate the unit costs of the inputs used. The costs are then estimated by multiplying unit costs by the quantities. The advantage of bottom-up is the data collected from study subjects which provide more reliable data. The disadvantage is the data needed and available will vary with the scope of the study. The top-down strategy of aggregating population sample data is examined for data completeness and sources for mortality, morbidity, hospitalizations,
1249
1250
71
The Economic Burden of Rotavirus Diarrhea: Taiwan Perspectives
ambulatory visits, pharmaceuticals and the like. The advantages of top-down are one can avoid the risk that the sum of treatment of individual diseases, but are likely to present misallocation of costs. The disadvantage of top-down is to underestimate or over-estimate total direct costs and the exclusion of cost categories that are not included in national health care expenditures (e.g., transportations). Retrospective versus prospective studies: In retrospective COI studies, all the relevant events have already occurred when the study is initiated. Conversely, in prospective COI studies the relevant events have not already occurred when the study is initiated. The major advantage of retrospective COI is that they are less expensive and time consuming that those performed prospectively. However, retrospective COIs can only be carried out when sufficient data are available. The appropriate method for measuring the dollar value of lives lost owing to morbidity and mortality has been the subject of much controversy. The approach adopted depends on the purpose of the analysis. COI is estimated by identifying the cost-generating components and by attributing a monetary value to them. The monetary value is the opportunity cost, the value of the forgone opportunity to use in a different way those resources that are used or lost due to illness. COI studies are typically divided into two main categories: core cost which resulting directly from the illness and other related cost including non-health costs of illness. Within each category, there are direct costs (for which payments are made) and indirect costs (for which resources are lost) (Rice, 1994). Direct costs are generally estimated as product of two components: number of services and unit price or charges. Indirect costs include morbidity and mortality costs. Morbidity costs are the value of reduced or lost productivity due to the disease. Mortality costs are the product of the number of deaths from the disease and the expected value of an individual’s future earnings with gender and age taken into account. COI analysis is different from any other economic evaluation analysis because it basically does not compare costs with outcome. The main objectives are (Ernst and Hay, 1994; Rice, 1994; Tarricone, 2006): (1) To assess the economic burden of illness to society. This would give information about the amount of scarce resources consumed because of illness and, along with epidemiological data on morbidity and mortality may help ranking diseases according to global burden. (2) To identify the main cost components and their incidence over total costs. This would help health policy maker defining and/or limiting. (3) To identify the actual clinical management of illness at a national level. This would help policy makers and mangers to analyze the production function used to combine inputs and/or intermediate services to deliver the final output that could range from a single product, such as hospitalization to an entire therapeutic pattern which encompasses multiple medical services. Inefficient and/or ineffective functions can then be put forward and may represent the base for the re-engineering of the whole process in case of inefficiencies, and for the reassessment of the clinical strategy as it happens for the evidence-based medicine. Clinical guidelines may for instance be one of the final products in this case, that is whenever the identification of the clinical management of illness is judged ineffective or too diverse in the same country. (4) To explain the variability of costs. In this case, statistical analysis can be performed to check whether there is a relationship between costs variability and variables, such as those related the illness (e.g., severity), to the patient (e.g., demographic variables) or to the health care providers (e.g., teaching hospitals versus district hospitals). Estimating the burden of specific diseases by means of COI studies is done frequently. However, the usefulness of COI studies has been disputed repeatedly. Some authors argue that this type of study does not deliver relevant information for healthcare policy
The Economic Burden of Rotavirus Diarrhea: Taiwan Perspectives
71
(Shiell et al., 1987), while others are more optimistic (Hodgson, 1989). However, these results would help managers to feed the planning process with more accurate information as to the future provision of service, since knowing the cost drivers that explain – at least partially – the consumption patterns of services can be very useful when planning the provision of health care services. Socioeconomic studies have their limitations and caution must be exercised in their interpretation. The cost per unit of service as well as the value of resources used in the treatment of diarrhea is often hard to quantify with sufficient accuracy. Calculating the cost of an illness requires information on all major direct expenditures as well as indirect costs. In industrialized countries, mortality due to intestinal infections is minimal; as a result, the decision to introduce rotavirus vaccines will be based on its ability to reduce morbidity and health care costs associated with rotavirus infection. As effective rotavirus vaccines become available, policy makers will have to make decisions regarding the relative cost and benefit of vaccination, in addition to considering its clinical effectiveness. In doing so, they must systematically consider the economic burden of disease, the impact of vaccination on health and economic outcomes and the net cost of vaccination and compare the costs of vaccination to health benefits.
10
Rotavirus Disease and Socio-Economic Burden in Taiwan
10.1
Background
Taiwan government-run national health insurance program, in which enrollment is mandatory, was implemented in 1995. By 1999, approximately 96% of Taiwan’s population was covered by the program (Bureau of National Health Insurance, 2006). The national health insurance program provides comprehensive coverage, including inpatient care, ambulatory care, laboratory tests, prescription drugs, and certain nonprescription drugs, dental services, traditional Chinese medicine, and certain preventive services. A co-payment is required for ambulatory care, inpatient care, and pharmaceuticals. However, services for catastrophic diseases, child birth, and preventive health care as well as medical services are offered in specific mountain areas or on offshore islands for low-income households and veterans who are exempt from the co-payment. The Bureau of National Health Insurance (BNHI) in Taiwan maintains an extensive database on all inpatient and outpatient visits, including their medical diagnoses and medical expenditure.
10.2
Outpatient
Lu et al. (2006) extracted the medical and administrative records of all pediatric patients Table 71-3).
1253
1254
71
The Economic Burden of Rotavirus Diarrhea: Taiwan Perspectives
. Figure 71-1 Flow chart of cost estimated (From Chen et al. (2007) with permission)
10.3.2 Economic Burden Costs were calculated for the different etiological groups: only rotavirus positive (n = 1,015), only bacteria positive (n = 202), rotavirus- and bacteria- positive (n = 80), and any cause (n = 2,600). > Table 71-4 illustrates the total cost of hospitalization among the 2,600 children. The estimated total burden of medical care between April 2001 and March 2003 represents US $686,750, $195,352, $77,929 and $1,699,399 among children with only rotavirus positive, only bacteria positive, rotavirus- and bacteria-positive, and any cause, respectively. Of these, 45% were caused by rotavirus-associated diarrhea and 16% by bacteria-associated diarrhea. Further US $188,057, $52,215, $20,886 and $470,774 represents the indirect costs among children with only rotavirus positive, only bacteria positive, rotavirus- and bacteria- positive, and any cause, respectively. Thus, up to 45% (US $973,620) can be considered the total burden of rotavirus-associated diarrhea in the study population. > Table 71-5 provides a breakdown for costs at the single patient level. The estimated single burden of medical care in hospitalized children between April 2001 and March 2003 is US $676 (381), $967 (543), $975 (549) and $653 (367) among children with only rotavirus positive, only bacteria positive, rotavirus- and bacteria-positive, and any cause, respectively. Further, US $185 (145), $258 (202), $261 (205) and $181 (141) represents the indirect costs in a child with only rotavirus positive, only bacteria positive, rotavirus- and bacteria-positive, and any cause, respectively. For patients with positive rotavirus test results, the mean (SD) total social cost of a rotavirus-associated hospital admission was US $861 (484) derived from the mean (SD) total direct cost of US $676 (381) and the mean (SD) total indirect cost of US $185 (145). The mean (SD) total SH cost was US $569 (320) and the mean total family expenditure was US $294 (181). Thus, up to 43% (US $294) was paid by the family. This cost represents ~40% of the monthly salary (US $720/month) of an unskilled or service worker,
The Economic Burden of Rotavirus Diarrhea: Taiwan Perspectives
71
. Table 71-1 Demographic characteristics, clinical symptoms, and hospital of admission for 2,600 children Figure 72-3 shows, quality of life followed
. Figure 72‐3 Quality of life during a dengue illness episode (adapted from Lum et al. 2008)
1271
1272
72
Disease Burden of Dengue Fever and Dengue Hemorrhagic Fever
a U-shaped curve. A radical decrease in the quality of life was observed in all of the patients from the onset of symptoms, with the lowest point (equivalent to a loss of 60% of the equivalent of perfect health) between the third and seventh days of illness. After the lowest point, patients experienced a progressive recovery of the quality of life. The average duration of the impairment was well beyond the number of days of fever (5 and 7 days, respectively), and the quality of life was 9 days for ambulatory patients (n = 83) and 13 days for hospitalized patients (n = 124). As shown in > Figure 72‐3 ambulatory patients had, on average, a lesser decrease in their quality of life and a faster recovery to normality. Despite these recent empirical studies on quality of life, the calculation of disease burden in DALYs requires additional refinements, such as conversion from visual analogue scales (e.g., the EuroQOL’s 100-point thermometer visual scale used in these studies) to time tradeoff utilities (O’Leary et al., 1995; Salomon and Murray, 2004). Furthermore, as dengue also affects children, it is necessary to validate the ability of parents to act as a child’s proxy to quantify quality of life (Prosser et al., 2007).
2.2.2
B. Major Gaps in Dengue Case Reporting
2.2.2.1
Lack of Uniform Criteria to Report Dengue Cases to WHO
In the Americas region, dengue cases are reported to WHO stratified by severity (e.g., DF or DHF). However, in Southeast Asia and the Western Pacific regions, dengue cases are reported without classification by severity, because case ascertainment is based on hospital admission (World Health Organization (WHO), 2006d). Thus, while some countries report only severe dengue cases, others attempt to report all, and still others report only laboratory confirmed cases (World Health Organization (WHO), 2000). This lack of uniform reporting makes it difficult to perform meaningful international comparisons and aggregations. 2.2.2.2
Limited Role of Surveillance and Reporting Systems
The majority of national surveillance systems depend mainly on hospital-based reporting with individual case reporting by hospital staff whose primary responsibilities lie elsewhere. Less information is obtained from clinics, and still more limited data are reported from private sector medical practices. In Southeast Asia, for example, notification is barely enforced and the number of cases of communicable diseases, such as, the proportion of the dengue disease burden dealt with in the private sector is unknown (Zaidi et al., 2004). In the Americas, surveillance systems are also generally passive. While this system may be sufficient to detect outbreaks, the data gathering is unlikely to represent the true incidence of disease in these countries (World Health Organization (WHO), 2006a). 2.2.2.3
Underreporting of Non-fatal Dengue Cases
There are a number of factors that contribute to the underreporting of dengue. First, the time required to collect a sample and complete a report form may be a disincentive to a busy healthcare provider. Second, there is little benefit to the evaluating physician or individual patient because results will often not be available for days to weeks after the acute visit that led to the sample collection. Third, while notification of suspected dengue to public health authorities (communicable disease units) may be legally required in most of the affected countries, is it rarely enforced. Finally, surveillance systems usually have logistic and budgetary constraints and all samples may not be tested. In addition, quality of testing techniques may
Disease Burden of Dengue Fever and Dengue Hemorrhagic Fever
72
vary, calling into question the reliability of reported results. Thus, underreporting of dengue cases, and probably even of deaths due to dengue, is a major concern. Evidence of underreporting of dengue cases is evident in studies conducted in Brazil, Indonesia, Puerto Rico, and Thailand. In Belo Horizonte, Southeastern Brazil, the level of reporting of hospitalized suspected dengue cases was estimated to be 63% between 1997 and 2002 (Duarte and Franc¸a, 2006). As the cases recorded in the reporting system were the more severe, the overall case fatality rate may have been consequently overestimated. In Indonesia, the number of reported cases was compared against medical records of hospitalized DHF cases that were admitted in four major hospitals in Bandung during 1994. Only 31% of those cases were captured in the report (Chairulfatah et al., 2001). Similar underreporting was found in Puerto Rico, where only 28.4% of the hospitalized DHF cases were detected by any of the surveillance and reporting systems (Rigau-Pe´rez, 1999). In another study, the same author tried to measure the burden of dengue in Puerto Rico during 1984–1994 (Meltzer et al., 1998). To deal with the existing underreporting, it was estimated that for every dengue case reported among children, there were about 10 additional cases not reported. Among adults, it was estimated that for every case reported, 27 cases went unreported. In a recent study in Thailand, underreporting was recognized and the true number was estimated as 10-fold the number reported (Clark et al., 2005). 2.2.2.4
Misclassification in Dengue Case Reporting
Dengue cases can be misclassified as DF, DHF, and DSS at the time of diagnosis because of the lack of familiarity of medical providers with dengue as a disease or the complexity of the WHO classification system (Deen et al., 2006). This is important because the severity of dengue disease is a predictor of the use of health care services and of medical care costs (Suaya et al., 2003). Correct classification is important epidemiologically because the WHO suggests calculating the dengue case-fatality rate (CFR) by dividing the number of deaths by the number of DHF cases (World Health Organization (WHO), 1997). If classification is not uniform, CFR comparisons across countries may be misleading. In a study in Puerto Rico, only 17 DHF and 3 DSS cases were identified among 986 dengue disease hospitalizations reported through the surveillance system during 1990–1991. A review of the hospital records of those patients, however, found that 88 and 14 of them had clinical diagnosis of DHF and DSS, respectively (Rigau-Pe´rez, 1997). The reviews of medical records identified about five times more severe dengue cases than were reported to the routine surveillance system. Another review of the medical records of dengue patients during the 2002 epidemic in Taiwan, found that 71% of the patients admitted with DHF were discharged with a diagnosis of dengue fever only (Lee et al., 2006). If appropriate allocation of resources to address issues of dengue reporting is to occur, better recognition and accuracy in classification of dengue disease and improved reporting is needed. 2.2.2.5
Underreporting of Fatal Dengue Cases
Reports of dengue deaths may be intuitively assumed more accurate than reports for non-fatal dengue cases. During the 1998 dengue epidemic in Puerto Rico, there were 17,000 reported cases of dengue and 19 deaths for which dengue was confirmed or probable. For the same year, however, only five dengue deaths are shown in the WHO DengueNet, the WHO-sponsored internet-based system for the global surveillance of dengue fever and DHF (World Health Organization (WHO), 2006c). This single time-point finding indicates a four fold underreporting to WHO of laboratory-diagnosed dengue deaths. In addition, there were 31 more deaths for which dengue was initially suspected but could not be confirmed because the virus
1273
1274
72
Disease Burden of Dengue Fever and Dengue Hemorrhagic Fever
was not identified by virus isolation or immunohistochemical staining of tissue, the patient died before seroconversion, and no other explanation for the death was identified (RigauPerez et al., 2002). Had these deceased patients had a convalescent sample it is highly likely that some of them would have seroconverted and classified as dengue. Difficulties in reporting and classification that occurred in Puerto Rico are likely to occur in other dengue endemic countries. Additional factors that may further interfere with confirmation and reporting of dengue deaths to WHO might include political and economic concerns regarding the possible impact on tourism.
2.2.3
C. Major Gaps in Understanding Special Regions and Populations at Risk of Dengue
2.2.3.1
Limited Public Knowledge in Major Regions at Risk
Data on DDB is limited for dengue endemic regions that have a significant portion of the world’s population: India, China, and sub-Saharan Africa. India. One billion people (15% of the world’s population) reside in India. India’s population is twice that of Southeast Asia (the region containing Thailand, Cambodia, and Vietnam) which currently reports the most dengue-related deaths world-wide. Despite that India has many regions with comparable environmental conditions, the number of reported dengue cases and deaths is only a fraction of those reported in Southeast Asia. Additionally, an increasing number of laboratory-diagnosed dengue cases in India are secondary infections (Vijayakumar et al., 2005). Despite generally being considered an urban disease, outbreaks of dengue are increasingly reported in rural areas, increasing the population that is at risk (Dhar et al., 2006; Jamaluddain and Saxena, 1997; Kumar et al., 2001; Ratho et al., 2005; Sharma et al., 2000). Prior to 2005, dengue surveillance had been very limited in India (World Health Organization (WHO), 1994). A study concerning the dengue epidemic in Chennai in 2001 suggested that the surveillance system was unlikely to generate proper information on the epidemiology of the disease (Kabilan et al., 2005). In 2004, a WHO initiative called for promoting improvement on dengue surveillance as part of the Integrated Disease Surveillance Program in India, strengthening laboratory networking and quality assurance, and reviewing case definitions (World Health Organization (WHO), 2003). In 2005 dengue case reporting became mandatory. Although improvements are being made, the national surveillance system is still passive, with only 12,317 dengue cases and 184 dengue deaths provisionally reported in 2006 (National Vector Borne Disease Control Programme (NVBDCP), 2008). These figures represent a dengue fever incidence of 1 per 100,000, a significantly lower rate than reported in Southeast Asian countries. The current gaps in epidemiologic data and surveillance in India make the dengue disease burden in India uncertain. Despite this, dengue is one of the listed causes of death and hospitalization among children in India (Vector Control Research Center ICoMR, 2006). China. One billion three hundred million people (20% of the world’s population) live in China. Roughly one fifth of China’s land mass, including some of the more densely populated regions, lies in tropical climes where dengue transmission is known to occur. Published reports on dengue outbreaks detailed the reemergence of dengue in the 1980s and 1990s (Fan et al., 1989; Qiu et al., 1991, 1993). However, since 2003, data from WHO does not include cases in China, (World Health Organization (WHO), 2006c) making the documentation of DDB in this country difficult.
Disease Burden of Dengue Fever and Dengue Hemorrhagic Fever
72
Sub-Saharan Africa. The DDB in Africa remains poorly defined. Travelers visiting and military personnel stationed in Africa have been identified with laboratory-diagnosed dengue infections, indicating that the virus is circulating (Centers for Disease Control and Prevention (CDC), 2005; Sharp et al., 1995). Several seroprevalence and fever studies conducted in sub-Saharan Africa have identified past evidence of the dengue virus transmission in many sub-Saharan countries including Senegal, Kenya, Cameroon, Djibouti, Sudan and Burkino Faso. These studies reported lower seroprevalence rates, 10–34%, than those seen in other tropical countries, such Haiti, Brazil, and Thailand (Halstead et al., 2001; Teixeira et al., 2002; Tuntaprasart et al., 2003). As expected, higher prevalence levels are noted among urban residents compared to rural residents (Collenberg et al., 2006; Ibrahim et al., 2002; Kuniholm et al., 2006; Morrill et al., 1991; Rodier et al., 1996; Saluzzo et al., 1986). In addition periodic outbreaks of dengue fever have been reported in the region (Ibrahim et al., 2002; Johnson et al., 1982; Rodier et al., 1996; Saluzzo et al., 1986). Without systematic surveillance and serosurveys with appropriate sample schemes to give a fair representation of the disease burden in the population, current DDB in Africa may remain poorly defined. Moreover if dengue is an endemic problem in sub-Saharan Africa, increasing urbanization will increase the burden of disease as large populations are placed at risk. 2.2.3.2
Limited Knowledge About Dengue in Travelers
According to the World Tourism Organization, 125.4 million international tourists visited countries where they may be at risk for acquiring dengue infection (World Tourism Organization, 2005). Depending on the population studied and the laboratory methods used, serological evidence of recent dengue infection was found in 7–45% of febrile travelers returning from endemic areas (Lo´pez-Ve´lez et al., 1996; Schwartz et al., 1996; Yabe et al., 1996), confirming that dengue is an important cause of fever in this group. Because it explains a substantial proportion of fevers in returning travelers, dengue may have a significant economic impact in non-dengue endemic countries due to days lost from work. However, given the spectrum of clinical illness, not all patients may seek medical attention or receive diagnostic testing. As a result, underreporting of dengue infection occurs even in developed countries. Moreover, not all travelers diagnosed with dengue may be reported to public health authorities. In the US where dengue fever reporting is not required nationally, less than 100 patients with laboratory-diagnosed dengue are reported to the US Centers for Disease Control and Prevention each year. However, we can estimate the level of underreporting using the available literature. During the period of 1 Jan 2001 through 31 December 2004, seven residents of the United States were diagnosed with dengue after returning from Thailand (Centers for Disease Control and Prevention (CDC), 2005). According to the World Tourism Organization, 2,012,077 US tourists visited Thailand during the same time period, (World Tourism Organization, 2006) giving a rate of 3.5 dengue infections per 1 million visitors to Thailand. However, among a prospective cohort of Dutch travelers, 0.7% of travelers from Southeast Asia experienced symptomatic, laboratory-diagnosed (anti-dengue IgM seroconversion) dengue infections (Cobelens et al., 2002). If the risk of acquiring dengue infection is similar for Dutch and US travelers visiting Southeast Asia, for each dengue case reported to the US Centers for Disease Control and Prevention there may be 5,000 additional unreported but symptomatic dengue cases. This is, however, a conservative estimate since the US, unlike Holland, shares a border with a dengue endemic country, Mexico. Personnel deployed in dengue endemic areas during humanitarian emergencies and conflicts are at a higher risk of dengue infection than regular travelers, since they usually
1275
1276
72
Disease Burden of Dengue Fever and Dengue Hemorrhagic Fever
live in areas with higher transmission rates and usually stay in those areas longer than a tourist. For example, during a 5-month deployment as part of the United Nations Mission in Haiti, 32% of 249 personnel with febrile illness had dengue infection (Gambel et al., 1999).
3
Conclusions
Existing knowledge about global dengue disease burden is limited. Although data for every health condition are imperfect, the constraints on data are particularly important for conditions such as dengue for which morbidity is relatively important compared to mortality and for which confirmatory diagnostic tests are difficult, expensive and not widely available. Registration data for deaths, while imperfect, usually tend to be better than those for morbidity. Also, censuses provide a check on overall death rates not available for morbidity data. Because dengue occurs in epidemics and treatment requires highly skilled medical personnel, the health system may become overloaded. As a result, treatment of other conditions requiring the same personnel and facilities may be adversely affected. Though population-based studies with active surveillance are the gold standard for burden estimations, their large cost and limited generalizability to other populations, settings and time, make alternative methods necessary. Several complementary study designs may provide reliable estimates of disease burden in dengue endemic countries. These include: (1) capturerecapture studies from incomplete, but independent reports of dengue cases; (2) estimating expansion factors from both populations under active surveillance and comparisons between data in medical records and aggregate statistics from representative institutions to correct for the level of misreporting to public health authorities; (3) treatment patterns from outbreak estimations; and (4) laboratory analyses of cases and deaths classified under other cases that could have included dengue. Reliable data obtained from these different approaches could provide the foundation for modeling to correct for under-diagnosis, misdiagnosis, underreporting and other limitations of existing dengue incidence. Well-planned standardized protocols for dengue disease burden studies in multiple countries could provide more complete and comparable data across countries, as was recently noted for rotavirus (World Health Organization (WHO), 2006b). Dengue disease burden estimates will facilitate cost-effectiveness analyses of new diagnostic, preventive, and therapeutic technologies.
Summary Points
Dengue is a viral acute disease transmitted by a mosquito. Dengue is a growing public health problem in tropical and sub-tropical regions. The affected countries have a current population in excess of three billion. Quantifying the epidemiological and economic burden of dengue is key to formulating policy decisions on research priorities, prevention programs, clinical training for management of the disease, and the introduction of new technologies such as vaccines, vector control, diagnostics or drugs. There is limited knowledge of global dengue disease burden. There are a number of major challenges in measuring its disease burden.
Disease Burden of Dengue Fever and Dengue Hemorrhagic Fever
72
The challenges include: lack of uniform application of the WHO case definition, limited
access to or standardization of dengue diagnostic tests, misdiagnosis, lack of systematic reporting of dengue cases to WHO, limited surveillance and reliable reporting systems, and limited public awareness in endemic regions and incidence of infection among travelers. Underreporting of dengue cases (both fatal and non-fatal) is probably the most important barrier to obtaining accurate assessments.
Acknowledgments The scientific paper (Suaya et al., 2007b), which served as a background document for this chapter was supported in part by the Pediatric Dengue Vaccine Initiative (PDVI). The preparation of this chapter was supported by the endowment of the Schneider Institutes for Health Policy, Brandeis University. For comments and assistance, the authors are indebted to Chrisann Newransky, MS, Clare Hurley, MM, Rana Sughayyar, MS, and Binod Sah, MD (Brandeis University, U.S.A.), Lucy Lum, MD (University Malaya, Malaysia), Bill Letson, MD (PDVI, Korea) and Axel Kroeger, MD (Special Programme for Research and Training in Tropical Diseases, (TDR), WHO Switzerland).
References Anderson KB, Chunsuttiwat S, Nisalak A, et al. (2007). Lancet. 369: 1452–1459. Armien B, Suaya JA, Quiroz E, et al. (2008). Am J Trop Med Hyg. 79(3): 364–371. Balmaseda A, Hammond SN, Pe´rez MA, et al. (2005). Am J Trop Med Hyg. 73: 1059–1062. Beatty ME, Hunsperger E, Long, et al. (2007). Emerg Infect Dis. 13(2): 308–310. Blacksell SD, Doust JA, Newton PN, Peacock SJ, Day NP, Dondorp AM. (2006). Trans R Soc Trop Med Hyg. 100: 775–784. Cattand P, Desjeux P, Guzma´n MG, Jannin J, Kroeger A, et al. (2006). Tropical Diseases Lacking Adequate Control Measures: Dengue, Leishmaniasis, and African Trypanosomiasis, Chap. 23. Disease Control Priorities in Developing Countries, 2nd ed. Oxford University Press, New York, pp. 451–466. Centers for Disease Control and Prevention (CDC). (2001). MMWR Morb Mortal Wkly Rep. 50: 57–59. Centers for Disease Control and Prevention (CDC). (2005). MMWR Morb Mortal Wkly Rep. 54: 556–558. Chairulfatah A, Setiabudi D, Agoes R, van Sprundel M, Colebunders R. (2001). Acta Trop. 80: 111–115. Clark DV, Mammen MP Jr, Nisalak A, Puthimethee V. (2005). Am J Trop Med Hyg. 72: 786–791. Cobelens FG, Groen J, Osterhaus AD, LeentvaarKuipers A, Wertheim-van Dillen PM, Kager PA. (2002). Trop Med Int Health. 7: 331–338.
Collenberg E, Ouedraogo T, Ganame´ J, Fickenscher H, Kynast-Wolf G, et al. (2006). J Med Virol. 78: 683–692. Deen JL, Harris E, Wills B, et al. (2006). Lancet. 368: 170–173. Dhar S, Malakar R, Ghosh A, Kundu R, Mukhopadhyay M, Banerjee R. (2006). Indian J Dermatol. 51: 57–59. Donoso Mantke O, Lemmer K, Biel SS, et al. (2004). J Clin Virol. 29: 105–112. Duarte HH, Franc¸a EB. (2006). Rev Saude Publica. 40: 134–142. Epidemic and Pandemic Alert and Response (EPR), World Health Organization (WHO). (2006). Chikungunya and Dengue in the South West Indian Ocean. http://www.who.int/csr/don/2006_03_17/ en/index.html. Accessed 30 May 2008. Epidemic and Pandemic Alert and Response (EPR), World Health Organization (WHO). (2008). Dengue/ dengue Haemorrhagic Fever in Brazil. http://www. who.int/csr/don/2008_04_10/en/index.html. Accessed 30 May 2008. EuroQOL Group. (1990). Health Policy. 16: 199–208. Fan WF, Yu SR, Cosgriff TM. (1989). Rev Infect Dis. 11 (Suppl. 4): S847–853. Gambel JM, Drabick JJ, Swalko MA, Henchal EA, Rossi CA, Martinez-Lopez L. (1999). Mil Med. 164: 300–302. Gubler DJ, Meltzer M. (1999). Adv Virus Res. 53: 35–70.
1277
1278
72
Disease Burden of Dengue Fever and Dengue Hemorrhagic Fever
Guzma´n MG, Kourı´ G. (2004). Int J Infect Dis. 8: 69–80. Guzma´n MG, Pelegrino JL, Pumariega T, et al. (2003). Rev Panam Salud Publica. 14: 371–376. Halstead SB. (1988). Science. 239: 476–481. Halstead SB. (1997). In: Gubler DJ, Kuno G (eds.) Dengue and Dengue Hemorrhagic Fever. CAB International, Wallingford, UK, pp 23–44. Halstead SB. (1999). Lancet. 353: 1100–1101. Halstead SB. (2005). Emerg Infect Dis. 11: 740–741. Halstead SB, Streit TG, Lafontant JG, et al. (2001). Am J Trop Med Hyg. 65: 180–183. Ibrahim SA, Mustafa OM, Mukhtar MM, et al. (2002). Trop Med Int Health. 7: 442–449. International Federation of Red Cross and Red Crescent Societies. (2007). Cambodia: Dengue Fever. Information Bulletin. July 10, 2007. http://www.ifrc.org/ docs/appeals/rpts07/KHdg07100701. pdf Accessed 30 May 2008. Jamaluddain M, Saxena VK. (1997). J Commun Dis. 29: 169–170. Johnson BK, Ocheng D, Gichogo A, et al. (1982). East Afr Med J. 59: 781–784. Kabilan L, Balasubramanian S, Keshava SM, Satyanarayana K. (2005). Indian J Pediatr. 72: 919–923. Kumar A, Sharma SK, Padbidri VS, Thakare JP, Jain DC, Datta KK. (2001). J Commun Dis. 33: 274–281. Kuniholm MH, Wolfe ND, Huang CYH, et al. (2006). Am J Trop Med Hyg. 74: 1078–1083. LaRocque RC, Breiman RF, Ari MD, et al. (2005). Emerg Infect Dis. 11: 766–769. Lee MS, Hwang KP, Chen TC, Lu PL, Chen TP. (2006). J Microbiol Immunol Infect. 39: 121–129. Levett PN, Branch SL, Edwards CN. (2000). Am J Trop Med Hyg. 62: 112–114. Lo´pez-Ve´lez R, Pe´rez-Casas C, Vorndam AV, Rigau J. (1996). Eur J Clin Microbiol Infect Dis. 15: 823–826. Lum LCS, Suaya JA, Tan LH, Sah BK, Shepard DS. (2008). Am J Trop Med Hyg. 78(6): 862–867. Malik EM, Eltahir HG, Ahmed ES. (2005). East Mediterr Health J. 11: 753–761. Meltzer MI, Rigau-Pe´rez JG, Clark GG, Reiter P, Gubler DJ. (1998). Am J Trop Med Hyg. 59: 265–271. Morrill JC, Johnson BK, Hyams C, et al. (1991). J Trop Med Hyg. 94: 166–168. Murray CJ. (1994). Bull World Health Organ. 72(3): 429–445. National Vector Borne Disease Control Programme (NVBDCP). (2008). Directorate General of Health Services, Ministry of Health and Family Welfare, Government of India. Dengue Cases and Deaths Since 2001. http://www.nvbdcp.gov.in/dengueC&D. html. Accessed 26 Sept 2007. O’Leary JF, Fairclough DL, Jankowski MK, Weeks JC. (1995). Med Decis Making. 15: 132–137.
Pediatric Dengue Vaccine Initiative (PDVI). (2006). PDVI diagnostic programs. http://www.pdvi.org/Newsletter_ files/PDVIDiagnostics.htm. Accessed 26 Sept 2006. Phuong HL, de Vries PJ, Nga TT, Giao PT, Hungle Q, et al. (2006). BMC Infect Dis. 6: 123. Prosser LA, Hammitt J, Keren R. (2007). Pharmacoeconomics. 25: 713–726. Qiu FX, Chen QQ, Ho QY, Chen WZ, Zhao ZG, Zhao BW. (1991). Am J Trop Med Hyg. 44: 364–370. Qiu FX, Gubler DJ, Liu JC, Chen QQ. (1993). Bull World Health Organ. 71: 349–359. Ratho RK, Mishra B, Kaur J, Kakkar N, Sharma K. (2005). Indian J Med Sci. 59: 519–527. Rigau-Pe´rez JG. (1997). Rev Panam Salud Publica. 1: 381–388. Rigau-Pe´rez JG. (1999). P R Health Sci J. 18: 337–345. Rigau-Perez JG. (2006). Lancet Infect Dis. 6: 297–302. Rigau-Perez JG, Ayala-Lopez A, Garcia-Rivera EJ, et al. (2002). Am J Trop Med Hyg. 67: 355–362. Rigau-Pe´rez JG, Gubler DJ. (1999). Lancet. 353: 1101. Rodier GR, Gubler DJ, Cope SE, et al. (1996). Trans R Soc Trop Med Hyg. 90: 237–240. Salomon JA, Murray CJL. (2004). Health Econ. 13: 281–290. Saluzzo JF, Cornet M, Castagnet P, Rey C, Digoutte JP. (1986). Trans R Soc Trop Med Hyg. 80: 5. Schwartz E, Mendelson E, Sidi Y. (1996). Am J Med. 101: 516–520. Sharma SN, Raina VK, Kumar A. (2000). J Commun Dis. 32: 175–179. Sharp TW, Wallace MR, Hayes CG, et al. (1995). Am J Trop Med Hyg. 53: 89–94. Shepard DS, Suaya JA, Halstead SB, et al. (2004). Vaccine. 22: 1275–1280. Suaya JA, Chai PF, Shepard DS, Lum L. (2003). Resource Utilization of Hospitalized Dengue Infected Children in Malaysia. Paper presented at the Annual Meeting, Academy for Health Services Research and Health Policy, TN, USA, June 2003. Suaya JA, Shepard D, Armien B, et al. (2007a). Am J Trop Med Hyg. 77: 9. Suaya JA, Shepard DS, Beatty ME. (2007b). Dengue burden of disease and cost of illness. In: Scientific Working Group: Report on Dengue. World Health Organization on behalf of the Special Programme for Research and Training in Tropical Diseases. Geneva, pp. 35–49. TDR/SWG/08. Teixeira MG, Barreto ML, Costa MCN, Ferreira LDA, Vasconcelos PVC, Cairncross S. (2002). Tro Med Int Health. 7: 757–762. Tuntaprasart W, Barbazan P, Nitatpattana N, Rongsriyam Y, Yoksan S, Gonzalez JP. (2003). Southeast Asian J Trop Med Public Health. 34: 564–568.
Disease Burden of Dengue Fever and Dengue Hemorrhagic Fever Vector Control Research Center ICoMR. (2006). Dengue and Dengue Hemorrhagic Fever. http://www.pon. nic.in/vcrc/deng.html. Accessed 15 October 2006. Vijayakumar TS, Chandy S, Sathish N, Abraham M, Abraham P, Sridharan G. (2005). Indian J Med Res. 121: 100–107. World Health Organization (WHO). (1994). Division of Communicable Diseases Partnership for Child Development. International Conference on Dengue Haemorrhagic Fever. Report of WHO/ICMR/DBT meeting and National Brainstorming Session on Dengue, National Institute of Virology. World Health Organization, Geneva, p. 7. WHO/CDS/ MIM/94.1. World Health Organization (WHO). (1997). Dengue Haemorrhagic Fever: Diagnosis, Treatment, Prevention and Control, 2nd ed. WHO, Geneva. World Health Organization (WHO). (2000). Dengue and Dengue Hemorrhagic Fever. Global Surveillance of Epidemic Prone Infectious Diseases. WHO, Geneva. World Health Organization (WHO). (2006a). DengueNet Implementation in the Americas. Report of a WHO/PAHO/CDC Meeting, San Juan, Puerto Rico, 9–11 July 2002. http://www.who.int/csr/resources/ publications/dengue/WHO_CDS_CSR_GAR_2003_8/ en/index.html. Accessed 26 Sept 2006. World Health Organization (WHO). (2003). Communicable Diseases and Disease Surveillance Dengue Report. Proceedings of the Meetings on Vector Borne Disease Surveillance with Special Focus on DengueNet at Delhi and Bangalore Organized by NICD in Collaboration with WR India, CSR/ Hq. SEARO under USAID/WHO/Hq.
72
World Health Organization WHO/WPRO/SEARO. (2004). Wkly Epidemiol Rec. 79(6): 57–62. http:// www.who.int/wer/2004/en/wer7906.pdf. World Health Organization (WHO). (2006b). Wkly Epidemiol Rec. 81: 349–356. World Health Organization (WHO). (2006c). DengueNet http://www.who.int/csr/disease/dengue/denguenet/en/index.html. Accessed 25 Sept 2006. World Health Organization (WHO). (2002). Dengue and Dengue Haemorrhagic Fever. Fact sheet N 117. April 2002; http://www.who.int/mediacentre/factsheets/fs117/en/print.html. Accessed 26 Sept 2006. World Health Organization (WHO). (2006d). Tropical Diseases Research (TDR). Strategic Direction for Research: Dengue. http://www.who.int/tdr/diseases/ dengue/direction.htm. Accessed 26 Sept 2006. World Health Organization (WHO). (2008). World Health Survey Instruments and Related Documents. World Health Organization, Geneva. http://www. who.int/healthinfo/survey/instruments/en/index.html. Accessed 15 Jan 2008. World Tourism Organization. (2005). Tourism Highlights. 2005. World Tourism Organization. (2006). Compendium of Tourism Statistics. www.world-tourism.org/statistics/index.htm. Accessed 25 Sept 2006. Yabe S, Nakayama M, Yamada K, et al. (1996). Kansenshogaku Zasshi. 70: 1160–1169. Zaidi AKM, Awasthi S, DeSilva HJ. (2004). BMJ. 328: 811–815. Zeckhauser R, Shepard D. (1976). Law Comtemp Probl. 40(4): 5–45.
1279
73 Cost-Effectiveness of a Dengue Vaccine in Southeast Asia and Panama: Preliminary Estimates D. S. Shepard . J. A. Suaya 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1282 2 Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1283 3 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1289 4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1291 5 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1293 Summary points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1294
#
Springer Science+Business Media LLC 2010 (USA)
1282
73
Cost-Effectiveness of a Dengue Vaccine in Southeast Asia and Panama: Preliminary Estimates
Abstract: To ascertain the economic feasibility of a pediatric tetravalent dengue vaccine, we developed and calibrated a cost-effectiveness model of vaccinating children at 15 months in Southeast Asia and Panama using a societal perspective. We assumed that full immunization would require two doses. In Southeast Asia, the gross cost per vaccinee would be US $8.28. Due to projected savings in dengue treatment, the net cost per capita would be 89% below the gross cost. The cost per disability adjusted life year (DALY) saved by a pediatric vaccine would be $50, making the potential vaccine highly cost-effective. In Panama, where we assumed a higher price per dose ($5), the gross cost per vaccinee would be $19.70 for infants and $26.70 in a catch up program for adults. As projected treatment costs are higher, however, averaging $154.27 when discounted to the time of vaccination, the estimated cost offsets would be 178% of vaccination costs for infants and 138% for adults. Thus the dengue vaccination appears to be cost saving under our most likely assumptions. Many uncertainties remain, however, until more epidemiological and economic data are obtained and vaccine development proceeds further. List of Abbreviations: CS, cost saving; DALYS, disability adjusted life years; DF, dengue fever; DHF, dengue hemorrhagic fever; DPT3, diphtheria pertussis tetanus (dose 3); DSS, dengue shock syndrome; GAVI, Global Alliance for Vaccines and Immunization; GNI, gross national income; HBV, hepatitis B vaccine; HIB, hemophilus influenza B; PAHO, Pan American Health Organization; SE Asia, Southeast Asia; WHO, World Health Organization
1
Introduction
Dengue is a mosquito-borne disease which threatens half the world’s population (Strobel and Lamaury, 2001). Epidemiological publications and surveillance reports have documented the sobering increase in cases and deaths from dengue (Strobel and Lamaury, 2001) and its spread to new areas, such as the Middle East and the US states of Texas (Gubler, 1998), (Centers for Disease Control, 2002) and Hawaii (Centers for Disease Control, 2002), and Paraguay (BBC, 2007). In the Americas, major outbreaks occurred in 2002 and 2007 with about one million cases reported in each outbreak (Pan American Health Organization, 2008) greatly increasing the importance of effective control programs. As of 2008, the National Institutes of Health have a product in Phase 1 testing, and both Glaxo Smith Kline and Sanofi Pasteur have candidate vaccines in Phase 2 clinical testing, and other developers are conducting laboratory research (Pediatric Dengue Vaccine Initiative (PDVI), 2008). To help guide stakeholders around continued investment in the development and potential use of a dengue vaccine, this paper uses and extends the authors, previous work on its cost-effectiveness in Southeast Asia (SE Asia) (Shepard et al., 2004). Periodically, economic analyses are conducted to guide public support for vaccine development in both industrialized (Stratton et al., 2001) and developing countries (Institute of Medicine, 1986), (Shepard et al., 1995), including previous cost-effectiveness studies of dengue (Shepard and Halstead, 1993), (Shepard et al., 2004). Insufficient economic analysis may have delayed the adoption of Hepatitis B vaccine (HBV) in developing countries (Muraskin, 1995). To minimize such risks with a pneumococcal pneumonia vaccine, a study on its cost-effectiveness is informing plans to greatly expand its use to low- and middle-income countries (Sinha et al., 2007). Assuming a pediatric dengue vaccine proves safe and effective, here we first examine whether it would be economically viable in the regional context of SE Asia. This region, the
Cost-Effectiveness of a Dengue Vaccine in Southeast Asia and Panama: Preliminary Estimates
73
focus of our earlier study (Shepard et al., 2004), has traditionally witnessed the highest incidence; the disease primarily strikes children and has been the site of major epidemics in the past 20 years (World Health Organization (WHO), 1997). The ten countries in SE Asia (Brunei, Cambodia, Indonesia, Laos, Malaysia, Myanmar, Philippines, Singapore, Thailand, and Vietnam) had a combined population in 1999 of 529 million persons (Population Reference Bureau, 2001). From 1997 through 2001, these countries officially reported to the World Health Organization (WHO) 1.2 million cases of dengue fever (DF) or dengue hemorrhagic fever (DHF), a life-threatening form of the disease, while the actual burden is projected to be far higher (Gubler and Meltzer, 1999). Our target population in SE Asia, the cohort of 11.6 million children who reached the age of 15 months sometime in 2001, was based on births in 1999 (United Nations, 2000, 2001). Our analysis next shows how our earlier framework (Shepard et al., 2004) can be applied to a national context. As a recent publication about Panama provides a comprehensive assessment of the cost of dengue in that country (Armien et al., 2008), we have selected it for illustration. Panama also illustrates the growing problem of dengue in the Americas. In Panama, dengue was re-introduced in 1985 and the country has experienced transmission since 1993. In 2005, the country experienced its worst epidemic since 1993, with over 5,482 dengue cases, 7 dengue hemorrhagic fever cases, and 5 deaths reported (Armien et al., 2008). For both areas, this paper (1) estimates the health benefits of offering dengue vaccination to annual child cohorts in that region, (2) determines the annual cost of that strategy, and (3) projects potential revenues to vaccine producers from sale of a pediatric dengue vaccine (Mahoney and Maynard, 1999). Our analysis is designed to inform the many stakeholders in vaccine development, including national governments, donors, research institutes, vaccine producers, and financing intermediaries such as the Vaccine Fund, which supports the activities of the Global Alliance for Vaccines and Immunization (GAVI) (Global Alliance for Vaccines and Immunizations, 2002). The stakeholders interact because investment decisions by donors and producers on vaccine development depend, in part, on the potential utilization and revenue from a vaccine if it is developed. Through this paper, we illustrate the choices facing policy makers. They must decide not only whether to recommend use of a dengue vaccine if available, but also the appropriate target population – specifically adults and infants. Addressing these questions highlights the needs and challenges in obtaining the needed data and predictions. Uncertainties affect many key parameters – the degree of under-reporting of dengue, the exact loss in quality of life, the mixture of treatment sites, and the cost, efficacy, period of protection, and side effects of a potential vaccine. For this reason, our approach is primarily didactic and our results and sensitivity analyses are necessarily preliminary. To address these limitations, we also indicate a research agenda for more definitive findings.
2
Methods
We conducted this economic analysis using standard approaches to cost-effectiveness analysis (Shepard and Thompson, 1979), (Gold et al., 1996) to derive the cost per Disability Adjusted Life Year (DALY) saved (World Bank, 1993). Each DALY saved represents 1 year of healthy life gained due to postponement of mortality and/or reduction in rate or severity of morbidity. To estimate the potential benefits of vaccination, we first constructed a conceptual model of
1283
1284
73
Cost-Effectiveness of a Dengue Vaccine in Southeast Asia and Panama: Preliminary Estimates
. Figure 73-1 Disease model of dengue and transition rates for SE Asia. The disease model shows the possible transition from a member of the general population to death from dengue hemorrhagic fever, with probabilities of favorable and unfavorable outcomes at each branch
dengue and calibrated it based on existing literature (> Figure 73-1) (Shepard et al., 2004). This proved important because terminology between dengue infections and dengue cases or dengue fever versus dengue hemorrhagic fever is sometime inconsistent. It also provided the basis for disease modeling, where each rectangle represents a model state. In adapting this framework to Panama, we used the number of reported cases in 2005, 2006, and 2007 by age as of mid-2008 (Panama Ministry of Health, 2008). Due to reporting lags, data for 2007 were still preliminary. Reported cases are widely believed to understate the true burden. We calculated projected cases by multiplying the reported cases times six, the expansion factor estimated by the Ministry of Health (Armien et al., 2008). As 2005 was an epidemic year for Panama, the two subsequent years had fewer cases. Nevertheless, dengue has been increasing steadily in the Americas (Pan American Health Organization, 2008). For that reason, we chose to calibrate this model from the most recent three years, rather than earlier data. In 2005, Panama had a population of 3.2 million persons and a per capita gross national income (GNI) of $4,630; it projected 32,934 clinical cases, of which 456 were hospitalized and 5 were fatal (Armien et al., 2008). To convert numbers of cases into population-based rates, we projected the population of Panama for 2006 and 2007 based on its growth rate of 1.528% per year (Central Intelligence Agency, 2008). As long term projections showed that the proportion of the population in the age group in which most dengue cases fall (ages of 15–59) would barely change through the year 2050 (United Nations, 2005), we applied the overall population growth rate to each group for our limited projection to the year 2006. We chose that year as it was not an epidemic year and had data for the full year.
Cost-Effectiveness of a Dengue Vaccine in Southeast Asia and Panama: Preliminary Estimates
73
. Table 73-1 Selected dengue indicators for Panama by year, 2005–2007 Indicator
2005
2006
2007
Reported dengue cases
5,489
4,323
3,527
Projected dengue cases
32,934
25,938
21,162
3,228,186
3,277,513
3,327,593
Estimated population Projected dengue cases per 100,000 population
1,020
791
636
Projected dengue infections per person per year
4.3%
3.3%
2.6%
Life expectancy at birth (years)
75.0
75.0
75.0
Projected lifetime infections per person (undisc)
3.188
2.473
1.987
Projected lifetime cases per person (undiscounted)
0.765
0.594
0.477
Projected lifetime cases per person (discounted)
0.293
0.230
0.188
> Table 73-1 reports selected indicators for Panama by year from 2005 through 2007. As the life expectancy at birth of Panama is 75 years, we projected lifetime cases by multiplying the average incidence of projected clinical cases times the life expectancy. As only 24% of infections are estimated to be clinical cases (Shepard et al., 2004), we projected lifetime infections by dividing the clinical cases by 24%. > Figure 73-2 shows the rates of reported dengue cases by age and year for 2005 through 2007. Whereas in SE Asia dengue affects primarily children (Shepard et al., 2004), in Panama the highest incidence is in adults. To understand the lifetime pattern of dengue, we calculated the cumulative number of projected dengue cases by age for Panama. This involved using the aforementioned expansion factor and the period of risk within each age interval. As an approximation, we ignored deaths within each age group except the last, so each closed interval had a 5-year period of risk. For the last, open-ended interval, we assumed a 10-year period of risk, approximating to the discounted life expectancy at age 70. > Figure 73-3 plots these cumulative rates by year. While the cumulative rates differ among the 3 years, the relative patterns by age are extremely stable. As we wanted to examine the effect of vaccinations at different ages, we needed to estimate the proportion of expected lifetime average clinical cases that occurred after each specified age category after birth, without adjustment for mortality over the lifespan. As shown in > Figure 73-4, this proportion declines steadily from 100% at birth to 0% by age 65. As the benefits of trying to vaccinate persons aged 65 and above would be very low; however, this simplifying assumption of no adjustment for mortality is reasonable. If a more complex analysis were conducted with adjustment for mortality, the proportion would decline to a small fraction, but not necessarily zero, at age 65. While these proportions could be structured by age group, the cumulative proportions differed by no more than one percentage point among age groups, so average proportions were used here instead. Our time horizon for the analysis was the lifetime for the cohort, which, for Panama, we set to the country’s life expectancy of 75 years (World Bank, 2008). With the expected average annual dengue infection rate of 3.5%, Panama has two thirds of the corresponding rate for SE
1285
1286
73
Cost-Effectiveness of a Dengue Vaccine in Southeast Asia and Panama: Preliminary Estimates
. Figure 73-2 Rate of reported dengue cases in Panama by age and year. With square marks for 2005, circles for 2006, and triangles for 2007, the graph shows that all 3 years show similar patterns of age-specific incidence rates of reported dengue cases, with the highest rates being among those aged 55–59 years
Asia of 5%. We adopted a societal perspective for evaluating costs and benefits. As Panama’s official currency is the US dollar, we kept all economic analyses in 2005 US dollars. Our reference for all economic analyses was the current situation, i.e., no dengue vaccination exists and dengue remains a threat despite ongoing vector control programs. The vector control program entails varying levels of public education, inspection of potential breeding sites, elimination or appropriate management of larval habitats, and insecticidal measures against larvae and adults. Because of the predominance of adult cases, we considered two alternative vaccination scenarios. The first would vaccinate children as they pass through the routine health care system at about age 1 year, as was assumed for SE Asia (Shepard et al., 2004). The second is a catch up program, in which vaccination is offered to persons aged 15–64. In adapting our model from SE Asia to Panama, there were three major changes. First, the major difference in the share of cases being hospitalized is substantially lower in Panama. Second, we were able to estimate treatment costs from the recent study from Panama (Armien et al., 2008), rather than the literature. The resulting costs per case proved higher than the earlier estimates for SE Asia based on earlier international literature (Shepard et al., 2004). Third, the annual rate from ‘‘population’’ to ‘‘infection,’’ based on 2006 data, was 3.3% in Panama, compared to 5% in SE Asia.
Cost-Effectiveness of a Dengue Vaccine in Southeast Asia and Panama: Preliminary Estimates
73
. Figure 73-3 Modeled projected dengue cases in Panama up to specified age, by year. With square marks for 2005, circles for 2006, and triangles for 2007, the lines show that the modeled cumulative rate of dengue infections rises in a similar way among age categories for the 3 years. An individual could potentially experience up to four infections, one of each dengue serotype
Other factors remained the same as our previous model from SE Asia. We assumed that vaccination would be 95% efficacious in providing lifetime protection against each of the four dengue virus serotypes (Shepard et al., 2004), reducing the recipient’s annual risk of infection to 0.165%. Vaccination would have negligible side effects. The quality of life values were based on losses per day (compared to perfect health) of 0.81 and 0.85, for DF and DHF, respectively (Gubler and Meltzer, 1999). In both areas, we assumed that children would be vaccinated with two doses. In SE Asia, we projected a mixed distribution system (both public and private) and twotiered prices from experience with HBV. For children vaccinated in public sector, the first dose would coincide with the visit for measles vaccination, generally given at the average age of 9–12 months (Centers for Disease Control, 2002). The second, requiring a separate visit, would occur around 18 months, the age just prior to the period of risk. For children vaccinated in the private sector, we assumed that neither dose would coincide with an existing vaccination and each thus dose would require a separate visit. Based on the current vaccination coverage for DPT3 in SE Asia, we assumed that 85% coverage would be achieved at full implementation, and 0.25 additional doses would be procured for each dose administered to compensate for an estimated 20% wastage rate (UNICEF, 1998).
1287
1288
73
Cost-Effectiveness of a Dengue Vaccine in Southeast Asia and Panama: Preliminary Estimates
. Figure 73-4 Modeled percentage of dengue cases beyond each specified age category per person alive (based on 2005–2007 data). This graph shows the percentage of cases beyond each age category, normalized to the average rate, falls with age. The pattern is the inverse of the modeled cumulative incidence rate
In Panama, a middle-income country in a region where governments traditionally buy vaccines through a revolving vaccine fund, we assumed a single national price. Our base case assumption used a bulk price of $5 per dose to parallel the assumed bulk price for pneumococcal conjugate vaccine for children (Sinha et al., 2007). While the price of the pneumonia vaccine is substantially higher in industrialized countries, a price of $5 per dose should allow widespread adoption in a middle-income country, such as Panama, and, in large volume, may cover the marginal cost of production. Financing could come from any combination of government, households, donors, insurers, and social insurance systems. As fair pricing is one of the criteria affecting the societal contribution, ranking and reputation of international vaccine manufacturers (Menou et al., 2008), companies have good reasons to set affordable prices in developing countries. On top of the vaccines, users would need syringes and services. To adjust for inflation and higher prices in Panama than SE Asia, we estimated that syringes would cost $0.10 each. Costs per contact of labor, overhead, vaccine distribution, and storage for administering the vaccine would be $7 per contact, twice the original level (Shepard et al., 1995). The model ‘‘discounted’’ health benefits (expressed as DALYs) and costs saved from vaccination at an annual real (net of inflation) rate of 3% per year.
Cost-Effectiveness of a Dengue Vaccine in Southeast Asia and Panama: Preliminary Estimates
73
We performed sensitivity analyses of key variables and assumptions in SE Asia. First, we varied the vaccine price per dose in SE Asia over the range of $0.25 (the current price of HBV under the Vaccine Fund) to $10, which is twice the base price. It should be noted that the price at which HBV started to be used in the public sector in developing countries was $1.50 (Mahoney and Maynard, 1999). Second, we varied the rate of clinical disease above and below our best estimate based on alternative values of Gubler and Meltzer’s (1999) ‘‘multiplier’’ that projects actual cases from reported cases. Third, we also examined a ‘‘least cost-effective scenario’’ with the lowest infection rate, a low vaccine efficacy (85%), the highest public sector vaccine price ($1.50), and only 85% of children served through the public sector. In Panama, we considered a price range of $1–10 per dose, where the lower bound is comparable to the public sector price for long established vaccines, while the upper bound is twice the estimated bulk price. We also considered alternative expansion factors, assuming that the projected number of cases per reported case could range from 2 (giving the lowest number of cases) to 10, based on the broad range of expansion factors in the international literature (Suaya et al., 2007). To best adapt our model to available data from Panama, we focused our model of impact around projected clinical cases. Thus, averaging across hospital and ambulatory cases, both fatal and non-fatal we derived the direct cost of treating one clinical case (the sum of medical and non-medical direct costs) and the indirect cost per case. We then discounted these to present value based on the age of the vaccines (1 year for infants, and 15–64 for adults, weighted according to their shares in the Panama population in 2005). We used the incidence rate in 2006 (our middle year), alternative expansion factors for under-reporting, and assumed vaccine characteristics to estimate the number of cases averted and associated health benefits (in DALYs) and dollars gained.
3
Results
The baseline disease burden of dengue in SE Asia for the cohort is 0.42 DALYs per 1,000 population, of which 52% is due to premature mortality and 48% to acute morbidity. The baseline gross cost of treatment is $99 per 1,000 population per year. Although DHF constitutes only 6% of clinical cases, it represents 68% of the disease burden and 67% of the treatment costs. In SE Asia, the dengue vaccination program would cost $50 per DALY saved under the base assumptions (> Table 73-2). This cost-effectiveness ratio is the cost to buy . Table 73-2 Cost-effectiveness of dengue vaccination in SE Asia under alternative DHF rates and costs of vaccine in the public sector (US$/DALY gained)a Public sector vaccine price per dose
a
DHF per 100,000 population
$0.25
$0.50
$1.00
$1.50
36
$438
$499
$622
$683
72
$19
$50
$111
$41
108
CS
CS
CS
CS
The base case values for this sensitivity analysis are bolded. CS denotes cost saving (i.e., vaccination costs less than the status quo)
1289
1290
73
Cost-Effectiveness of a Dengue Vaccine in Southeast Asia and Panama: Preliminary Estimates
a year of good health. The lower that ratio, the more favorable the health intervention. The calculated ratio also means that each $1 million invested in dengue vaccination would save 20,000 DALYs. The sensitivity value shows that the cost-effectiveness ranges from 0 to 100,000. In SE Asia, under our most likely assumptions, vaccination reduces both the mortality and morbidity burdens of disease by 82%, saving 0.34 DALYs per 1,000 population per year. The incremental cost of vaccinating one child against dengue would be $4.85 in the public sector, $39.10 in the private sector, and $8.28 overall. Since each child would receive two doses, the cost per dose would be $4.14. The gross cost per 1,000 population of the vaccination program would be $154. This cost is relatively low because children must be vaccinated during just 1 year of their life, but that cost is allocated over the entire population. The net cost per 1,000 population, $17, is 89% below the gross cost because of the savings in health care costs from fewer dengue cases. In aggregate, when fully implemented in SE Asia, dengue vaccination would cost $81.7 million per year, save $72.7 million in treatment, and have a net cost of $9.0 million. It would save 182,000 DALYs per year. In Panama, the incremental cost of vaccinating one person against dengue would be $19.70 for an infant and $26.70 for an adult (> Table 73-3). The adult cost is higher because the adult needs two incremental contacts with the health care system, whereas the child needs only one. Even ignoring the cost offsets, the vaccine is relatively cost-effective, at $2,069 and $2,574 per DALY in infants and adults, respectively. The cost offsets are more substantial than those in SE Asia as the costs of dengue treatment are higher, based the recent empirical study (Armien et al., 2008). The cost offsets represent 178% of vaccination costs for infants and 138% for adults. Thus, with the cost offsets included, the vaccine becomes cost saving. . Table 73-3 Economic evaluation of dengue vaccination in Panama under most likely assumptions Item
Adults
Expansion factor for cases (projected/reported)
6
6
Projected lifetime cases per person (discounted)
0.2304
0.2510
Vaccine efficacy
a
Infants
95%
95%
Projected lifetime cases averted per person vaccinated (disc.)
0.2189
0.2384
Disease burden per discounted case (DALYs)
0.0435
0.0435
Disease burden averted per person vaccinated (DALYs)
0.0095
0.0104
Gross cost per person vaccinated (US$)
$19.70
$26.70
Direct cost per discounted case
$154.27
$154.27
Projected lifetime direct cost averted per person vaccinated
$33.77
$36.78
Net cost per person vaccinated
-$14.07
-$10.08
Total cost per discounted case (direct and indirect)
$359.14
$359.14
Projected lifetime total cost averted per person vaccinated
$78.62
$85.63
Cost-effectiveness, ignoring offsets (US$/DALY)
$2,069
$2,574
Cost-effectiveness, including offsets (US$/DALY)
CSa
CSa
Benefit-cost ratio
4.0
3.2
CS denotes ‘‘cost saving,’’ so vaccination offsets more than its costs
73
Cost-Effectiveness of a Dengue Vaccine in Southeast Asia and Panama: Preliminary Estimates
. Table 73-4 Economic analysis of dengue vaccination in Panama under alternative expansion factors and vaccine prices Vaccine price per dose Expansion factor
Include cost offsets
$1 Infants
$5
$10
Adults
Infants
Adults
Infants
Adults
Cost-effectiveness (US$/DALY gained) 2 6a 10
Yes
CS
$1,284
$2,660
$4,177
$6,598
$7,792
No
$3,056
$4,831
$6,207
$7,723
$10,145
$11,339
Yes
CS
CS
CS
CS
CS
$233
No
$1,019
$1,610
$2,069
$2,574
$3,382
$3,780
Yes
CS
CS
CS
CS
CS
CS
No
$611
$966
$1,241
$1,545
$2,029
$2,268 0.73
Benefit-cost ratio 2
Yes
2.7
1.7
1.3
1.1
0.81
6a
Yes
8.1
5.1
4.0
3.2
2.4
2.2
10
Yes
13.5
8.5
6.7
5.3
4.1
3.6
a
The most likely values corresponding to the most likely assumptions are bolded
> Table 73-4 reports the sensitivity analysis for Panama. For most values of the vaccineprice and expansion factor, the vaccine is cost saving. Even under the least favorable combination of the expansion factor (two times reported cases) and vaccine price ($10 per dose), the vaccine would have borderline cost-effectiveness for a middle income country.
4
Discussion
Cost-effectiveness. Cost-effectiveness interpretations should first relate to other health interventions as a whole. The cost-effectiveness ratio for a dengue vaccine in SE Asia ($50 per DALY saved) is comparable to the most favorable public health programs for children, which each cost less than $100 per DALY saved (i.e., control of respiratory infections, perinatal conditions, diarrheal disease, pertussis, polio, measles, tetanus, malaria, malnutrition, and vitamin A deficiency). It is 20 times more favorable than the treatment of leukemia (World Bank, 1993). Next, a dengue vaccine should be judged against other approaches to prevention, particularly environmental control. By contrast to the cost per DALY of $50 for vaccination, the published cost-effectiveness ratio for an environmental management approach to dengue prevention and control ($3,139 per DALY), derived from the Singapore experience, was far less favorable (Shepard and Halstead, 1993). While a pediatric dengue vaccine would be expected to confer virtual lifelong immunity, environmental management, must be delivered repeatedly each year to the entire population as part of an integrated vector control strategy, with associated high recurrent costs. With few exceptions, sustained control or prevention of dengue virus transmission by vector control has not been achieved in recent decades, mainly due to operational constraints including weaknesses of program delivery, continuity and
1291
1292
73
Cost-Effectiveness of a Dengue Vaccine in Southeast Asia and Panama: Preliminary Estimates
coverage, and over-reliance on insecticidal control methods, especially those targeting adult mosquitoes. Control programs have generally failed to engage communities in sustainable behavior change that prevents the creation of, or controls larval habitats. We estimate that after many cohorts have been vaccinated, vaccinations would be at least as effective as current vector control programs, and could supplant them for purposes of preventing dengue transmission. To estimate the potential savings, we assembled the most recent available data on dengue vector control. As dengue is primarily an urban disease, more urbanized countries had higher spending. Annual cost per 1,000 population was $15 in Indonesia (1998), $81 in Thailand in 1994 (Sornmani and Okanurak, 1995), $188 in Thailand in 1998, $240 in Malaysia (2002), and $2,400 in Singapore (2000) (unpublished data, World Health Organization for Indonesia and Thailand; Ministry of Health, Malaysia; Ministry of the Environment, Singapore). For 1997, among 14 Latin American countries, corresponding spending on control ranged from $20 to 3,560 with a median $260 (Pan American Health Organization (PAHO), 1999). For 17 Caribbean islands in 1990, the corresponding expenditure ranged from $140 to 8,490 with a median of $1,340 (Nathan, 1993). As control costs were predominantly for personnel, we used the same methodology applied for vaccination costs to compute the average expenditure on environmental control for SE Asia. In 2001 prices, this unweighted average was $66 per 1,000 population. Thus, if vaccination allowed at least a one third reduction in spending on environmental control, it would be cost saving. In Panama, as noted, we found that the dengue vaccination would be cost saving under the most likely values. An upper threshold of three times the GDP per capita is recommended by WHO to identify interventions that are ‘‘cost-effective’’ in low- and middle-income countries (World Health Organization, 2008). The factor of three incorporates the human element in cost-effectiveness – societies value years of healthy life for more than just their productivity value. Even the least favorable cost-effectiveness ratio, $11,339, is less than three times Panama’s per capita GNI. Vaccination costs. One of the most critical questions in pharmaco-economics concerns the price of the product. That price must satisfy the opposing needs of the consumers and the producers of a product. It must be high enough to satisfy the needs of the producer. That is, it must cover at least the amortized development and production costs plus a return competitive with other alternatives. On the other hand, it must be low enough that the public sector could purchase the vaccine on a large scale. This analysis for SE Asia suggested that our base case price per dose ($0.50 in public sector and $10 in private sector) could meet these two objectives. The proposed public price exceeds the cost per dose of older vaccines in the Expanded Programme on Immunization, but is below the price of those most recently introduced (e.g., HBV and Hib), suggesting it is a plausible average over the vaccine’s anticipated lifetime. The hypothesized private price is comparable to existing private sector prices in SE Asia, e.g., $9.51 per dose for Hib and $13.59 for the Japanese encephalitis in Malaysia manufactured by Biken (Cardosa, M.J. personal communication, June 2003). Both industry and government representatives at a major dengue conference understood the rationale for two-tiered prices of these magnitudes (Almond et al., 2002). Under a national program, governments would purchase vaccines from producer via tender, with negligible commissions and fees. The full cost per dose (weighted average of the public and private sectors, including labor and overhead) of $4.14 for SE Asia lies within the range ($0.50–6.00) of reported cost per dose in a systematic review of studies of expanding vaccination coverage (Pegurri et al., 2005). The analysis indicated that the vaccine price was also a critical variable in Panama. Without considering treatment costs, the cost-effectiveness tripled as the assumed price rose. With Panama being a middle-income country, dengue vaccination is clearly cost-effective
Cost-Effectiveness of a Dengue Vaccine in Southeast Asia and Panama: Preliminary Estimates
73
at the most favorable and most likely prices assumed in this analysis, but may not necessarily be cost-effective at the least favorable (higher) assumed price. Limitations and results of sensitivity analyses. The greatest limitations in our model are uncertainties in its parameters. Multivalent vaccines have not yet reached the level of development to know their efficacy, duration and type of protection, and side effects. Incomplete reporting leads to uncertainty about the current burden of dengue. Our sensitivity analysis showed that even with less favorable vaccine performance and disease incidence, vaccination remained cost-effective in SE Asia. Sensitivity analyses showed that the two factors that most affected the cost-effectiveness of a dengue vaccine were: (1) the price per dose of vaccine, and (2) the incidence of DHF. The low incidence of DHF in SE Asia (36 per 100,000 population per year) corresponds to half of the base case values for each age group, with an overall annual infection rate of 2.5%. The intermediate value (72) corresponds to the central values described in > Figure 73-1. The high incidence (108) corresponds to an infection rate of 6% (16% for those under 15 years, and 3% for those 15 and above) and 30% of infections being clinical cases. Higher rates of infection would be impossible over the long run, as they would require reinfections with the same serotype. If the incidence of DHF were high enough and the price were low enough, savings in treatment of DF and DHF would more than offset the costs of vaccination. With the highest incidence of DHF, vaccination is actually cost saving (CS) for all four prices shown and, in fact, for any price per dose up to $1.73. The highest incidence and lowest price achieve the greatest savings. The cost-effectiveness ratio under least favorable assumptions in > Table 73-2 ($683 per DALY) and the cost-effectiveness ratios for poorer vaccine performance ($788 and $960 per DALY) all fall in the same category as several other public health programs in adults (e.g., control of diabetes and motor vehicle injuries) (World Bank, 1993). Moreover, these cost-effectiveness ratios are less than the region’s per capita GNI of $1,083 – another benchmark for cost-effective interventions. Extension to other regions and population groups. The cost-effectiveness of vaccinations, like that of other preventive programs, is roughly proportional to the disease burden. While dengue has been endemic in SE Asia for decades, it returned to the Americas in 1980 and spread in Brazil, Venezuela, Mexico, and other countries. In the United States, 102 cases of DF were reported from Hawaii in 2001, the first outbreak on that island in 56 years (Centers for Disease Control, 2002). Similar cost-effectiveness analyses could examine the extension of vaccination to other countries, to older children and various adult ages in endemic countries, as well as to travelers, and members of the military from non-endemic countries.
5
Conclusions
Several factors contributed to preliminary results favoring the development and subsequent use of a dengue vaccine in SE Asia. First, at a projected price of $0.50 per dose in the public sector, the dengue vaccine would be relatively inexpensive. This projection was based on the recent success of large public sector procurement through the Vaccine Fund in driving down the price of HBV to half of this level. It also counted on donor involvement in effectiveness testing, large public sector procurement, with concomitant low marketing overhead, to provide the producer with a means to reduce unit costs of production and distribution, thereby increasing overall profitability. Second, by assuming that dengue vaccination would be linked with the last dose of the Expanded Program on Immunization, a high (85%) coverage rate seemed reasonable.
1293
1294
73
Cost-Effectiveness of a Dengue Vaccine in Southeast Asia and Panama: Preliminary Estimates
In Panama, the similarity of preliminary results between targeting adults and infants as vaccines was striking. Several differences between the two age groups happened to counterbalance one another. An adult program was assumed to accept any adult aged 15–64. The average age of those vaccinated would be about 40 years, just prior to the age of highest incidence (age 50). The use of discounting favors focusing on adults because their lag between vaccination and peak disease is small, but for infants is large. This advantage for adults is offset by a smaller fraction of projected lifetime clinical cases averted, due to a substantial portion of the disease occurring before adult vaccination. If a vaccine did not provide protection that lasts over multiple decades, then adult vaccination would clearly be preferable. Several types of research are needed to address the data uncertainties. More studies on disease burden, utilization of health services, and cost of illness, particularly population-based investigations, can help clarify incidence, patterns and costs of treatment, and outcomes. Definitive data on vaccine performance must await Phase III and IV field studies. Vaccine prices depend on manufacturers’ judgments about achieving financial and public recognition goals.
Summary points We developed and calibrated a cost-effectiveness model of vaccinating children at 15 months in SE Asia and Panama.
We selected a societal perspective. We assumed that full immunization would require two doses. As our most likely assumptions, we assumed prices per dose of $0.50 and $10 each in the public and private sectors, respectively, in SE Asia, and a uniform price of $5 per dose in Panama.
The gross cost per 1,000 population (of all ages) of the vaccination program would be US $154 in SE Asia.
Projected savings in dengue treatment were projected to offset 89% of vaccination cost in SE Asia, reducing the net cost per capita to only $17 per 1,000 population.
The cost per disability adjusted life year (DALY) saved by a pediatric vaccine would be $50 in SE Asia, making the potential vaccine highly cost-effective.
In Panama, due to higher treatment cost, a moderately priced vaccine would be cost saving for both infants and adults; the result is even more favorable than being highly costeffective. From a benefit cost perspective, a moderately priced dengue vaccine in Panama would also be economically advantageous, with estimated benefits equal to 4.0 and 3.2 times the costs for infants and adults, respectively. In countries such as Panama where dengue affects primarily adults, both infant and adult vaccinations are advantageous strategies. Eventually, vaccination may allow countries to reduce spending on environmental control.
Acknowledgments The preparation of this chapter was supported in part by the endowment of the Schneider Institutes for Health Policy, Brandeis University, Waltham, MA, USA. The authors are indebted to Gladys Guerrero, MD, MPH, and Dennys Rodriguez of the Department of
Cost-Effectiveness of a Dengue Vaccine in Southeast Asia and Panama: Preliminary Estimates
73
Epidemiological Surveillance of the Panama Ministry of Health for providing epidemiological data from Panama, to Blas Armien, MD, of the Gorgas Memorial Institute, Panama, for assistance with analysis of these data, and to Clare L. Hurley, MM, of Brandeis University for editorial assistance.
References Almond J, Clemens J, Engers H, Halstead SB, Khiem HB, Pablos-Mendez A, Pervikov Y, Tram TT. (2002). Vaccine. 20: 3043–3046. Armien B, Suaya J, Quiroz E, Sah B, Bayard V, Marchena L, Campos C, Shepard D. (2008). Am J Trop Med Hyg. 79: 364–371. BBC. (2007). Dengue sparks Paraguay emergency. London http://news.bbc.co.uk/2/hi/americas/6407287. stm. Accessed 2 March 2007. Centers for Disease Control. (2002). Notice: dengue fever in Hawaii. http://www.cdc.gov/travel/other/denguehawaii-oct2001.htm. Accessed 23 January 2002. Central Intelligence Agency. (2008). Panama. The world factbook. Langley, VA, Central Intelligence Agency. https://www.cia.gov/library/publications/the-worldfactbook/geos/pm.html. Accessed June 25, 2008. Global Alliance for Vaccines and Immunizations. (2002). http://www.vaccinealliance.org/ Accessed 29 January 2002. Gold MR, Siegel JE, Russell LB, Weinstein MC. (1996). Cost-Effectiveness in Health and Medicine. Oxford University Press, New York. Gubler D, Meltzer M. (1999). Adv Virus Res. 53: 35–70. Gubler DJ. (1998). Clin Microbiol Rev. 11: 480–496. Institute of Medicine. (1986). New Vaccine Development: Establishing Priorities, vol. II: Diseases of Importance in Developing Countries. National Academy Press, Washington, DC. Mahoney RT, Maynard JE. (1999). Vaccine. 17: 646–652. Menou V, Hornstein A, Lipton-McCombie E. (2008). Access to medicine index (Report prepared by the Innovest Healthcare Team). Netherlands Access to Medicine Foundation, Haarlem. http://atmindex. org/download/67. Muraskin W. (1995). The War Against Hepatitis B: A History of the International Task Force on Hepatitis B Immunization. University of Pennsylvania Press, Philadelphia. Nathan MB. (1993). J Am Mosq Control Assoc. 9: 1–7. Pan American Health Organization (PAHO). (1999). A Blueprint for Action for the Next Generation: Dengue Prevention and Control. http://www.paho. org/English/HCP/HCT/hct-136–99.doc Accessed 29 January 2002. Pan American Health Organization. (2008). Dengue: Progress Report. Provisional Agenda Item 4.10, document
CE142/17 (Eng.) for 142nd Session of the Executive Committee, Washington, D.C., USA, 23–27 June 2008. Pan American Health Organization, World Health Organization, Washington, DC. http://www.reliefweb. int/rw/RWFiles2008.nsf/FilesByRWDocUnidFilename/ ASIN-7F3PF4-full_report.pdf/$File/full_report.pdf. Accessed 22 May 2008. Panama Ministry of Health. (2008). Dengue cases by health region and age group, 2005–2007. [Casos de dengue segu´n region de salud por grupo de edad, 2005–2007]. Directorate of General Health, Department of Epidemiology, Panama. Pediatric Dengue Vaccine Initiative (PDVI). (2008). Dengue vaccines. Pediatric Dengue Vaccine Initiative, Seoul, Korea. http://www.pdvi.org/vaccines/vaccine. htm www.pdvi.org. Pegurri E, Fox-Rushby JA, Walker D. (2005). Vaccine. 23:1624–1635. Population Reference Bureau. (2001). http://www.prb. org/Content/NavigationMenu/Other_reports/ 2000–2002/sheet1.html. Accessed 18 August 2002. Shepard D, Suaya J, Halstead S, Nathan M, Gubler D, Mahoney R, Wang D, Meltzer M. (2004). Vaccine. 22: 1275–1280. Shepard D, Thompson M. (1979). Public Health Rep. 94: 535–543. Shepard D, Walsh J, Kleinau E, Stansfield S, Bhalotra S. (1995). Vaccine. 13: 707–714. Shepard DS, Halstead SB. (1993). In: Jamison DT, Mosley WH, Measham AR, Bobadilla JL (ed.) Disease Control Priorities in Developing Countries. The World Bank, Oxford University Press, New York, pp. 303–320. Sinha A, Levine O, Knoll M, Muhib F, Lieu T. (2007). Lancet. 369: 389–396. Sornmani S, Okanurak K. (1995). Social and economic impact of dengue haemorrhagic fever in Thailand. Social and Economic Units, Faculty of Tropical Medicine Mahidol University. Stratton K, Durch JS, Lawrence RS (ed.). (2001). Vaccines for the 21st Century: A tool for Decision Making. National Academy Press, Washington DC. Strobel M, Lamaury I. (2001). Rev Med Interne. 22: 638–647. Suaya JA, Shepard DS, Beatty ME. (2007). Dengue: burden of disease and costs of illness. Working
1295
1296
73
Cost-Effectiveness of a Dengue Vaccine in Southeast Asia and Panama: Preliminary Estimates
paper 3.2. In: Report of the Scientific Working Group meeting on Dengue, Geneva, 1–5 October 2006 (TDR/SWG/07). Special Programme for Research and Training in Tropical Diseases (TDR), WHO, www. who.int/tdr, Geneva, Switzerland, pp 35–49. UNICEF. (1998). The state of the world’s children. http://www.unicef.org/sowc98/tab1.htm. Accessed 29 January 2002. United Nations. (2000). Expected births for 2001. In. Statistics Division. http://unstats.un.org/cgi-bin/ cdbdemo.exe. Accessed 29 January 2002. United Nations. (2001). Social indicators. Statistics Division. http://www.un.org/depts/unsd/social/population.htm. Accessed 29 January 2002. United Nations. (2005). World Population Prospects. The 2004 Revision (ESA/P/WP.193). United Nations,
New York. http://www.un.org/esa/population/publications/WPP2004/2004Highlights_finalrevised.pdf. World Bank. (1993). World Development Report: Investing in Health. World Bank, Washington, DC. World Bank. (2008). World development indicators. www.worldbank.org. Accessed 1 June 2008. World Health Organization (WHO). (1997). Dengue Hemorrhagic Fever. Diagnosis, Treatment, Prevention and Control. World Health Organization, Geneva. World Health Organization. (2008). Cost-Effectiveness Thresholds. Choosing Interventions that are Cost Effective (WHO-CHOICE). In World Health Organization, Geneva. http://www.who.int/choice/costs/ CER_thresholds/en/index.html. Accessed 29 June 2008.
74 Burden of Sexually Transmitted Chlamydia trachomatis Infections L. M. Niccolai . D. Berube 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1298
2
Global Burden . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1299
3
Burden in Developed World . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1300
4 Disparities in Burden by Sex, Age, and Race . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1303 4.1 Women Have a Higher Observed Prevalence of Infection . . . . . . . . . . . . . . . . . . . . . . . . . 1303 4.2 Adolescents and Young Adults are Most Affected by C. trachomatis Infections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1304 4.3 Racial and Ethnic Minorities are Disproportionately Affected by C. trachomatis Infections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1304 5
Burden of Repeat C. trachomatis Infections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1306
6
Increasing Burden of C. trachomatis Infections? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1307
7 Burden of Associated Sequelae . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1308 7.1 Women Suffer Negative Long-Term Consequences as a Result of C. trachomatis Infections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1308 7.2 C. trachomatis is Associated with High Economic Burden . . . . . . . . . . . . . . . . . . . . . . . . . . 1309 7.3 Infection with C. trachomatis Causes Increased HIV Transmission and Acquisition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1309 8
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1309 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1310
#
Springer Science+Business Media LLC 2010 (USA)
1298
74
Burden of Sexually Transmitted Chlamydia trachomatis Infections
Abstract: Chlamydia trachomatis (C. trachomatis) infections are the most common bacterial sexually transmitted disease in the world, and therefore these infections constitute an enormous public health problem. Globally, over 90 million new infections occur annually and a higher burden of disease is found in resource poor countries, perhaps due to limited access to care and treatment. In developed countries, C. trachomatis infection is the most common reportable disease, and risk also varies along socioeconomic factors. Younger individuals have higher burden of disease due to a variety of factors including physiological, psychological, and structural reasons, and racial/ethnic minorities have a higher burden for reasons that are not entirely clear but likely to be multifaceted. Women bear a disproportionate burden of negative health consequences of C. trachomatis infections including pelvic inflammatory disease, chronic pelvic pain, and infertility. The cost associated with treating infections and their sequelae is substantial. Infection with C. trachomatis also contributes increased HIV transmission. Recent evidence of increasing C. trachomatis > prevalence has been reported, and this necessitates immediate efforts to better understand these trends in order to implement effective control measures. Continuing and improving existing control efforts are required to reduce the burden of C. trachomatis infections. List of Abbreviations: CDC, Centers for Disease Control and Prevention; C. trachomatis, Chlamydia trachomatis; HIV, Human Immunodeficiency Virus; LGV, Lymphogranuloma Venereum; NAAT, Nucleic Acid Amplification Test; PID, Pelvic Inflammatory Disease; STD, Sexually Transmitted Disease; STI, Sexually Transmitted Infection; UK, United Kingdom; US, United States
1
Introduction
The sexually transmitted disease known as chlamydia is caused by infection with Chlamydia trachomatis bacteria. The C. trachomatis genus is sub-divided into different biovars: the lymphogranuloma venereum (LGV) biovar that causes a different sexually transmitted infection (STI) also known as LGV, and the trachoma biovar that includes the genotypes that cause genital tract and ocular diseases. The focus of this chapter will be on the sexually transmitted trachoma biovar that causes urogenital chlamydial disease. While LGV is also an important STI, its burden is less than that of chlamydia and has different epidemiology, so it will not be discussed in detail. > Sexual transmission of C. trachomatis can occur during vaginal, oral, or anal sexual contact. The same genotypes can also be transmitted by a mother to her newborn during delivery, causing eye conjunctivitis and/or pneumonia. Transmissibility in sexual partnerships is estimated to be approximately 70% from both men to women and women to men (Quinn et al., 1996). C. trachomatis infections are often asymptomatic, up to 75% of the time, and more often asymptomatic in women than in men. When symptomatic, C. trachomatis produces non-specific and generally mild symptoms. In women, signs and symptoms may include cervicitis, urethritis, endometritis, and acute salpingitis. In men, signs and symptoms may include urethritis, epididymitis, prostatitis, and proctitis. Long term consequences in women include pelvic inflammatory disease, tubal scarring, chronic pelvic pain, and infertility, and in men, Reiter’s syndrome. Diagnosis of urogenital chlamydial infections may be performed using swab specimens from the endocervix or vagina in women and urethral swabs in men, or using urine specimens
Burden of Sexually Transmitted Chlamydia trachomatis Infections
74
from both men and women. On these specimens, C. trachomatis may be detected by culture, antigen detection (including direct immunoflourescence or enzyme immunoassays), DNA hybridization, and nucleic acid amplification tests (NAATs). While cell culture is often considered the ‘‘gold standard,’’ its use is often impractical owing to the need for technical expertise, specimen storage and transport, and a cell culture system. The other methods are both more practical and less expensive, and the NAATs have superior sensitivity and specificity compared to the non-amplification methods. The recommended treatment regimen for uncomplicated chlamydial infections is either azithromycin (1 g single dose orally) or doxycycline (100 mg twice a day for 7 days orally). Both treatments have equal efficacy (>95%) and rare adverse events. Though doxycycline is less expensive, azithromycin is often preferred because of the simplicity of the single dose regimen, ensuring higher compliance. Because of the high frequency of asymptomatic infection and the long-term negative health consequences, > screening for C. trachomatis been an important focus of chlamydia control programs in many parts of the world. Recent advances in nucleic acid amplification tests that can be performed on urine specimens (and thus outside of clinical settings) and the availability of single dose treatment have made this more feasible. The goal of chlamydia control has been to shorten the duration of infection, thereby improving reproductive health for women and interrupting disease transmission.
2
Global Burden
Sexually transmitted C. trachomatis infections are a major source of morbidity throughout the world in developing and industrialized countries alike. According to the World Health Organization, the estimated number of new cases among adults in 1999 exceeded 90 million, making C. trachomatis the most common bacterial STI pathogen (Gerbase et al., 1998). The greatest burden of disease is found in South and South East Asia (>40 million new cases), in large part reflecting the dense populations of large countries in this area of the world, followed by sub-Saharan Africa (>15 million new cases) (> Figure 74-1). In addition to a greater burden of disease due to larger population sizes, prevalence estimates also suggest that individuals in developing countries are at greater risk for C. trachomatis infections. Though variability in diagnostics, > surveillance, reporting, and targeted populations make international comparisons difficult, a recent report from World Health Organization summarized key differences across a variety of countries (Gerbase et al., 1998; WHO, 2007). Studies of pregnant women from around the world indicate a greater risk in S. and S. East Asia and Africa compared to Europe and North America. For example, prevalence estimates among pregnant women in the five African countries of Cape Verde, South Africa, Gabon, Central African Republic, and Tanzania range from 6 to 13%, and estimates in the two Asian countries of Thailand and India are 6 and 17%, respectively. In contrast, prevalence estimates among pregnant women in seven European countries range from 3 to 8%, and in the US, the prevalence is estimated to be 8% (CDC, 2007). Perhaps the best evidence for regional differences throughout the world comes from a single study that used similar methodology to estimate C. trachomatis prevalence in women from ten different areas on four continents (Franceschi et al., 2007). Cross-sectional > population-based surveys were conducted between 1993 and 2004 among non-pregnant women age 15–44 years in Nigeria, Columbia, Argentina, Vietnam (two areas), China, Thailand (two areas), Korea, and Spain. Over 5,000 women were included. Findings revealed C. trachomatis prevalence ranged
1299
1300
74
Burden of Sexually Transmitted Chlamydia trachomatis Infections
. Figure 74-1 Global burden of C. trachomatis infections by world region (WHO, 2007). The burden of C. trachomatis infections is highest in South and Southeast Asia and Sub-Saharan Africa due to the populous nature of countries in these regions
from Figure 74-2). Over half of these infections are estimated to occur among persons age 15–24 years.
Burden of Sexually Transmitted Chlamydia trachomatis Infections
74
. Figure 74-2 Estimated total burden of C. trachomatis in United States. The total number of reported C. trachomatis infections in the US exceeds 1 million, but this is just a fraction of the total estimated burden due to undiagnosed infections
> Table 74-1 shows examples of research studies, organized by population and setting, providing estimates of the burden of C. trachomatis. This list is not comprehensive or systematically selected; rather, key studies that are qualitatively representative of the larger body of literature were selected. As shown in the table, prevalence varies by setting and population. Among regionally or nationally representative population-based samples, C. trachomatis > positivity estimates typically range from approximately 2 to 11%. For example, among a nationally representative sample of sexually active young adults in the US, prevalence was estimated to be 4% (Miller et al., 2004). In London UK, the estimate was 11%; the higher prevalence in this study likely reflects the selective focus on an urban area (Fenton et al., 2001). Among clinic-based samples, prevalence estimates are somewhat higher, typically ranging from 7 to 18%. Higher prevalence estimates in clinics compared to population-based representative samples may be due to higher representation of people seeking care for symptoms or other known risk factors for infection. For example, among men in four US cities seeking care at health clinics, drug treatment or detention centers, prevalence was 7% (Gaydos et al., 2006). Among youth recruited from similar settings (e.g., health clinics, detention centers) in Washington State, prevalence was 9% among women and 5% among men (Marrazzo et al., 1997). Prevalence of C. trachomatis is often equally high, e.g., 7–16%, among individuals that are considered to be in ‘‘high-risk’’ groups, defined behaviorally and sociodemographically. For example, among men and women entering prisons in California the prevalence is estimated 7% (Bernstein et al., 2006). Among disadvantaged young women entering a national job training program in US, the prevalence was estimated 10–12% (Joesoef and Mosure, 2006a). These groups have higher prevalence estimates than population-based representative samples as well, perhaps due to the over-representation of racial/ethnic minorities and individuals from impoverished backgrounds which are associated with STI risk and prevalence.
1301
1302
74
Burden of Sexually Transmitted Chlamydia trachomatis Infections
. Table 74-1 Burden of C. trachomatis infections in select populations Reference
Population
Prevalence
Population-based samples Bakken et al. (2006)
Population-based registry in Central Norway 1990–2003 examining results at time of first test (n = 28,599)
6–11% among women 15–23% among men
Fenton et al. (2001)
Population-based sample of young adults in London, UK (n = 143)
11%
Miller et al. (2004)
Nationally representative sample of sexually active young adults age 18–26 in AddHealth study US (n = 14,322)
4%
Klausner et al. (2001)
Population based sample of young, low-income women in northern California US (n = 1,439)
3%
van Valkengoed et al. (2000)
Random sample of patients of 16 general practices in Amsterdam Netherlands mailed test kits (n = 4,810)
3% among women 2% among men
Clinic- or community-based samples Gaydos et al. (2006)
Men attending adolescent primary care clinics, school-based health clinics, street-based outreach venues, communitybased organizations, drug treatment centers, and detention centers in Baltimore, Denver, San Francisco and Seattle US (n = 17,712)
7%
Griesinger et al. (2007)
Female urban adolescents in Berlin Germany presenting for gynecologic care (n = 521)
7%
Marrazzo et al. (1997)
Youth age 21 years or less at nontraditional sites including CBOs, school-based health centers, teen clinics, detention centers in Washington US (n = 10,118)
9% among women
Niccolai et al. (2003)
Pregnant adolescent recruited from ten health care clinics in Connecticut US, 1998–2000 (n = 203)
18%
Rowhani-Rahbar et al. (2006)
Men attending public STD clinics in Connecticut US (n = 4,990) 15%
5% among men
High-risk samples Bernstein et al. (2006)
Men and women newly entering six California US prisons (n = 1,270)
7%
Joesoef and Mosure, (2006a)
Disadvantaged women age 16–24 entering a national job training program in US (n = 106,377)
10–12%
Joesoef and Mosure, (2006b)
Disadvantaged men age 16–24 entering a national job training program in US (n = 51,478)
8%
Kahn et al. (2005)
Adolescents in 14 juvenile detention centers in US (n = 131,296)
16% among women 6% among men
74
Burden of Sexually Transmitted Chlamydia trachomatis Infections
. Table 74-1 (continued) Reference
Population
Menon-Johansson Young male inmates in UK et al. (2005)
Prevalence 13%
Other Burstein et al. (1998)
Students at three urban public middle schools attend school- 21% among based health clinic in Maryland US (n = 213) women 2% among men
Cohen et al. (1999) Junior and senior high school students at eight urban public 12% among schools in Louisiana US (n = 4,805) women 6% among men Debattista et al. (2002)
Men who have sex with men attending gay venues in Brisbane 4% Australia (n = 184)
Renton et al. (2006)
Women presenting for termination of pregnancy in UK (n = 863)
9%
The prevalence of C. trachomatis infection varies by setting and population. UK United Kingdom; US United States
4
Disparities in Burden by Sex, Age, and Race
4.1
Women Have a Higher Observed Prevalence of Infection
A variety of data sources demonstrate a higher burden and greater risk of C. trachomatis infection in women compared to men. US CDC surveillance data from 2006 demonstrate that the overall burden of reported chlamydial infection among women was almost three times as high as the rate among men (516 cases per 100,000 females vs. 173 cases per 100,000 males) (CDC, 2007). As shown in > Table 74-1, when similar screening patterns are applied to both men and women, the prevalence of infection is often higher in women. Among high school students in Louisiana, the prevalence was 12% in females and 6% in males (Cohen et al., 1999). Among disadvantaged young adults in the US entering a job training program, observed prevalence among women was 10–12% compared to 8% among men (Joesoef and Mosure 2006a,b). There are at least two reasons for the observed differences between men and women. First, sex differences in the burden of many STI partly reflect increased susceptibility in women due to physiological differences between males and females (Ehrhardt et al., 1999). More efficient transmission from men to women may be due to more extended contact with pathogens after sexual exposure among women (e.g., infected semen remains in the vagina) and increased exposed surface area of susceptible cells (e.g., cervix vs. urethra). Increased biologic susceptibility among women can explain the higher prevalence observed in several studies when men and women undergo similar screening patterns. (However, it should also be noted that the gender difference in transmissibility for C. trachomatis may not be as great as for other STI.) Another likely significant contributor to the higher observed burden of C. trachomatis among women is different screening patterns that result in a > detection bias and underestimation of the true burden of infection among men. This explains the greater number of reported cases among women than men (reflecting observed burden, not true risk as described above). Because of the disproportionate burden of negative health effects in women and because women are more likely to have contact with the health care system than men for reproductive
1303
1304
74
Burden of Sexually Transmitted Chlamydia trachomatis Infections
health services, women are more likely to be tested for C. trachomatis and therefore detected with infection than men. However, the advent of urine based testing for infection that can be used in both men and women is likely to increase the frequency of diagnosis in men because testing can be done outside of health care settings by non-clinicians.
4.2
Adolescents and Young Adults are Most Affected by C. trachomatis Infections
Differences in burden of C. trachomatis are also often observed by age, with younger age groups having a higher burden. Surveillance data based on case reports in the US indicate the highest age-specific rates of chlamydia in 2006 were among women age 15–19 years (2,863 cases per 100,000 females) and 20–24 years (2,797 cases per 100,000 females) (CDC, 2007). Rates drop off quickly after those age groups (e.g., to 1,141 cases per 100,000 females among women age 25– 29 years). A similar pattern by age is observed for men though the rates are lower overall. Similar patterns also emerge from non-surveillance data reports. For example, among female inmates entering California prisons in 1999, women age 25 years had a prevalence of 8.9%, women age 26–30 years had a prevalence of 4.3%, and women age >30 years had a prevalence of 2.6% (Bernstein et al., 2006). In another study of women entering a national job training program in USA from 1998 to 2004, prevalence estimates of C. trachomatis were 13% among women age 16–17 years, 11% among those age 18–19 years, 10% among ages 20–21, and 7% among ages 20–24 (Joesoef and Mosure, 2006a). The greater burden of C. trachomatis among younger women may partially reflect differential screening. Screening guidelines in the US recommend annual testing of sexually active women age 25 years or less regardless of risk factors [CDC, 2007], so this population probably undergoes more frequent screening. However, there are also biological, psychological, social, and structural reasons why younger women have higher rates of many STI and C. trachomatis in particular. Younger women may be biologically more susceptible to STI due to cervical ectopy (Lee et al., 2006). This is the condition in which the cervix still displays exposed columnar epithelium, not yet replaced by squamous epithelium as adolescents transcend through puberty into adulthood, which are the target cells for C. trachomatis. Cognitively, adolescents are less able to conceptualize the long-term impact of current actions that may increase risk-taking sexual behaviors, and many adolescents do not perceive themselves to be at risk for STI (Shrier, 2004). Behaviorally, adolescents are less likely to use condoms consistently and more likely to have higher rates of partner turnover (Shrier, 2004). Finally, many adolescents do not access sexual healthcare services due to lack of transportation, inability to pay, or concerns about confidentiality (McKee et al., 2006; Thrall et al., 2000). This can result in longer duration of infection and continued transmission in networks of younger populations. For all of these reasons, adolescents and young adults may be at true increased risk for C. trachomatis infections. This is of concern because younger women have the greatest potential loss from the negative reproductive health consequences of C. trachomatis infections.
4.3
Racial and Ethnic Minorities are Disproportionately Affected by C. trachomatis Infections
It is virtually universally observed that racial/ethnic minority groups (e.g., African-Americans, Latinos) have higher rates of C. trachomatis infections compared to whites. For example, data
Burden of Sexually Transmitted Chlamydia trachomatis Infections
74
from the National Longitudinal Study of Adolescent Health (Add Health) during 2001–2002, which includes a nationally representative sample of young adults age 18–26 years living in the US, reported an overall prevalence of 4.2% but dramatic differences by race (Miller et al., 2004). This study found that the prevalence among black women was 14% compared to 2.5% among white women. Other studies have shown similar patterns: data for men and women in the national job training program for disadvantaged youth show a 4-fold higher prevalence among black men compared to white men and a 2-fold greater difference for black women compared to white women (Joesoef and Mosure, 2006a,b). Rates were also elevated for other racial/ethnic minority groups including Asian/Pacific Islanders, American Indians, and Latinos. These consistent and dramatic differences in the burden of C. trachomatis infections are of tremendous concern. These differences are likely to be real rather than artifacts of screening as described above for sex and age. For example, differences that have been observed in representative US samples are not subject to detection or reporting biases because all selected individuals are similarly screened (Miller et al., 2004). Reasons for these disparities are not always clear, but are likely to be complex and the result of many factors. Since there is no reason to suspect biological differences in susceptibility, it is more likely that race and ethnicity are markers of fundamental determinants of health status, such as poverty, access to care, and health seeking behaviors. There is some evidence to support this. Data from the Add Health study that includes a nationally representative sample of young US adults were examined with respect to health insurance coverage, health care seeking behaviors, and chlamydial infections (Geisler et al., 2006). Data were also available by race/ethnicity. This study found remarkable differences in C. trachomatis positivity by race: 13% among black and Native Americans, 7% among Hispanic, and 2% among white and Asians. This study also found that health insurance coverage differed significantly by race/ethnicity for both men and women, and that coverage was inversely associated with chlamydia prevalence. For example, blacks were more likely to not have health insurance than whites (20 vs. 15% among women and 25 vs. 20% among men). In this same sample, not having health insurance was also significantly associated with C. trachomatis; men who had health insurance for greater than 12 months had 43% reduced likelihood of C. trachomatis positivity compared to those with no insurance, and the reduced likelihood was 36% among women. The same pattern was observed for health care seeking behaviors assessed by asking about usual source of care (e.g., primary care, emergency room, etc.). Another explanation for differences in C. trachomatis prevalence by race/ethnicity lies in a network approach. Because STI are transmitted in the context of a sexual dyad, and sexual dyads are often changing, understanding the broader connections between individuals in sexual networks is often necessary to understand transmission dynamics of STI. In other words, to understand which individuals are vulnerable to STI, it is important to examine not only their own risks but also the risks of their partners, their partners’ partners, and others in the sexual networks. Using this network analytic approach, investigators have found that African Americans’ higher STI rate can be explained in part by the patterns of sexual networks within and between different racial/ethnic groups (Laumann and Youm, 1999). They identified two sexual network factors pertaining to partner selection that explain higher rates of STI among AfricanAmericans. First, they found that partner choice among African-Americans was more highly dissortative compared to whites, meaning that among African-Americans there is more mixing between core individuals (those with more sex partners) and peripheral individuals (those with fewer sex partners) compared to whites; this results in higher prevalence among AfricanAmericans. Second, STI prevalence remains high within African-American populations because
1305
1306
74
Burden of Sexually Transmitted Chlamydia trachomatis Infections
their choice of sex partners was more segregated, that is, African-American are more likely to have African-American sex partners and to remain segregated from other (lower prevalence) racial/ ethnic groups. These findings remained consistent after controlling for behavioral risk factors (e.g., number of partners). Thus, there is compelling evidence for a network explanation for higher STI rates among African-Americans. Regardless of the cause, racial disparities in STI are dramatic and troubling, and therefore warrant additional research for an increased understanding and a targeted public health response.
5
Burden of Repeat C. trachomatis Infections
Though estimates vary depending on populations and measures, it is clear from a substantial body of literature that repeat C. trachomatis infections are also common. In other words, many individuals who are diagnosed with C. trachomatis will be diagnosed with another infection at some point in time (see > Table 74-2). . Table 74-2 Burden of repeat C. trachomatis infections in select populations Reference
Population
Prevalence
Whittington et al. (2001)
Young women seeking health care in five US cities 7% at 4 months (n = 1,194)
Golden et al. (2005)
Men and women in a treatment trial in Washington 13% at 19 weeks (in US (n = 2,751) control arm)
Fortenberry et al. (1999)
Adolescent females seeking care in Indiana US (n = 490)
18% within 6 months
Blythe et al. (1992) Adolescent females receiving GYN care in Indiana US (n = 1,308)
38% within mean 9.7 months
LaMontagne et al. Women age 16–24 in UK (n = 592) (2005)
24% per year
Hillis et al. (1994)
Retrospective analysis of case report data in Wisconsin US
7–14% within 12 months
Niccolai et al. (2007)
Adolescent women in Connecticut US (n = 411)
57% within mean 4.7 years
The prevalence of repeat C. trachomatis infections is high. UK United Kingdom; US United States
For example, in a five-city study of teenage and young adult women attending reproductive health, STD, and adolescent medicine clinics, 7% had repeat C. trachomatis 4 months after the initial diagnosis (after excluding the possibility of treatment failures by conducting a testof-cure visit at 1-month after the initial diagnosis) (Whittington et al., 2001). In a study of male and female STD clinic patients, 20% had repeat C. trachomatis during a mean follow-up of 335 days (Rietmeijer et al., 2002). In another study of female adolescents with an initial infection attending STD and community-based primary care clinics, 18% were C. trachomatispositive again within 6 months, and 20% were C. trachomatis-positive within 7–12 months (Fortenberry et al., 1999). Rates of > repeat infections range from 20 to 50 per
Burden of Sexually Transmitted Chlamydia trachomatis Infections
74
100 person-years of follow-up (LaMontagne, 2005; Niccolai et al., 2007; Rietmeijer et al., 2002; Whittington et al., 2001). The burden of repeat infections is also high: 26–53% of all diagnoses are repeat infections (Blythe et al., 1992; Niccolai et al., 2007; Rietmeijer et al., 2002). The amount of time between initial and repeat infections is relatively short, with estimates ranging from 5 to 11 months (Fortenberry et al., 1999; Niccolai et al., 2007; Whittington et al., 2001). A recent report using multiple data sources for a cohort of young women followed for a mean of 4.7 years found that 57% of study participants experienced a repeat infection (Niccolai et al., 2007). Clearly, repeat C. trachomatis infections comprise a substantial health burden among women and possibly higher than generally recognized.
6
Increasing Burden of C. trachomatis Infections?
Recently, there have been several reports of increases in the burden of C. trachomatis from countries with active screening programs. One of the earliest reports of this trend came from Sweden in 2002. Using both case reports and laboratory data, Gotz et al. reported an increase in positivity from 4.1% in 1994 to 5.4% in 1999 (Gotz et al., 2002). In the United States, similar trends have been reported. US data from women age 15–24 years in family planning clinics reported increases in 5 of 10 US Public Health Service planning regions (CDC, 2007). The most dramatic increases were observed in the region that includes the states of Alaska, Idaho, Oregon, and Washington, where widespread screening and treatment programs for chlamydia began in 1988. During the first 9 years of the program, from 1988 to 1996, chlamydia positivity among women age 15–24 declined from 10.3 to 4.0%, a greater than 60% decline (CDC, 2007). However, during the following 8 years, there was a 46% increase in chlamydia positivity from 3.9% in 1997 to 5.7% in 2004. Whether these trends represent a true increase in burden of C. trachomatis or reflect some other change is an important and unanswered question. In addition to a true increase in prevalence, increases in positivity could be due to improved laboratory testing technology reflected in the switch to using tests with higher sensitivities (NAATs), or a shift toward screening higher-risk women. Several studies have examined these potential effects to better understand the trend of increasing positivity. For example, in the Swedish study mentioned above, increasing positivity occurred both in labs that had and had not switched to more sensitive tests, therefore, the increase could not be attributed to greater detection due to improved laboratory techniques (Gotz et al., 2002). The authors concluded that the increase in the number of reported infections reflected ‘‘a real increase in prevalence.’’ In another analysis of data from the US, investigators were able to examine the risk characteristics of the tested population over time and to adjust for the type of test used (Fine et al., 2008). They observed that the proportion of tested women with demographic, clinical, or sexual risk behavior characteristics associated with an increased risk of chlamydial infection remained stable or decreased over time, and they adjusted chlamydia positivity to account for use of more sensitive laboratory tests. After carefully controlling for the potential effects of changing laboratory test characteristics and demographic and sexual risk behaviors of tested populations, they showed an independent 5% per year increase in chlamydia positivity over this 8-year period. Thus, the authors concluded that a true increase in chlamydia prevalence may have occurred. Additional studies provide further evidence of a true increase in C. trachomatis prevalence. Another explanation for a true increase that is supported by data from the experience in British Columbia, Canada. Since 2002, trends of increasing case reports have been observed (Brunham
1307
1308
74
Burden of Sexually Transmitted Chlamydia trachomatis Infections
et al., 2005). In 2005, Brunham et al. reported a decline in the case rate from 216/100,000 in 1991 to 104/100,000 in 1997, followed by an increase to 193/100,000 in 2003. They also reported a similar trend for repeat C. trachomatis infections, with an increase from approximately 2/100,000 during 1989–1995 to 53/100,000 in 1996. This change reflects a 4.6% increase per year since 1989. This increase in case reports that appears to be driven by increases in repeat infections suggests a possible explanation that has been termed the ‘‘arrested immunity hypothesis’’ (Brunham and Rekart, 2008). This hypothesis puts forth the idea that rising case rates are due to earlier detection and treatment of infection that interferes with the development of protective immune responses. In Canada, as in many parts of the developed world, control programs have involved aggressive screening efforts for early detection and treatment of infection since the 1980s. Earlier detection and treatment has likely reduced the duration of infection in many individuals, resulting in lowered immune response and lower antibody production. This, in turn, may have resulted in greater susceptibility to reinfection at the population level and hence the increased number of reported chlamydia cases. Other possible explanations for a true increase in prevalence could be changes in individual behaviors or sexual network dynamics. Reasons remain unclear at this point, and investigation into these important shifts should be a priority.
7
Burden of Associated Sequelae
7.1
Women Suffer Negative Long-Term Consequences as a Result of C. trachomatis Infections
The burden of sequelae associated with C. trachomatis infections are summarized in > Table 74-3. C. trachomatis infections are an important preventable cause of pelvic inflammatory disease (PID), chronic pelvic pain, tubal infertility, and ectopic pregnancies among women (Stamm, 1999). Lower genital tract infection with C. trachomatis frequently ascends to the upper reproductive tract where permanent damage (e.g., scarring) can occur. It is estimated that inadequately treated chlamydial infections may cause PID in 20–40% of . Table 74-3 Burden of sequelae associated with C. trachomatis infections among women Examples Health
Chronic pelvic pain Tubal scarring Pelvic inflammatory disease Infertility Ectopic pregnancy
Psychosocial
Stigma Concern about health effects Anxiety about sex partner reaction
Economic
Estimated 1.5 billion (US 1994)
HIV risk
2- to 5-fold increased risk of HIV acquisition
Burden of Sexually Transmitted Chlamydia trachomatis Infections
74
women (Stamm, 1999), and that among women with PID, 20% may develop tubal infertility (Westrom et al., 1992). There is evidence that the long-term negative health effects of genital C. trachomatis are more closely linked to repeat infections than to initial infections. Epidemiologic studies have also shown increased risk for ectopic pregnancies and PID among women with repeat infections. In one study, the risk for PID among women with two infections was increased 4-fold compared to women with one infection, and the risk among women with three or more infections was increased 6-fold (Hillis et al., 1997). Women also suffer immediate psychological consequences after a C. trachomatis diagnosis, including stigma associated with STI, uncertainty about reproductive health effects, and anxiety about partners’ reactions to the diagnosis (Duncan et al., 2001).
7.2
C. trachomatis is Associated with High Economic Burden
The economic burden associated with C. trachomatis is substantial. In 1994 US dollars, it was estimated that the direct cost of C. trachomatis infections was $1.5 billion (IOM, 1997). This estimate includes the cost of health care services, laboratory services, treatment for infections and associated sequelae. The direct cost of pelvic inflammatory disease which is often caused by C. trachomatis is $3.1 billion.
7.3
Infection with C. trachomatis Causes Increased HIV Transmission and Acquisition
Several studies have documented increased HIV risk associated with C. trachomatis infections (Wasserheit, 1992). The increased risk is often estimated to be 2- to 5-fold. For example, in a study among female sex workers in Africa, presence of C. trachomatis infection was associated with a 3.6-fold increase in likelihood of HIV seroconversion after controlling for potential confounding by sexual behavior (Laga et al., 1991). It is also noteworthy that the proportion of HIV infections attributable to non-ulcerative STI like C. trachomatis is likely to be greater than the proportion attributable to ulcerative STI despite lower relative risk because of their higher prevalence in the population.
8
Conclusion
In sum, it is clear from a large and growing body of evidence that C. trachomatis infections constitute an important public health problem due to their burden, disproportionate impact on vulnerable populations, and their sequelae. The possibility that the burden of infections is increasing in countries that have been implementing screening programs for control is also of concern. Continued efforts to understand biological and behavioral risks and the larger social and cultural contexts in which infections are transmitted are critical, and applying this knowledge to prevention and control is a public health priority. Improved access to screening and treatment in developing countries is also needed to reduce the high burden in these settings. Reducing the burden of C. trachomatis infections will have a positive impact on women’s reproductive health in particular and the overall health of general populations as well.
1309
1310
74
Burden of Sexually Transmitted Chlamydia trachomatis Infections
Summary Points C. trachomatis is the most common bacterial sexually transmitted disease in the world, and therefore these infections constitute an enormous public health problem.
The true burden of C. trachomatis infections is unknown due to substantial number of undiagnosed infections and/or under-reporting.
The burden of C. trachomatis is especially high in resource poor countries, perhaps due to limited access to care and treatment.
In developed countries, C. trachomatis infections are most common among adolescents and young adults, and among racial/ethnic minorities.
Women bear a disproportionate burden of negative health consequences of C. trachomatis infections including pelvic inflammatory disease, chronic pelvic pain, and infertility.
Additional burdens include high cost on the health care system and increased HIV transmission.
Recent evidence of increasing C. trachomatis prevalence has been reported.
References Bakken IJ, Norbdo SA, Skjeldestad FE. (2006). Sex Transm Dis. 33: 26–30. Bernstein KT, Chow JM, Ruiz J, Schachter J, Horowitz E, Bunnell R, Bolan G. (2006). Am J Public Health. 96: 1862–1866. Blythe MJ, Katz BP, Batteiger BE, Ganser JA, Jones RB. (1992). J Pediatrics. 121: 487–493. Brunham RC, Pourbohloul B, Mak S, White R, Rekart ML. (2005). J Infect Dis. 192: 1836–1844. Brunham RC, Rekart ML. (2008). Sex Transm Dis. 35: 53–54. Burstein GR, Waterfield G, Joffe A, Zenilman JM, Quinn TC, Gaydos CA. (1998). Sex Transm Dis. 25: 395–402. Centers for Disease Control and Prevention. (2007). US Department of Health and Human Services, Atlanta GA. Cohen DA, Nsuami M, Martin DH, Farley T. (1999). Pediatrics. 104: 1281–1285. Debattista J, Clementson C, Mason D, Dwyer J, Argent S, Woodward C, Dean J, Buks L, Copley M, Hinwood G, Benfield C, Walton P. (2002). Sex Transm Dis. 29: 216–221. Duncan B, Hart G, Schoular A, Bigrigg A. (2001). BMJ. 322: 195–199. Ehrhardt AA, Bolan G, Wasserheit JN. (1999). In: Holmes KK et al. (eds.) Sexually Transmitted Diseases. McGraw-Hill, New York. Fenton KA, Copas A, Mitchell K, Elam G, Carder C, Ridgway G, Wellings K, Erens B, Field J, Johnson AM. (2001). Sex Transm Infect. 77: 194–198.
Fine D, Dicker L, Mosure D, Berman S. (2008). Sex Transm Dis. 35: 47–52. Fortenberry JD, Brizendine EJ, Katz BP, Wools KK, Blythe MJ, Orr DP. (1999). Sex Trans Dis. 26: 26–32. Franceschi S, Smith J, Brule van den A, Herrero R, Arslan A, Anh PTH, Bosch FX, Hieu NT, Matos E, Posso H, Qiao YL, Shin HR, Sukvirach S, Thomas JO, Snijders PJF, Munoz N, Meijer CJLM. (2007). Sex Transm Dis. 34: 563–569. Gaydos CA, Kent CK, Rietmeijer CA, Willard NJ, Marrazzo JM, Chapin JB, Dunne EF, Markowitz LE, Klausner JD, Ellen JM, Schillinger JA. (2006). Sex Transm Dis. 33: 314–319. Geisler WM, Chyu L, Kusunoki Y, Upchurch DM, Hook EW III. (2006). Sex Transm Dis. 33: 389–396. Gerbase AC, Rowley JT, Heymann DHL, Berkley SFB, Piot P. (1998). Sex Transm Inf. 74: S12–S16. Golden MR, Whittington WLH, Handsfield HH, Hughes JP, Stamm WE, Hogben M, Clark A, Malinski C, Helmers JRL, Thomas KK, Holmes KK. (2005). N Eng J Med. 352: 676–685. Gotz H, Linback J, Ripa T, Arneborn M, Ramstedt K, Ekdahl K. (2002). Scand J Infect Dis. 34: 28–34. Griesinger G, Gille G, Klapp C, von Otte S, Diedrich K. (2007). Clin Microbiol Infect. 13: 436–439. Hillis SD, Owens LM, Marchbanks PA, Amsterdam LE, MacKenzie WR. (1997). Am J Obstet Gynecol. 176: 103–107. Institute of Medicine. (1997). National Academy Press, Washington DC. Joesoef MR, Mosure DJ. (2006a). Sex Transm Dis. 33: 571–575.
Burden of Sexually Transmitted Chlamydia trachomatis Infections Joesoef MR, Mosure DJ. (2006b). Sex Transm Dis. 33: 636–639. Kahn RH, Mosure DJ, Blank S, Kent CK, Chow JM, Boudov MR, Brock J, Tulloch S. (2005). Sex Transm Dis. 32: 255–259. Klausner JD, McFarland W, Bolan G, Hernandez FM, Lemp GF, Cahoon-Young B, Morrow S, Ruiz J. (2001). J Infect Dis. 183: 1087–1092. Laga M, Nzila N, Goeman J. (1991). AIDS. 5: S55–S63. LaMontagne SD, Baster K, Emmett L. (2005). Determinants of chlamydia re-infection: the role of partner change and treatment [abstract]. In: Program and Abstracts of the 16th Biennial Meeting of the International Society for Sexually Transmitted Diseases Research, Amsterdam, The Netherlands, Abstract TO-404. Laumann EO, Youm Y. (1999). Sex Transm Dis. 26: 250–261. Lee V, Tobin JM, Foley E. (2006). J Fam Plann Reprod Health Care. 32: 104. Marrazzo JM, White CL, Krekeler B, Celum CL, Lafferty WE, Stamm WE, Handsfield HH. (1997). Ann Intern Med. 127: 796–803. McKee MD, Fletcher J, Schechter CB. (2006). J Adolesc Health. 39:183. Menon-Johansson AS, Winston A, Matthews G, Portsmouth S, Daniels D. (2005). Int J STD AIDS. 16: 799–801. Miller WC, Ford CA, Morris M, Handcock MS, Schmitz JL, Hobbs MM, Cohen MS, Harris KM, Udry JR. (2004). JAMA. 291: 2229–2236. Niccolai LM, Ethier KA, Kershaw TS, Lewis JB, Ickovics JR. (2003). Am J Obstet Gynecol. 188: 63–70. Niccolai LM, Hochberg AL, Ethier KA, Lewis JB, Ickovics JR. (2007). Arch Pediatr Adol Med. 161: 246–251.
74
Quinn TC, Gaydos C, Shepherd M, bobo L, Hook EW III, Viscidi R, Rompalo A. (1996). JAMA. 276: 1737–1742. Renton A, Thomas BM, Gill S, Lowndes C, TaylorRobinson D, Patterson K. (2006). Int J STD AIDS. 17: 443–447. Rietmeijer CA, Van Bemmelen R, Judson FN, Douglas JM Jr. (2002). Sex Transm Dis. 29: 65–72. Rowhani-Rahbar A, Niccolai LM, Dunne DW, Green S, Jenkins H, Khoshnood K. (2006). Int J STD AIDS. 17: 453–458. Shrier LA. (2004). Adolesc Med Clin 15: 215. Stamm WE (1999). In: Holmes KK (eds.) et al. Sexually Transmitted Diseases. McGraw-Hill, New York, NY. Thrall JS, McCloskey L, Ettner SL, Rothman E, Tighe JE, Emans SJ (2000). Arch Pediatr Adolesc Med. 154: 885. van Valkengoed IGM, Morre SA, Brule van den AJC, Meijer CJLM, Deville W, Bouter LM, Boeke AJP. (2000). Sex Transm Infect. 76: 375–380. Wasserheit JN. (1992). Sex Trans Dis. 19: 61–77. Weinstock H, Berman S, Cates W Jr. (2004). Perspect Sex Reprod Health. 36: 6–10. Westrom L, Joesoef R, Reynolds G, Hagdu A, Thompson SE. (1992). Sex Transm Dis. 19: 185–192. Wilson JS, Honey E, Templeton A, Paavonen J, Mardh P-A, Stary A, Stray-Pedersen B. (2002). Hum Reprod Update. 8: 385–394. Whittington WL, Kent C, Kissinger P, Oh MK, Fortenberry JD, Hillis SE, Litchfield B, Bolan GA, St. Louis ME, Farley TA, Handsfield HH. (2001). Sex Transm Dis. 28: 117–23. World Health Organization Department of HIV/AIDS. (2007). http:/www.who.int/docstore/hiv/GRSTI/003. html.
1311
75 Disease Burden from Group A Neisseria meningitidis Meningitis in Hyperendemic Countries of the African Meningitis Belt C. Suraratdecha . C. Levin . F. M. LaForce 1 Group A Neisseria meningitidis Meningitis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1314 2 Incidence, Case-Fatality, and Disability Rates of Group A Meningococcal Meningitis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1316 3 Disease Burden Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1317 4 Economic Burden of Group A Meningococcal Meningitis . . . . . . . . . . . . . . . . . . . . . . . . . 1318 5 Disease Burden Estimates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1318 6 Methodological and Data Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1318 7 Applications of the Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1320 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1321
#
Springer Science+Business Media LLC 2010 (USA)
1314
75
Disease Burden from Group A Neisseria meningitidis
Abstract: Neisseria meningitidis meningitis is a devastating illness characterized by the sudden onset of intense headache, high fever, nausea, vomiting, photophobia, and stiff neck. The disease has a high mortality rate and is associated with long-term neurological defects such as deafness and recurrent seizures. > Epidemics of group A meningococcal meningitis continue to pose an important public health problem for sub-Saharan Africa. The last major epidemic occurred in 1996–1997, with more than 180,000 reported cases and 20,000 deaths. During 2006 and 2007 Burkina Faso suffered more than 45,000 cases of group A meningococcal meningitis. Using the population-based incidence and bacteriologic data from Niger to estimate disease burden in seven hyperendemic countries (Mali, Burkina Faso, Niger, Nigeria, Chad, Sudan, and Ethiopia), over a 10-year period, group A N. meningitidis is estimated to cause about 1.1 million cases of meningitis, 133,000 deaths, 317,000 disabilities, and 12 million disability-adjusted life years lost in the population 0- to 40-year-old > cohorts. Over the last few years a major effort has been under way to control epidemic meningococcal disease in Africa through the development, testing, licensure, and introduction of new conjugate meningococcal > vaccines. The Meningitis Vaccine Project, a partnership between PATH and the World Health Organization, is developing an affordable meningococcal A > conjugate vaccine that will be introduced at public health scale in 2009. Wide-scale introduction of the meningococcal A conjugate vaccine is expected to eliminate these epidemics. List of Abbreviations: CFR, > case-fatality rate; DALYs, disability-adjusted life years; Men A, meningococcal A; Men Ps, meningococcal polysaccharide; PCR, > polymerase chain reaction; WHO, World Health Organization; WHO-CHOICE, choosing interventions that are cost effective; YLD, years lost due to disability; YLL, years of life lost
1
Group A Neisseria meningitidis Meningitis
Bacterial meningitis continues to be a major health problem in sub-Saharan Africa. Greenwood (2004) estimated that 200,000 cases and 70,000 deaths occur annually from bacterial meningitis due to Haemophilus influenzae, Streptococcus pneumoniae, and Neisseria meningitidis in African children under 5 years of age. Of these three agents, only the meningococcus has the ability to cause large epidemics, particularly in the African meningitis belt (Mar et al., 1979; Makela et al., 1992; Moore, 1992). Meningococcal meningitis is an infection of membranes surrounding the brain and spinal cord and presents clinically with acute onset of intense headache, high fever, nausea, vomiting, photophobia, and stiff neck (> Table 75-1). The disease carries a high mortality rate, particularly if treatment is delayed, and long-term neurological defects. The first outbreak of meningococcal meningitis was described in Geneva, Switzerland, in 1805, but it was not until 1887 that N. meningitidis was isolated and identified as the causative agent. The first outbreak in Africa occurred in soldiers in Algiers in 1840 (Chalmers and O’Farrell, 1916). The disease appeared in North Africa and the sub-Saharan region of Africa in the 1880s in an area that became known as the African meningitis belt, which extends from Ethiopia in the east to Senegal in the west (> Figure 75-1; WHO, 1998). For more than 100 years, sub-Saharan Africa has suffered periodic meningitis epidemics in this so-called ‘‘meningitis belt.’’ The human toll from these epidemics has been enormous; for example, the 1996 epidemic resulted in more than 180,000 reported cases and 20,000 deaths. The primary goal of this chapter is to provide an estimate of the disease burden due to group A Neisseria meningitidis in the hyperendemic countries of the African meningitis belt.
1315
75
Disease Burden from Group A Neisseria meningitidis
. Table 75-1 Key features of acute bacterial meningitis in sub-Saharan Africa 1. Bacterial meningitis is a severe acute infection of the membranes lining the brain and spinal cord 2. Bacterial meningitis is a very serious infection; despite antimicrobial treatment, mortality rates are about 10% 3. Most bacterial meningitis is by three bacteria: pneumococci, meningococci, and Haemophilus influenzae 4. Only meningococci are capable of causing large epidemics 5. Sub-Saharan Africa is the home for 95% of global systemic meningococcal infections 6. Over 85% of meningococcal isolates in sub-Saharan Africa belong to one group, group A 7. There is no group A meningococcal disease in Europe or North America 8. The currently available vaccine to protect against meningococci is meningococcal polysaccharide 9. Conjugate vaccines have been shown to induce a better and longer-lasting immune response than polysaccharide vaccines. Conjugate vaccines against group A meningococci are being developed This table lists the key facts of bacterial meningitis in sub-Saharan Africa due to meningococci
. Figure 75-1 African meningitis belt countries. The figure illustrates hyperendemic countries and non-hyperendemic countries in the African meningitis belt where group A meningococcal meningitis has been reported, and their population size
1316
75
Disease Burden from Group A Neisseria meningitidis
The chapter will review published data and present a method used to estimate the disease burden of group A meningococcal meningitis in hyperendemic countries. Estimates will be summarized, as well as limitations of the method.
2
Incidence, Case-Fatality, and Disability Rates of Group A Meningococcal Meningitis
Quantifying the meningococcal disease burden in Africa is not a simple task. While there have been many descriptive studies of meningitis in sub-Saharan Africa, only a few are population based and even fewer link bacteriologic information with individual cases over time. Most published data are hospital based and/or descriptions of outbreaks of meningococcal meningitis. These epidemics occur over a relatively short period of time and are characterized by very high attack rates. For example, using the data from a sentinel site in Khartoum, Sudan, Salih et al. (1990) reported an annual incidence rate of 1,679 per 100,000 inhabitants during a large epidemic due to group A N. meningitidis (February–August 1988). About a third of the cases were under 9 years old. In the Savannes region of Togo between December 1996 and May 1997, the cumulative attack rate from group A N. meningitidis was 581 per 100,000 population (Aplogan et al., 1997). The Upper East, Upper West, and Northern regions of Ghana reported an attack rate of meningococcal disease of 550 per 100,000 population at the end of the epidemic in 1997 (Woods et al., 2000). Even in the absence of dramatic epidemics the > endemic case rates are high when compared to Europe and the United States. Data from the Bobo-Dioulasso region of Burkina Faso collected during the periods of May 2002–April 2003 and March 2004–February 2005 showed that annual incidence rates of group A meningococci were 2.7, 6.3, 7.8, and 2.6 per 100,000 in children under 1 year, 1–4 years, 5–14 years, and 15 years and older, respectively (Traore et al., 2006). The attack rates in various areas of Zaria, Northern Nigeria, from March to May 1977 ranged from 0 to 11.9 per 1,000 inhabitants, with an average of 3.6 per 1,000 (Greenwood et al., 1979). Using the retrospective data from 16 health centers in two Northern districts (Atacora and Donga) of Benin from 1998 to 2001, Fourn et al. (2004) estimated an increase in crude incidence rates from 85 to 567 per 100,000 in Atacora and 71 to 619 per 100,000 in Donga. During the huge 1996 outbreak in Nigeria with more than 115,000 cases, almost three quarters of the patients were under 15 years; 11% were above 30 (Mohammed et al., 2000). These data show the enormous variability in rates, depending largely on whether data were collected during an epidemic. The only comprehensive population-based analysis of bacterial meningitis in the African meningitis belt was done at Centre de Recherche Me´dicale et Sanitaire (CERMES) in Niamey, Niger, after the meningitis epidemic in 1995–1996 (Campagne et al., 1999). The bacteriology laboratory at CERMES is the national reference laboratory for Niger, with more than 30 years of experience in the bacteriologic analysis of > cerebrospinal fluid specimens. Data are rigorously maintained during epidemic (1994–1995) and inter-epidemic (1981–1993 and 1996–1997) periods and reviewed on a regular basis. Baseline incidence rates are high in the population 0–29 years of age, and during epidemics the rates soar ten fold in this age group. About five percent of all cases of meningococcal meningitis in epidemic and inter-epidemic periods occur in infants, while 90% of cases occur in 1- to 29-year-olds. The incidence of group A N. meningitidis varies annually. While the crude meningitis incidence rate is 100.8 per 100,000 per year, the rate for N. meningitidis is 55.3 cases per 100,000 per year, with 85.6% of meningococcal infections due to Serogroup A (> Table 75-2).
1317
75
Disease Burden from Group A Neisseria meningitidis
. Table 75-2 Incidence rates per 100,000, case-fatality and disability rates (%) of group A Neisseria meningitidis in Niamey, Niger (1981–1996) Total per 100,000
Age (years) Table 75-2. All 0- to 40-years age cohorts were followed over the 10-year period to account for variation in incidence rates during epidemic and inter-epidemic years. The model assumed that an epidemic cycle occurs every 10 years. Since the model used the population data in 2009 and the last major epidemic occurred in 1996–1997, the model replaced the age-specific incidence rates in the 8th year in the cycle in the model by epidemic rates. The susceptible population was adjusted over time for infections and deaths from other causes using the age-specific mortality rates to avoid double counting of meningococcal meningitis cases. Disability-adjusted life years (DALYs) were calculated as the sum of the years of life lost (YLL) and years lost due to disability (YLD) associated with chronic sequelae associated with deafness, seizure disorder, and motor deficits (Lopez et al., 2001). YLLs were estimated from the WHO life tables for hyperendemic countries (WHO, 2006). Age-specific mortality rate was calculated and weighted average among seven hyperendemic countries. The disability weight for average chronic sequelae due to meningitis of 0.183 was used in YLD calculation.
1318
75 4
Disease Burden from Group A Neisseria meningitidis
Economic Burden of Group A Meningococcal Meningitis
The medical cost associated with meningococcal meningitis was projected over a 10-year period, consisting of diagnostic studies, outpatient care, and inpatient care. All cost estimates are presented in 2007 US dollars. The cost of treating a case of meningitis is $80 and is based on a Ministry of Health study conducted in Burkina Faso during the 2002 meningitis epidemic (Ministry of Health and Burkina Faso, 2003) and the Choosing Interventions that are Cost Effective (WHO-CHOICE) model. The total costs per case at the district level ranged from $67 to 90 and included initial consultation visit, the hospitalization and costs of supporting the hospitalized family member, laboratory, and transportation costs. Using the WHO-CHOICE model (WHO, 2002), and assuming the ratio of urban to rural patients is 3:1, an outpatient visit in the urban area costs $10.42 and each hospital day costs $32.08 per day. In the rural area, one outpatient visit costs $6.14 and an inpatient day costs $18.85 per day. The model further assumes that all cases were seen (outpatient), all urban cases were hospitalized, and only one third of the total rural patients were hospitalized. Cost of antibiotics for the non-hospitalized cases is $10. An average estimated cost is $77.68 per case. To estimate the diagnostic costs, the following guidelines were followed. The clinical guideline in Burkina Faso is to perform a lumbar puncture in all suspected meningitis cases in endemic years. In the model, an average cost of diagnostic is $15. It is assumed that lumbar puncture is performed in 50% of endemic cases and in 10% of epidemic cases. A lumbar puncture kit costs $2, Gram stain costs $0.60, latex agglutination costs $15.60, and polymerase chain reaction (PCR) costs $6 for a negative sample or $12 for a positive sample (Chanteaux, 2005, unpublished).
5
Disease Burden Estimates
The total > high-risk group (up to 40 years old) in hyperendemic countries in 2009 is projected at 201 million. Extrapolation of the Niger data and adjustment of age-specific incidence rates during endemic and epidemic years to reflect the epidemiological pattern of meningococcal meningitis that appears in cycles of approximately 8–12 years produced a burden estimate of group A meningococcal meningitis of 1,105,836 cases over the 10-year period. When case-fatality rates, disability rates, and DALY formula were applied to the total case estimates at 0% discount rate, 133,586 deaths, 316,657 disabilities, and 12,172,860 DALYs were derived. The estimate of disease burden from group A meningococcal meningitis is 110,584 cases annually, ranging from 44,409 during non-epidemic years to 669,835 in epidemic years (year 8 in > Figure 75-2). The decennial estimate of economic cost as a result of these cases is $111,213,294 (or $11 million per year, on average) from treatment, laboratory, and diagnostic testing costs. > Table 75-3 summarizes disease burden and economic burden of group A meningococcal meningitis in discounted and undiscounted values.
6
Methodological and Data Considerations
Epidemic meningitis is a major public health problem in sub-Saharan Africa. Virtually all of the meningitis epidemics with incidence rates greater than 100 per 100,000 have been caused by a specific pathogen, group A Neisseria meningitidis. Despite the general acceptance of the
1319
75
Disease Burden from Group A Neisseria meningitidis
. Figure 75-2 Group A meningococcal meningitis cases by year. This figure shows the number of group A Neisseria meningitidis cases estimated from the disease burden model by year. The model assumed that the epidemic would occur in year 8 of the 10-year cycle
. Table 75-3 Estimated disease burden and economic burden of group A Neisseria meningitidis in hyperendemic countries over the 10-year period YLL
YLD
DALYs
Economic burden (USD)
0% discount
8,404,432
3,768,428
12,172,860
111,213,294
3% discount
4,409,696
1,963,576
6,373,272
109,982,969
This table provides the estimates of discounted (3% discount rate) and undiscounted YLL, YLD, DALYs, and economic burden due to group A N. meningitidis in population 0–40 years old over a 10-year period. DALYs disability- adjusted life years; YLD years lost due to disability; YLL years of life lost
public health importance of epidemic meningitis, ascertaining and quantitating the meningococcal disease burden is not simple. There have been many descriptive studies of meningitis in sub-Saharan Africa, but, as mentioned earlier, only a few are population-based with bacteriologic confirmation of cases, and there was only one study that spanned epidemic and interepidemic periods. Meningitis surveillance in African meningitis belt countries varies greatly in quality. For the most part, African surveillance systems can identify major epidemics of meningococcal meningitis but cannot precisely determine either the total number of cases during an epidemic or the number of meningitis cases during inter-epidemic periods. Country surveillance data are often fragmented, poorly codified, and of limited value to epidemiologists. Some countries such as Burkina Faso and Niger have well developed systems that rapidly identify meningitis epidemics; many others have limited capabilities. To improve meningitis surveillance, major efforts have been made in recent years, and beginning in 2003, the Meningitis Vaccine Project has sponsored an effort at WHO’s Multi Disease Surveillance Center in Ouagadougou to collect, codify, and analyze meningitis data from Niger, Burkina Faso, and Mali. Comprehensive sets of data are now available from Niger (1986–2004), Burkina Faso (1996–2004), and
1320
75
Disease Burden from Group A Neisseria meningitidis
. Figure 75-3 Burkina Faso meningitis cases, 1996–2007. This figure shows the annual reported cases of meningitis in Burkina Faso from 1996 to 2007 (Ministry of Health and Burkina Faso, 2003, WHO)
Mali (1992–2004). > Figure 75-3 shows the meningitis incidence data for Burkina Faso from 1996 to 2007. These data clearly show the seasonal distribution and the wave-like nature of meningococcal epidemics in these countries. Epidemics occur across 2 or 3 years, with intervals of quiescence that last 5–7 years. Incidence rates during epidemics can soar to more than 500/100,000 population. The temporal characteristics of epidemics are very predictable; the hot, dry, and dusty conditions during the first 4 months of the calendar year serve as the climatologic backdrop for the annual burst of cases of meningococcal meningitis. While these new data are a major improvement, they do not provide information on disease rates by pathogen because of the absence of case-based surveillance data that link individual cases with bacteriologic data.
7
Applications of the Model
The analytic approach described in this paper is one such method that can be used to demonstrate the impact of disease control strategy through estimating the burden of group A meningococci that could be averted by implementing vaccination strategies. Over the last 20 years, control of epidemic meningitis has emphasized surveillance and reactive mass > immunizations with the meningococcal polysaccharide (Men Ps) vaccine. A successful reactive vaccination strategy requires major investments in infrastructure and resources such as real-time surveillance data, accurate bacteriologic information, availability of vaccine, and rapid implementation of district-based vaccinations. More often than not, immunizations are given in the latter phase or after the epidemic with discouraging results. Because Men Ps vaccine is ineffective in infants and toddlers under 2 years, and has no effect on colonization, these emergency mass vaccination campaigns must be repeated on a regular basis. Currently, it is estimated that meningitis belt countries spend about US $20 million annually responding to meningococcal outbreaks. Through an innovative partnership, the Meningitis Vaccine Project has developed an affordable (US $0.40/dose) Men A conjugate vaccine manufactured at the Serum Institute of India (LaForce et al., 2007). Conjugate vaccines are more immunogenic than > polysaccharide
75
1321
Disease Burden from Group A Neisseria meningitidis
vaccines, they prime immunological memory, can be confidently used in children under 1 year of age, and have been shown to induce herd immunity. The vaccine will be given as a single dose to all 1- to 29-year-olds, a strategy that is expected to generate strong herd immunity and rapid control of group A N. meningitidis infections. Introduction at public health scale is expected in 2009. It is important for countries to be aware of current disease burdens so that they can estimate the cost and potential savings that may accrue with the introduction of more potent vaccines such as Men A conjugate vaccine.
Summary Points Group A Neisseria meningitidis is the most important cause of epidemic meningitis in subSaharan Africa.
Reactive vaccination strategies with meningococcal
> polysaccharide vaccine have not eliminated epidemics. There are limited population-based data on the incidence and the etiology of bacterial meningitis in Africa. The most complete data set is from Niger (1981–1996), which shows a meningococcal group A incidence rate of about 50 cases per 100,000; comparable incidence rates for all meningococcal infections in the United States is 0.3 cases per 100,000. The model predicts that, over 10 years, there are more than 1.1 million cases of meningococcal group A meningitis. Introduction of a new conjugate meningococcal vaccine in mass vaccination campaigns of high-risk population (1- to 29-year-olds) is expected to generate herd immunity and eliminate these epidemics.
References Aplogan A, Batchassi E, Yakoua Y, Croisier A, Aleki A, Schlumberger M, Molina S, Sidatt M, Kaninda AV. (1997). Sante. 7(6): 384–390. Bovier P, Wyss K, Au H. (1999). Soc Sci Med. 48(9): 1205–1220. Campagne G, Schuchat A, Djibo S, Ousse´ini A, Cisse´ L, Chippaux JP. (1999). Bull World Health Organ. 77 (6): 499–508. Chalmers AJ, O’Farrell WR. (1916). J Trop Med Hyg. 29: 101–116, 117–129. Fourn L, Makoutode´ M, Ouendo M, Tounkara B. (2004). Sante. 14(3): 153–159. Greenwood B. (2004). In: Parry E (ed.) Principles of Medicine in Africa, 3rd ed. Cambridge University Press, Cambridge, UK, pp. 305–315. Greenwood BM, Bradley AK, Cleland PG, Haggie MH, Hassan-King M, Lewis LS, Macfarlane JT, Taqi A, Whittle HC, Bradley-Moore AM, Ansari Q. (1979). Trans R Soc Trop Med Hyg. 73(5): 557–562.
LaForce FM, Konde K, Viviani S, Preziosi MP. (2007). Vaccine. 25 (Suppl. 1): A97–A100. Lopez AD, Salomon J, Ahmad O, Murray CJL, Mafat D. (2001). Life Tables for 191 Countries: Data, Methods, and Results. Global Programme on Evidence for Health Policy Discussion Paper Series: No. 9. World Health Organization, Geneva. Makela PH, Takala AK, Peltola H, Eskola J. (1992). J Infect Dis. 165 (Suppl. 1): S2–S6. Mar ID, Denis F, Cadoz M. (1979). Pathol Biol. (Paris) 27 (9): 543–548. Ministry of Health, Burkina Faso. (2003). Recherche operationelle sur la gratuite de la prise en charge des patients atteints de meningite cerebro-spinale au Burkina Faso. Mohammed I, Nasidi A, Alkali AS, Garbati MA, AjayiObe EK, Audu KA, Usman A, Abdullahi S. (2000). Trans R Soc Trop Med Hyg. 94(3): 265–270. Moore PS. (1992). Clin Infect Dis. 14(2): 515–525.
1322
75
Disease Burden from Group A Neisseria meningitidis
Parent du Chaˆtelet I, Gessner BD, da Silva A. (2001). Vaccine. 19(25–26): 3420–3431. Salih MA, Ahmed HS, Karrar ZA, Kamil I, Osman KA, Palmgren H, Hofvander Y, Olce´n P. (1990). Scand J Infect Dis. 22(2): 161–170. Traore Y, Njanpop-Lafourcade BM, Adjogble KL, Lourd M, Yaro S, Nacro B, Drabo A, Parent du Chaˆtelet I, Mueller JE, Taha MK, Borrow R, Nicolas P, Alonso JM, Gessner BD. (2006). Clin Infect Dis. 43(7): 817–822. United Nations Population Division. (2007). World Population Prospects: The 2006 Revision. United Nations, New York. Woods CW, Armstrong G, Sackey SO, Tetteh C, Bugri S, Perkins BA, Rosenstein NE. (2000). Lancet. 355 (9197): 30–33.
World Health Organization. (1998). Control of Epidemic Meningococcal Disease. WHO Practical Guidelines, 2nd ed. WHO, Geneva. http://www.who.int/csr/ resources/publications/meningitis/whoemcbac983. pdf. Accessed 7 Mar 2008. World Health Organization. (2002). Statistical Information System (WHOSIS). Africa D Total population, births and mortality rates; Unit costs for base case analysis: WHO African Region – D, Country specific hospital costs, WHO CHOICE. World Health Organization. (2006). Life Tables for WHO Member States. WHO, Geneva. http://www. who.int/whosis/database/life_tables/life_tables.cfm. Accessed 7 Mar 2008.
76 DALYs in Chronic Hepatitis C : A Paneuropean Perspective U. Siebert . A. Conrads-Frank . R. Schwarzer . B. Lettmeier . G. Sroczynski . S. Zeuzem . N. Mu¨hlberger 1 Key Facts Hepatitis C Virus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1324 2 Review of Existing DALY Data and Calculation of the Contribution of Cirrhosis and Liver Cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1326 3 Revised Evaluation of the HCV Burden of Disease after Inclusion of Cirrhosis and Liver Cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1331 4 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1332 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1333
#
Springer Science+Business Media LLC 2010 (USA)
1324
76
DALYs in Chronic Hepatitis C : A Paneuropean Perspective
Abstract: > Hepatitis C virus (HCV) is one of the leading causes of chronic liver disease with serious sequelae such as end-stage cirrhosis and liver cancer. However, few burden of disease data that consider health-related quality of life are available for Europe. The objective of this evaluation was to summarize available data on the HCV-related burden of disease under consideration of the reduction in quality of life for countries of the WHO European region. We analyzed available data on disability-adjusted life years (> DALY) and life years lost to disability (> YLD). Literature and international health databases were systematically searched for HCV-related data on DALY and YLD. If no HCV-specific data were available, these were calculated via HCV-attributable fractions. HCV-specific DALY data were only available for hepatitis C without cirrhosis and liver cancer and for all cause cirrhosis and liver cancer. The calculation via HCV-attributable fractions yielded that approximately 1.2 million DALYs were lost due to HCV in 2002 in the WHO European region, of which about one sixth can be attributed to quality-of-life impairment. About 95% of the DALYs were accumulated by patients in advanced disease stages. Our results emphasize the public health relevance of hepatitis C and highlight the importance and the potential benefit of preventive antiviral treatment. The variation of DALY figures indicate the potential of inequality of health services across the countries of the WHO European region that should be investigated further. The finding that HCV-specific data are not available suggests that the crisis is not yet fully perceived on the European level. List of Abbreviations: AF, Attributable fractions, in this chapter used for the fraction of cases of liver cirrhosis and liver cancer, that are caused by HCV infection; > ALT, Serum alanine aminotransferase. A liver enzyme that is an indicator of liver disease; DALY, > Disability adjusted life years; GBD, WHO Global Burden of Disease Project (GBD); HBV, > Hepatitis B virus; HCV, Hepatitis C virus; > ICD-10, The International Statistical Classification of Diseases and Related Health Problems 10th; ICD-10 Code B20-B24, ICD-10 Code for human immunodeficiency virus (HIV) disease; ICD-10 Code C16, ICD-10 Code for malignant neoplasm of the stomach; RNA, Ribonucleic acid; > SF-36, Short Form 36 (health survey questionnaire), a generic quality-of-life profile instrument; WHO, World Health Organization; YLD, > Life years lost to disability; > YLL, Years of life lost from premature death
1
Key Facts Hepatitis C Virus
Hepatitis C virus (HCV) discovered in 1989. Hepatitis C is a leading cause of chronic liver disease with life-threatening sequelae such as end-stage liver cirrhosis and liver cancer. 15–25% of HCV infections progress to severe liver disease. Progression is often silent, and therefore, many cases are diagnosed at a late stage, when therapeutic options are already limited (‘‘Silent Killer’’). In late stages, liver transplantation is the only therapeutic option. With early detection, antiviral treatment can prevent about 60% of complications. Pegylated interferon combined with ribavirin is the state of the art treatment. Transmission is via blood to blood contact. Infections decreased with introduction of routine blood screening in 1991. Many patients have been infected prior to the 1990s (‘‘Awakening Giant’’).
DALYs in Chronic Hepatitis C : A Paneuropean Perspective
76
The hepatitis C virus (HCV) was discovered in 1989 and identified as one of the leading causes of chronic liver disease with serious sequelae such as end-stage cirrhosis and liver cancer (Lauer and Walker, 2001). According to expert estimates, HCVaccounts for 20% of cases of acute hepatitis, 70% of cases of chronic hepatitis, 40% of cases of end-stage cirrhosis, 60% of cases of hepatocellular carcinoma, and 30% of liver transplants (European Association on the Studies on the Liver, 1999) in industrialised countries. The World Health Organization (WHO) has estimated that as many as two thirds of the liver transplants in the developed world are due to HCV infection (World Health Organization, 1999a). In the US, a consensus conference named HCVas the primary reason for liver transplantation (National Institute of Health, 2002). These figures show that hepatitis C causes a major public health problem and poses a huge burden of disease on society. The burden of disease results from premature death through severe liver disease and from impairment in quality of life. Quality of life was initially assumed to be affected primarily in the minority of patients with advanced disease, where HCV may cause end-organ damage in a number of different organ systems. The vast majority of patients with only mild liver disease were described as ‘asymptomatic’ and their quality of life was assumed to be unaffected by the infection (Anonymous, 1997; European Monitoring Centre for Drugs and Drug Addiction, 2004; Spiegel et al., 2005). This position has been revised since a number of studies revealed that HCV per se causes a broad array of symptoms and diminishes health related quality of life even in the absence of advanced liver disease (Foster et al., 1998; Siebert et al., 2003; Siebert et al., 2005; Spiegel et al., 2005; Hollander et al., 2006; Lang et al., 2006; von Wagner et al., 2006). As Spiegel et al. summarize in a systematic review of quality of life studies, patients with HCV score lower than uninfected controls across all scales of the Short Form 36 (SF-36) regardless of liver histology or ALT (serum alanine aminotransferase) levels. Quality-of-life impairment is detectable in persons who have never used drugs, but was found to be greatest in intravenous drug users, which indicates that HCV is not the sole reason for quality of life reduction in patients with chronic hepatitis C (Foster et al., 1998; European Monitoring Centre for Drugs and Drug Addiction, 2004; Gjeruldsen et al., 2006). Most frequent complaints associated with HCV infection comprise neuropsychiatric and gastrointestinal disorders, algesia and dysesthesia (Lang et al., 2006). Quality of life was shown to improve significantly with successful antiviral treatment (Spiegel et al., 2005). Today, it is commonly accepted that HCV-related quality of life impairment must be taken into account in order to assess the true burden of the disease. Most estimates of burden of disease reflect provisional expert consensus opinion. Motivated by the uncertainty of present burden of disease estimates and a lack of reliable data with which to prioritize public health measures, an international working group was established to assist the WHO in estimating the global burden of disease associated with HCV infection (The Global Burden of Disease Working Group, 2004). However, important results from this working group are still preliminary or pending. A summary measure for the burden of disease, designed to compare premature death and disability from various diseases across countries, was introduced by the WHO Global Burden of Disease Study (GBD). This measure is the DALY, the disability-adjusted life years. We focused on this measure for our present study. In the following, we present a review that we performed with the objective to (1) search for and summarize the data on HCV-related burden of disease in terms of DALY that are presently available for countries of the WHO European region and (2) where the DALY do not include the impact of the late stages of cirrhosis and liver cancer, to estimate this contribution through HCV-attributable fractions from DALY for all-cause cirrhosis and all-cause liver cancer.
1325
1326
76 2
DALYs in Chronic Hepatitis C : A Paneuropean Perspective
Review of Existing DALY Data and Calculation of the Contribution of Cirrhosis and Liver Cancer
The primary goal of our review was to retrieve nationally representative data that are comparable across countries. Therefore, we performed a systematic literature search in Medline, PreMedline and Embase, combining search terms for disease (‘‘hepatitis C’’ or ‘‘HCV’’), burden of disease related outcomes and geographic regions. We defined the disease as a documented HCV infection, determined by circulating viral RNA without restrictions to co-infections. Additional index search terms related to the disease included ‘‘cirrhosis’’, ‘‘hepatocellular carcinoma’’, and ‘‘liver cancer’’. As outcome of interest related to burden of disease, we searched for quality of life and focused on data for DALY. We searched for publications concerning Europe, or one of 22 countries in the WHO European region (Austria, Belgium, the Czech Republic, Denmark, Finland, France, Germany, Greece, Hungary, Ireland, Italy, the Netherlands, Norway, Poland, Portugal, Romania, Russia, Spain, Sweden, Switzerland, Turkey, and the United Kingdom). Since HCV was first discovered in 1989, we restricted our search to documents published since then. A final update of the literature search was performed in October 2006. Documents in languages other than English or German were excluded. In addition to the systematic literature search, we reviewed reference lists of retrieved publications, searched websites of national and international organizations, e.g., Centers for Disease Control and Prevention (US), Deutsche Stiftung Organtransplantation – German Foundation for Organ Transplantation, European Association for the Study of the Liver, European Centre for Disease Prevention and Control, European Liver Transplant Registry, European Commission, The Statistical Office of the European Communities, Eurosurveillance, National Institutes of Health (US), Organization for Economic Co-operation and Development, WHO. We also consulted with experts from health organizations and pharmaceutical companies to obtain data from national sources. For multi-causal disease outcomes such as liver cirrhosis or liver cancer, the data were reported only in aggregated form, while cause-specific data were missing. Therefore, our approach was to use attributable fractions (AF) to calculate the number and proportion of cases specifically related to HCV infection. Using this methodology, the number of cases attributable to HCV was derived by multiplying the total number of cases of the multi-causal disease outcome with the HCV-attributable fraction. HCV-attributable fractions used in our work were derived from a recent publication by Perz et al. (Perz et al., 2006), who estimated the attributable fractions of cirrhosis and hepatocellular carcinoma due to HBV and HCV infections for WHO sub-regions. We preferred attributable fractions as derived by Perz et al. to AFs recently derived by other researchers (Parkin, 2006) because the attributable fractions derived by Perz et al. do not rely on estimates of the HCV prevalence and corresponding relative risks of exposure in the source populations, which at present are major sources of uncertainty. The use of region-specific AFs is a simplification that may not have a strong influence on regional burden of disease estimates, but could yield inaccurate results for single countries. To describe quality-of-life-related burden of disease, we used DALY and years lost due to disability (YLD) (Murray, 1994; Murray and Lopez, 1994; Murray and Lopez, 1997; Mathers et al., 2003; Anonymous et al., 2006; Sassi, 2006). The DALY was introduced by the WHO Global Burden of Disease Study (GBD) to compare death and disability from various disorders across countries. It is a summary measure of population health that combines years of life lost
DALYs in Chronic Hepatitis C : A Paneuropean Perspective
76
from premature death (YLL) and years of ‘healthy’ life lost by being in a state of poor health or disability (YLD). One DALY can therefore be thought of as one lost year of ‘healthy’ life and the goal of a health intervention is to minimize DALYs. DALYs for a disease are calculated as the sum of YLL in the population and YLD for incident cases of the health condition. The incidence perspective for YLD is chosen to match the perspective of YLL, which measures the incident stream of life years lost due to deaths. Our systematic literature search yielded several studies investigating health related quality of life in patients with hepatitis C. However, none of the studies estimated the burden of disease resulting from HCV-related quality-of-life impairment for a whole country or region in Europe. Burden of disease data considering quality of life were only reported by the WHO Global Burden of Disease Project (GBD) (World Health Organization). Available data comprised year 2002 country-specific DALY (World Health Organization, 2004a) and region-specific YLD estimates (World Health Organization, 2004b). The WHO GBD reported DALYs and YLD only for hepatitis C without cirrhosis and liver cancer (World Health Organization, 2004a). Therefore, DALYs and YLD resulting from HCVrelated cirrhosis and liver cancer were calculated by weighing given data for all cause cirrhosis and liver cancer with HCV-attributable fractions. Retrieved and calculated DALY figures along with country-specific absolute numbers and rates, both overall and separately for HCV-related hepatitis, cirrhosis and liver cancer are presented in an evidence table (> Table 76-1). In addition, attributable fractions used to derive HCV-specific values, data sources, and calculation steps are listed. Based on our calculations, almost 1.2 Million DALYs were lost in the WHO European region due to HCV in 2002, which corresponds to an overall rate of 134.54 DALYs per 100 000 residents. The majority of DALYs (81%) were lost due to HCV-related cirrhosis. HCV accounted for 35% of the cirrhosis and 30% of the liver cancer DALYs in the WHO European region. These figures vary substantially across countries. Among countries with data for HCV, cirrhosis and liver cancer, the lowest value is 25 DALY per 100 000 (Iceland) and the highest value is 443 DALY per 100 000 (Moldova). In addition, data are presented as a choropleth map of the WHO European region compiled in the software SAS (release 9.1 by SAS Institute Inc., Cary, NC, USA), where the map categories represent rounded data quartiles (> Figure 76-1). The map shows the distribution of HCV-related DALY rates in countries of the WHO European region. Countries with low HCV-related DALY rates (155 DALYs per 100 000) include Croatia, Germany, Hungary, Kazakhstan, Kyrgyzstan, Moldova, Romania, Russia, Slovenia, Tajikistan, Turkmenistan, Ukraine and Uzbekistan. The countries with high rates are located in the East (east of Finland and Belarus and north of the Black Sea, and in a range of countries between Slovenia and Ukraine), whereas countries with low rates are disseminated throughout the region. Germany, with a value of 156 DALY per 100 000 (just above the threshold), is the only Western European country in the group of the highest DALY range. Because the DALY measure combines aspects of mortality and quality of life, we looked at YLD (the disability component of the DALY measure) separately. YLD data were available for the WHO European region, but not for individual countries. HCV-related YLD were calculated similarly to HCV-related DALY, using the regional HCV attributable fractions for cirrhosis (35%) and liver cancer (30%) derived from our DALY calculation. According to
1327
796
Cyprus
160
1816
5,197
59,850
5,177
82,414
10,970
Finland
France
Georgia
Germany
Greece
85
445
7971
7
20
5,351
1,338
Denmark
64
100
263
97
656
89
527
398
35
7
81
n
Estonia
10,246
4,439
Croatia
Czech Republic
4,126
7,965
Bosnia and Herzegovina
10,296
Belgium
Bulgaria
8,297
9,940
Austria
Azerbaijan
8,111
Armenia
Belarus
69
3,072
Andorra
3,141
Population (’000)
Albania
Countries of the WHO European Region
7413
264492
21203
140576
12061
5724
14892
28833
371
19749
20362
8016
25332
26913
23463
24341
6694
134
n
Total DALYS
51 74.0917
73.8103
96.1512
86.9196
66.0514
93.4966
92.0596
17.7280
89.2549
88.1801 7209 139.2458
53419
4583
1946 145.4489
5659 105.7492
10956 106.9362
141
7505 169.0782
6923
2725
9626
9151
7977
9249 114.0353
2276
n
per 100,000
38
2817
25.6814
38 100507 121.9532
34
38
38
34
38
38
38
38
34
34
38
34
34
38
34
38
34
%
AF
DALYS attr. to HCV
Cirrhosis
13269
45230
4055
53237
2578
601
1912
7631
423
3284
7398
5583
4638
3965
5551
5435
2656
50
4960
n
Total DALYS
44
44
15
44
44
15
44
44
44
44
15
15
44
15
15
44
15
44
15
%
AF
5838
19901
608
23424
1134
90
841
3358
186
1445
1110
837
2041
595
833
2391
398
22
744
n
53.2245
24.1475
11.7497
39.1387
21.8241
6.7375
15.7216
32.7708
23.3892
32.5563
13.9329
20.2958
19.8225
5.9839
10.0361
29.4813
12.9672
31.6535
23.6883
per 100,000
DALYS attr. to HCV
Liver cancer
All
26.2708
per 100,000
88.2120
98.9423 88.6864
41.1173
9100
82.9596
128379 155.7728
7824 151.1245
78659 131.4282
5878 113.0885
2056 153.6581
6586 123.0644
14378 140.3361
327
9049 203.8786
8296 104.1517
3659
12323 119.6868
9835
9337 112.5361
12039 148.4245
2710
80 115.8265
825
n
DALYS attr. to HCV
76
DALYS attr. to HCV
Hepatitis
. Table 76-1 DALY and DALY rates related to HCV in countries of the WHO European region in 2002
1328 DALYs in Chronic Hepatitis C : A Paneuropean Perspective
84
5,067
2,329
3,465
Kyrgyzstan
Latvia
162
6158
5,398
1,986
40,977
8,867
Slovakia
Slovenia
Spain
Sweden
417
6592
27
0
Russian Federation
220
22,387
144,082
Romania
243
27
4,270
Moldova
1656
1097
10,535
10,049
Portugal
Serbia and Montenegro
38,622
Poland
90
361
1
11
12
97
1203
San Marino
4,514
Norway
34
Monaco
16,067
393
Malta
Netherlands
447
Luxembourg
Lithuania
40
15,469
Kazakhstan
735
11195
6,304
124
57,482
3,911
Ireland
4
187
Israel
287
Italy
9,923
Hungary
Iceland
6237
64195
12094
22621
24606
24
677009
159426
53605
29519
93738
3671
12723
62
242
1122
12646
6419
31659
75553
110240
3142
2448
82
93358
34
72.8766
18.9415
23.7904
10.8223
93.7094
82.5203
30.9093
30.0908
68.5757
23.4585
95.3598
54205 242.1225
18226 426.7966
11217 111.6304
31871
1395
4835
23
92
426
4300 124.0892
2183
10764 212.4291
25688 166.0636
41891
1194
930
31
31742 319.8883
38
38
38
34
38
33.5571
2370
24394
26.7296
59.5307
4596 231.4278
7691 142.4789
9
34 230183 159.7589
34
34
38
34
38
38
38
38
38
34
34
34
34
38
38
38
38
3136
32520
1118
3207
9174
14
84009
20090
3025
5502
17908
754
4167
24
98
211
1175
1171
2956
9489
70278
1486
1223
83
7594
44
44
44
15
44
15
15
15
44
15
44
44
44
44
44
15
15
15
15
44
44
44
44
15
1380
14309
492
481
6
12601
3014
454
2421
2686
332
1833
10
43
93
176
176
443
1423
30922
654
538
36
1139
15.5628
34.9186
24.7736
8.9125
23.0311
8.7460
13.4608
10.6249
24.0930
6.9552
7.3497
11.4108
30.4847
11.0163
20.7319
5.0874
7.5413
8.7506
9.2012
53.7947
10.3701
13.7567
12.6654
11.4799 24.7890 40.9750
40.7147
37.2949 43.7469 92.3152
40.2499
2.0876
57.1033
4167
46.9904
45295 110.5363
5115 257.5791
8172 151.3914
220
16
248942 172.7788
57381 256.3088
18923 443.1102
15294 152.2001
35654
1817
7029
35 102.4308
146
531 118.6912
4560 131.6078
2398 102.9673
11305 223.0992
28314 183.0427
84009 146.1472
2583
1592
71
33068 333.2497
DALYs in Chronic Hepatitis C : A Paneuropean Perspective
76 1329
59,068
25,705
877,887
United Kingdom
Uzbekistan
ALL
63496
5600
2414
919
1186
5913
2733544
132821
114459
224442
28496
50473
2365
23172
10302
n
Total DALYS
54.5874
per 100,000
24.4046
39.3128
73.6345
45159 175.6793
43494
76310 156.0483
9689 202.0972
17161
804
7878 127.1734
3915
n
DALYS attr. to HCV
35 957389 109.0561
34
38
34
34
34
34
34
38
%
AF
Cirrhosis
532544
9089
19702
26395
2853
15060
1908
899
3767
n
Total DALYS
1363
8669
3959
428
2259
286
135
1657
n
18.2505
5.3038
14.6764
8.0963
8.9275
3.2126
13.9942
2.1771
23.1094
per 100,000
DALYS attr. to HCV
30 160219
15
44
15
15
15
15
15
44
%
AF
Liver cancer
80.2770
per 100,000
55.1223 36.0261
92.3978 1181105 134.5395
52122 202.7670
54577
81188 166.0233
11302 235.7542
25333
1128
11720 189.1861
5757
n
DALYS attr. to HCV
All
The Hepatitis C Virus (HCV) can cause acute hepatitis and chronic hepatitis with potential long-term consequences cirrhosis and liver cancer. Disability adjusted life years (DALY) are derived from WHO GBD (World Health Organization, Global Burden of Disease Project) data. DALY resulting from hepatitis (column 3) represent original WHO data. These data excluded cirrhosis and liver cancer. Liver cancer and cirrhosis DALY attributable (attr.) to HCV represent WHO GDB data weighted by population attributable fractions (AF). Population size data used to calculate DALY rates were derived from the WHO GDB project as well. Figures highlighted in italics represent external input data, figures in standard font indicate calculated results Data sources:(a) WHO GBD data 2002 available at: http://www3.who.int/whosis/burden/estimates/2002Rev/2002RevCountries/DthDALY2002.zip(b) Population attributable fractions: Perz JF, Armstrong GL, Farrington LA, Hutin YJ and Bell BP. The contributions of hepatitis B virus and hepatitis C virus infections to cirrhosis and primary liver cancer worldwide. J Hepatol 2006;45:529–38
4,794
48,902
Turkmenistan
Ukraine
70,318
Turkey
37
3707
6,195
2,046
Tajikistan
185
n
7,171
Macedonia
Switzerland
Population (’000)
DALYS attr. to HCV
Hepatitis
76
Countries of the WHO European Region
. Table 76-1 (continued)
1330 DALYs in Chronic Hepatitis C : A Paneuropean Perspective
DALYs in Chronic Hepatitis C : A Paneuropean Perspective
76
. Figure 76-1 HCV-related DALY rates in countries of the WHO European region in 2002. Source: Calculated from WHO GBD data (World Health Organization, 2004a)
the calculation, HCV caused 200 104 YLD in the WHO European region in 2002. Of those 6250 (3%) were due to hepatitis C without cirrhosis or liver cancer, 191 537 (96%) were due to HCV-related cirrhosis and 2317 (1%) were due to HCV-related liver cancer.
3
Revised Evaluation of the HCV Burden of Disease after Inclusion of Cirrhosis and Liver Cancer
The inclusion of the estimated HCV-related fraction of DALY from all-cause cirrhosis and liver cancer changes the picture dramatically. According to our calculations, HCV caused approximately 1.2 million DALYs in the WHO European region in 2002. Comparing this figure to DALY reported by the WHO GBD study for other diseases reveals that hepatitis C is a major health problem, almost the size of HIV/AIDS (ICD-10 Code B20-B24) or stomach cancer (ICD-10 Code C16), each with about 1.4 million DALY in 2002 (World Health Organization, 2004c). A closer look on YLD (the disability component of the DALY measure) reveals that about one sixth of the HCV-related DALY burden (200 104 YLD) are due to quality-of-life impairment. However, the YLD account for disability in incident cases only. Quality of life reductions for patients, who suffer from the disease but have been detected in the past, are not included. DALY differ in a wide range between countries. High DALY figures are predominantly found in Eastern countries. Since YLD contribute only one sixth to the DALY, they are unlikely to explain the differences between countries. Differences in the second DALY component, YLL
1331
1332
76
DALYs in Chronic Hepatitis C : A Paneuropean Perspective
(life years lost due to premature death) are therefore a more likely reason for DALY variation. The strong geographic variation of DALY may be partially explained by differences in prevalence, but also younger age at death is likely an important contribution. Deviating health care standards or differences in the distribution of competing and synergistic risk factors should be considered as explanations for geographic variations in age at death. For example, HCV-related death rates might be lower in countries with high prevalence of hepatitis B (World Health Organization, 1999b), whereas they might be higher in countries with a high alcohol consumption (Peters and Terrault, 2002). Specifically, heavy alcohol consumption was shown to be associated with higher risks of cirrhosis, liver cancer and death in patients with chronic hepatitis C (Shepard et al., 2005; Perz et al., 2006). Involvement of other risk factors or comorbidities can not be excluded. Heterogeneous data quality may add to geographical variation. About 95% of the HCV-related DALYs were accumulated by patients in advanced disease stages (cirrhosis or liver cancer). This high percentage emphasizes the importance and potential benefit of preventive antiviral treatment. However, it also suggests that quality of life impairment caused by HCV infection per se (Foster et al., 1998; Spiegel et al., 2005; Hollander et al., 2006; Lang et al., 2006; von Wagner et al., 2006) might not be relevant from a burden of disease perspective. However, this cannot be concluded with certainty, as it is unclear what conditions and aspects are covered by the disability weight of 0.075 (Mathers et al., 2003) that was used for the calculation of hepatitis C related DALYs in the GBD study. It should also be realized that estimating YLD and subsequently DALYs was the most complex component of the GBD study, requiring a broad set of input data from a multitude of sources (Mathers, 2005). Results for single countries might therefore be surrounded with considerable uncertainty. A limitation of our analysis is the use of regional instead of country-specific HCV attributable fractions. This may not have a strong influence on regional burden of disease estimates, but may yield inaccurate results for single countries. Since all parameters of our calculations are revealed, though, country-specific estimates can easily be revised by the interested reader when better data become available. The considerable geographic variation of DALY should be examined in further research. Reliable and comparable data for prevalence and mortality across countries need to be identified for that purpose. Competing risks and differences in health care for hepatitis C patients, for example access to state-of-the-art antiviral treatment or availability of liver transplantations, should be evaluated as well (Lettmeier et al., 2008).
4
Conclusion
Presently, the only data allowing a consistent cross-country comparison of burden of disease resulting from HCV-related premature death and quality of life impairment are DALYs and YLD estimated by the WHO GBD study in 2002. According to our additional calculations based on these data and HCV-attributable fractions of cirrhosis and liver cancer, hepatitis C is a major health problem in the WHO European region, comparable in magnitude to HIV/AIDS or stomach cancer. Our results underline the importance and potential benefit of preventive antiviral treatment. The burden of disease differs strongly across the countries of the WHO European region, with highest values in eastern countries. This variation indicates the potential of inequality of health services across counties that should be investigated further. The lack of data demonstrates that the full extent of the hepatitis C crisis is not yet fully perceived in Europe.
DALYs in Chronic Hepatitis C : A Paneuropean Perspective
76
Summary Points According to our calculations, HCV caused approximately 1.2 Million DALYs (life years
lost through premature death or impairment of quality of life) in the WHO European region in 2002. One sixth of these DALYs can be attributed to quality of life impairment. Hepatitis C is a major health problem, comparable to HIV/AIDS or stomach cancer each with about 1.4 million DALY. About 95% of the HCV-related DALY were accumulated by patients in advanced disease stages (cirrhosis of liver cancer), which underlines the importance and potential benefit of preventive antiviral treatment. The burden of disease expressed in DALY differs strongly across countries, with highest values in eastern countries. Countries with highest HCV-related DALY rates (>155 DALYS per 100 000) are Croatia, Germany, Hungary, Kazakhstan, Kyrgyzstan, Moldova, Romania, Russia, Slovenia, Tajikistan, Turkmenistan, Ukraine and Uzbekistan. Further research regarding the causes of the variation is needed. The lack of data demonstrates that the full extend of the hepatitis C crisis is not yet fully perceived in Europe.
Acknowledgments This project was supported in part by an unrestricted research grant from Hoffmann La-Roche Ltd., Basel, Switzerland. The authors had complete and independent control over study design, analysis and interpretation of data, report writing, and publication, regardless of results.
References Anonymous. (1997). Hepatology. 26: 2S–10S. European Association on the Studies on the Liver (EASL). (1999). J Hepatol. 31(Suppl 1): 3–8. European Monitoring Centre for Drugs and Drug Addiction (EMCDDA), Jager J, Limburg W, Kretzschmar M, Postma M, Wiessing L. Hepatitis C and injecting drug use: impact, costs and policy options. Monograph 7 [monograph on the Internet]. Lisbon: European Monitoring Centre for Drugs and Drug Addiction; 2004 [cited 2007 Sept]. Available from: http://www.emcdda.eu.int/?nnodeid=428. Foster GR, Goldin RD, Thomas HC. (1998). Hepatology. 27: 209–212. Gjeruldsen S, Loge JH, Myrvang B, Opjordsmoen S. (2006). Nord J Psychiatry. 60: 157–161. Hollander A, Foster GR, Weiland O. (2006). Scand J Gastroenterol. 41: 577–585. Lang CA, Conrad S, Garrett L, Battistutta D, Cooksley WG, Dunne MP, Macdonald GA. (2006). J Pain Symptom Manage. 31: 335–344. Lauer GM, Walker BD. (2001). N Engl J Med. 345: 41–52.
Lettmeier B, Mu¨hlberger N, Schwarzer R, Sroczynski G, Wright D, Zeuzem S, Siebert U. (2008). J Hepatol. 49: 528–536. Lopez AD, Mathers CD, Ezzati M, Jamison DT, Murray CJL (2006). Global Burden of Disease and Risk Factors. Oxford University Press, New York. Also available from: http://www.dcp2.org/pubs/GBD. Mathers CD, Bernard C, Moesgaard Iburg K, Inoue M, Ma Fat D, Shibuya K, Stein C, Tomijima N, Xu H. Global Burden of Disease in 2002: data sources, methods and results. Global Programme on Evidence for Health Policy Discussion Paper No. 54 [monograph on the Internet]. World Health Organization, Geneva, Switzerland 2003, revised 2004 [cited: 2006 Oct]. Available from: http://www.who. int/healthinfo/paper54.pdf. Mathers CD. Uncertainty and data availability for the global burden of disease estimates 2000–2002 [monograph on the Internet]. 2005 [cited 2007 Mar]. Available from: http://www.who.int/healthinfo/publications/boduncertaintypaper2002.pdf.
1333
1334
76
DALYs in Chronic Hepatitis C : A Paneuropean Perspective
Murray CJ. (1994). Bull World Health Organ. 72: 429–445. Murray CJ, Lopez AD. (1994). Bull World Health Organ 72: 481–494. Murray CJ, Lopez AD. (1997). Bull World Health Organ 75: 377–381. National Institute of Health (NIH). (2002). Hepatology. 36: S3–S20. Parkin DM. (2006). Int J Cancer. 118: 3030–3044. Perz JF, Armstrong GL, Farrington LA, Hutin YJ, Bell BP. (2006). J Hepatol. 45: 529–538. Peters MG, Terrault NA. (2002). Hepatology. 36: S220–S225. Sassi F. (2006). Health Policy Plan. 21: 402–408. Shepard CW, Finelli L, Alter MJ. (2005). Lancet Infect Dis. 5: 558–567. Siebert U, Sroczynski G, on behalf of the German Hepatitis C Model (GEHMO) Group and the HTA Expert Panel on Hepatitis C. (2003). [Antiviral therapy for patients with chronic hepatitis C in Germany. Evaluation of effectiveness and cost-effectiveness of initial combination therapy with Interferon/ Peginterferon plus Ribavirin. Series of the German Institute for Medical Documentation and Information commissioned by the Federal Ministry of Health and Social Security]. Vol. 8. DIMDI, Ko¨ln. Siebert U, Sroczynski G, on behalf of the German Hepatitis C Model (GEHMO) Group and the HTA Expert Panel on Hepatitis C. (2005). Int J Technol Assess Health Care. 21: 55–65. Spiegel BM, Younossi ZM, Hays RD, Revicki D, Robbins S, Kanwal F. (2005). Hepatology. 41: 790–800. The Global Burden of Disease Working Group. (2004). J Clin Pharmacol. 44: 20–29.
von Wagner M, Lee JH, Kronenberger B, Friedl R, Sarrazin C, Teuber G, Herrmann E, Zeuzem S. (2006). J Viral Hepat. 13: 828–834. World Health Organization. The Global Burden of Disease project: results for 2002 and earlier years, methods, documentation and publications. Manuals, resources and software for carrying out national burden of disease studies [website on the Internet]. [cited 2007 Mar]. Available from: http://www.who. int/healthinfo/bodproject/en/index.html. World Health Organization. Hepatitis C [monograph on the Internet]. 1999a [cited 2006 Oct]. Available from: http://www.who.int/immunization/topics/ hepatitis_c/en/index.html/. World Health Organization. (1999b). J Viral Hepat. 6: 35–47. World Health Organization. Mortality and burden of disease estimates for WHO member states in 2002 [database on the Internet]. 2004a [updated N.N.; cited 2006 Oct]. Available from: http://www3.who. int/whosis/burden/estimates/2002Rev/ 2002RevCountries/DthDALY2002.zip. World Health Organization. Estimates of Years Lost due to Disability by sex, cause and WHO Region for 2002 [monograph on the Internet]. 2004b [cited 2007 Mar]. Available from: http://www.who.int/entity/healthinfo/statistics/gbdwhoregionyld2002.xls. World Health Organization. Estimates of DALYs by sex, cause and WHO Region for 2002 [monograph on the Internet]. 2004c [cited Mar 2007]. Available from: http://www.who.int/healthinfo/statistics/ gbdwhoregiondaly2002.xls.
77 Economics and Vaccines J. Bos . M. Postma 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1336
2
Modeling Economic Evaluations of Vaccines: Methodological Issues . . . . . . . . . . 1338
3 3.1 3.1.1 3.1.2 3.1.3 3.2 3.2.1 3.2.2 3.2.3 3.2.4 3.3 3.3.1 3.3.2 3.3.3 3.3.4
Cost-Effectiveness of Vaccine Interventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1340 Childhood Cluster Diseases Vaccination Programs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1340 Measles, Mumps and Rubella (MMR) Vaccination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1341 Diphtheria, Tetanus, Pertussis and Polio (DTPP) Vaccination . . . . . . . . . . . . . . . . . . . 1343 Varicella Vaccination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1343 Vaccination Campaigns Against Respiratory Diseases . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1345 Hemophilus influenzae Type B . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1345 Pneumococcal Vaccines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1346 Meningococcal Vaccines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1346 Influenza Vaccines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1347 Vaccination Campaigns Against Sexually Transmitted Infections and Other Vaccine-Preventable Viruses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1347 Hepatitis B . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1347 Human Papilloma Virus (HPV) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1348 Hepatitis A . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1348 Rotavirus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1349
4
Discussion and Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1349 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1350
#
Springer Science+Business Media LLC 2010 (USA)
1336
77
Economics and Vaccines
Abstract: Infectious diseases are an important cause of mortality and morbidity, causing approximately 27% of the total disease burden in > DALY. A large part of DALY lost due to infectious diseases could be prevented by improving existing vaccination programs for the population. Diseases such as > childhood cluster diseases, hepatitis A and B, respiratory infections caused by influenza, pneumococcal and meningococcal infections, and Hemophilus influenzae type B, are for a large part preventable by current vaccines. The implementation of new vaccine programs or/and strategies is often a costly process with long term consequences. Vaccination programs often concern a large part of the population, and have a large budget impact. Once a vaccination program has started, it is (due to equity reasons) extremely difficult to cease the program. To gain a better understanding of the potential impact on health benefits and costs of a vaccine intervention, health-economic evaluations are frequently used, which estimate the future impact on health gains and costs. Health-economic evaluations are mostly presented as one of the following four types of analysis: cost-minimization, cost-benefit, cost-effectiveness and cost-utility analysis. In this chapter, we provide an overview of the main techniques and challenges associated with health economic evaluations of vaccination programs, such as the choice of the model. Additionally, an overview of health economic evaluations that have been performed on currently implemented vaccination strategies is presented. From our analysis it follows that vaccine programs, and especially those against childhood cluster diseases, and vaccination of elderly against influenza are amongst the world’s most cost-effective interventions. List of Abbreviations: BCG, bacillus Calmette-Gue´rin; CBA, > cost benefit analysis; > CEA, cost effectiveness analysis; > CMA, cost minimization analysis; DALY, disability adjusted life year; > DTPP, diphtheria, tetanus, pertussis and poliomyelitis; > Hib, Hemophilus influenzae type b; > HPV, human papilloma virus; > IPV, inactivated poliomyelitis vaccine; > MMR, measles, mumps and rubella; > OPV, oral poliomyelitis vaccine; > QALY, quality adjusted life year; > WTP, willingness to pay; > VZV, varicella zoster virus
1
Introduction
Despite the discovery of vaccines that helped to eradicate smallpox, the launch of a global campaign to eradicate poliomyelitis, and support the control of measles, diphtheria and other diseases, infectious diseases remain the leading cause of death in developing countries (WHO, 2007). Worldwide, infectious diseases are an important cause of mortality and morbidity, causing approximately 27% of the total disease burden in DALY (Authors calculation based on WHO, 2004 data). A large part of DALY lost due to infectious diseases could be prevented by improving existing vaccination programs for the population. Diseases such as childhood cluster diseases, hepatitis B, TB, respiratory infections caused by influenza, pneumococcal and meningococcal infections, Hemophilus influenzae type B, are for a large part preventable by current vaccines but still lead to a large burden of disease and mortality. > Table 77-1 provides an overview of the global burden of disease caused by infectious diseases. The implementation of new vaccine programs or/and strategies is an often costly process with long term consequences, due to the fact that (1) vaccination strategies often involve vaccination of large parts of the population and (2) once a vaccination program has started, it is (due to equity reasons) extremely difficult to cease the program. Additionally, in most cases,
Economics and Vaccines
77
. Table 77-1 Overview of major infectious diseases and DALY burden worldwidea All Causes Communicable conditions Infectious and parasitic diseases
1.490.125.643 403.490.600 308.887.251
Tuberculosis
34.735.908
STDs excluding HIV
11.347.067
HIV/AIDS
84.457.784
Diarrheal diseases
61.966.183
Childhood-cluster diseases (Pertussis, Poliomyelitis, Diphtheria, Measles, Tetanus)
41.479.543
Meningitisa
6.191.790
Hepatitis B
2.170.326
Hepatitis C
1.003.682
Malaria
46.485.868
Tropical-cluster diseases: Trypanosomiasis, Chagas disease, Schitomiasis, Leishmaniasis, lymphatic filariasis, onchocerciasis
12.245.452
Leprosy
198.778
Dengue
615.529
Japanese encephalitis
709.219
Trachoma
2.328.780
Intestinal nematode infections
2.951.341
Respiratory infections Lower respiratory tract infections, Upper respiratory tract infections, Otitis media
94.603.349
Noncommunicable diseases
697.815.295
Injuries
181.991.119
Adapted from (Gwatkin and Guillot, 1998) a Total does not add up since we omitted DALY burden due to perinatal conditions and nutritional deficiencies
decisions with regards to vaccination programs are public health policy decisions, making rational decision making a must. A better understanding of the potential impact of an intervention will enable policy makers to prioritize intervention in public healthcare, so that the available funds are spend in such a way that health care gains are being maximized. As a result, in an increasing number of countries, for new interventions to successfully apply for reimbursement, not only their clinical effectiveness has to be proven, but also their costeffectiveness (Cookson and Mc Daid, 2003). For the estimation of potential costs and benefits of the introduction of a new public health program, health-economic evaluations are frequently used, which estimate the future impact on health gains and costs. Health-economic evaluations are mostly presented as one of four types of analysis: cost-minimization analysis (CMA), cost-benefit analysis (CBA), costeffectiveness analysis (CEA) and cost-utility analysis (CUA), of which the latter two are used most often. In cost-minimization analysis, the interventions under study are assumed to have
1337
1338
77
Economics and Vaccines
similar efficacy and equal effectiveness. Only relevant costs are compared, and the cheapest intervention is assumed to be the most efficient. In cost-benefit analysis, the benefits of the intervention are expressed in monetary values, usually as the individual’s willingness to pay (WTP) for a certain risk reduction in morbidity or mortality due to a new medical intervention (Johannesson and Meltzer, 1998). Concern over the monetary valuation of the resulting human lives, led to the development of cost-effectiveness analysis as an alternative. In costeffectiveness analysis, benefits of the intervention are expressed in a natural outcome measure, such as averted infections or life years gained. In cost-effectiveness analysis, differences in quality of life are not valued. Another form of cost-effectiveness analysis is cost-utility analysis, where health gains are corrected for the quality of life of the patient. Health benefits are then mostly expressed as quality adjusted life years (QALYs) or other denominators, such as disability adjusted life years (DALYs) (Johannesson and Meltzer, 1998). In a growing number of countries, such as the Netherlands, United Kingdom, Australia, and Canada, these evaluations are being used to guide policy making on the introduction on new public vaccination programs (Cookson and Mc Daid, 2003). For instance, the decision of the Netherlands to vaccinate all infants with meningococcal C vaccine on the age of 14 months was a direct effect of the health-economic analysis (Welte et al., 2005).
2
Modeling Economic Evaluations of Vaccines: Methodological Issues
In economic evaluations of healthcare interventions, net costs are related to health gains, and expressed as a ratio such as the cost-effectiveness ratio of net costs per life-year gained. Net costs are estimated by subtracting the savings on averted infectious diseases’ treatment costs from the costs of the vaccines and the investment costs in administration and infrastructure. Both costs and health gains are corrected for the time of occurrence using an annual > discount rate. Vaccine interventions usually have the following features, which might set them apart from other interventions in public health: 1. Preventative character: Although therapeutic vaccines are being developed against certain forms of cancer for instance, most vaccines are given as prophylaxis. Some infectious diseases, such as hepatitis B and HPV infections, have an impact on the occurrence of severe complications much later on in life. This might pose a challenge to the analyst, since often long time frames of analysis need to be modeled. 2. Large scale, hence often large budgetary impacts: Whereas economic evaluations of other pharmaceuticals often target individual patients (or small patient populations), vaccine interventions in most cases concern a larger part of the population, causing the intervention to have a large budgetary impact. Therefore, when assessing these interventions, budget impact and allocation of the total budget will form an important aspect of the final decision. 3. Targeted against infectious diseases: Although some exceptions exist, (for instance vaccines are being developed against cancer) most vaccines target specific infectious diseases. Due to the communicable character of the agents causing disease, all indirect effects of removal of the pathogen from the population on those not vaccinated should also be included in the model. This is a challenge for the analyst, since the transmission mechanism of the causing agent needs to be assessed and included in the analysis. Additionally, an
Economics and Vaccines
77
intervention against an infectious disease will gradually be subject to diminishing marginal returns. This occurs since vaccinating the population leads to a diminished > force of infection, causing the incidence of disease to drop. So, as the coverage of the vaccination program increases, more and more persons need to be vaccinated in order to prevent a single case of disease. This may well cause a paradox, in which eradication of disease is even more costly than disease control by selected vaccination strategies. 4. Incongruent timing between costs associated with the intervention and the resulting health benefits: Some diseases, such as HPV infections or hepatitis B infections have an impact on the occurrence of severe complications much later in life. Thus, a large time gap might exist between the costs of the intervention and the expected benefits of the intervention. Using equal discount rates for costs and health effects might have a large impact on the costeffectiveness ratio of the vaccine. In a recent consensus statement on vaccination programs for preventing Hepatitis B, Beutels et al. (2002) re-iterate that discounting health effects significantly and negatively affects estimated cost-effectiveness of vaccination programs with long term effects. In economic evaluations, it is good practice to use data from clinical trials. However, it has been argued that for the evaluation of vaccine programs, effectiveness data from clinical trials are not transferable to a real life setting (Clemens et al., 1996; Edmunds et al., 1999). Since vaccine trials are small in relation to population based vaccination programs, they can only provide an estimate of the individual efficacy of the vaccine and do not give a good estimate for the overall effectiveness of a mass vaccination campaign in the population. By vaccinating a large group in the population, the circulation of the bacteria or virus will diminish, leading to a reduction of disease beyond the direct effects of vaccination. As a consequence, the force1 of infection diminishes. So, by mass vaccination, a certain level of protection is also offered to those in the population who are not vaccinated (Edmunds et al., 1999). This phenomenon is called > ‘‘herd immunity’’ – not correcting for it may cause an underestimation of the effect of the intervention (Edmunds et al., 1999). Another consequence of a diminished force of infection is that infections will occur at a later age. This shift in agespecific incidence can have great implications for public health, especially for diseases that have worse outcome when occurring at a later age (for instance varicella zoster infections) (Beutels et al., 2002). Cost-effectiveness analyses of vaccination programs that assess the transmission of disease, use either a static or a dynamic transmission model. The difference between a static or a dynamic model lies in the assumption of the value of the force of infection. A static model assumes that the force of infection is a constant, while in a transmission dynamic model the force of infection is a function of the number of infectious individuals in the population (Edmunds et al., 1999). In the latter, herd immunity effects of mass vaccination on the force infection are taken into account. Also, in transmission dynamic models the effects of vaccination on for instance the age-specific incidence rate are assessed. In > Table 77-2, an overview is presented of the most important model types and characteristics.
1
The force of infection is defined as the per-susceptible rate of infection. The force of infection can be calculated approximately by dividing the incidence by the number of susceptibles
1339
1340
77
Economics and Vaccines
. Table 77-2 Key characteristics of the models (Bos et al., 2008) Key assumption Static
Constant force of infection
Model type Decision analysis/ Markov model
Dynamic Force of infection is a function (Age-structured) of the fraction infected compartmental model
Main features Simple Minimal data requirements Complex Needs detailed data on transmission dynamics Able to evaluate impact of herd immunity Accounts for shift in agespecific incidence rates Gives insight in shifts in costeffectiveness ratio over time
3
Cost-Effectiveness of Vaccine Interventions
In this section, we will provide an overview of most common vaccination programs in developed countries, their cost-effectiveness and the cost-effectiveness models used for the analysis. A large number of vaccines have been in use in public health interventions for quite some time, as is illustrated in > Table 77-3 (which is by no means exhaustive). The cost-effectiveness of an intervention is usually measured compared to the do-nothing alternative. A number of these vaccines have been in use for a long period of time, making the cost-effectiveness of these vaccines is difficult to measure. The effects of the vaccine on the epidemiology and transmission of disease depend highly on the fact that a vaccine has been given for a very long period in time before starting the analysis. Therefore, it is impossible to calculate the cost-effectiveness against the do-nothing scenario. Additionally, the decision to vaccinate against these diseases has been made a long time ago, in a time where health economic evaluations of public health interventions were not performed. However, costeffectiveness studies on adaptations of those vaccines, for instance different formulations with less side-effects and/or higher efficacy, are frequently used by policy makers.
3.1
Childhood Cluster Diseases Vaccination Programs
This category consists of vaccination programs against common childhood diseases, such as measles, mumps, diphtheria, tetanus, poliomyelitis, rubella, and pertussis. Cost-effectiveness analysis of the original interventions with Measles, Mumps, Rubella vaccine (MMR) and polio and Diphtheria, Tetanus and Pertussis (DTP) vaccines have only been performed to a limited extend. These childhood vaccines are commonly cited to be among the most cost-effective interventions, and have resulted in the saving of millions of lives of infants and young children (WHO, 2008). A recent study estimated that a one-week ‘‘supplemental immunization activity’’ against measles carried out in Kenya in 2002 – in which 12.8 million children were vaccinated – would result in a net savings of US$12 million over the following ten years; during
Economics and Vaccines
77
. Table 77-3 Introduction of first generation vaccines Vaccine
Year of introduction first generation vaccine for human use
Smallpox
1798
Rabies
1885
Plague
1897
Diphtheria
1923
Pertussis
1926
Tuberculosis (BCG)
1927
Tetanus
1927
Yellow fever
1935
Injectable Polio vaccine (IPV)
1955
Oral Polio vaccine (OPV)
1962
Measles
1964
Mumps
1967
Rubella
1970
Hepatitis B
1981
Influenza
1953
Hepatitis Aa
1995
Hemophilus influenza type B
1992
Pneumococcal conjugate vaccine
2000
Pneumococcal vaccine polysaccharideb
1948
Meningococcal C vaccine
1999
Meningococcal A, Y, W135 and C vaccine
2005
Adapted from Plotkin and Mortimer (1994) a and b Baker (2007) and Austrian (1981)
which it would prevent 3,850,000 cases of measles and 125,000 deaths. In the United States, cost-benefit analysis indicated that every dollar invested in a childhood disease vaccine dose saves US$2 to US$27 in health expenses (WHO, 2008).
3.1.1
Measles, Mumps and Rubella (MMR) Vaccination
Since the vaccines against measles, mumps and rubella are usually given in the MMR combination vaccine, we will discuss both health-economic analyses of single vaccine interventions, such as revaccination after an outbreak of measles, as well as the cost-effectiveness of MMR vaccine. Few studies have been undertaken to assess the cost-effectiveness of the MMR vaccine since it’s introduction in general vaccination programs in the second half of last century. Most studies focus on adaptations of existing formulations, or changes in target population. A single study from 1985 was found that compared vaccination with MMR
1341
1342
77
Economics and Vaccines
vaccine to the do-nothing scenario in the US, and found the intervention to be cost-saving (White et al., 1985). Without an immunization program, an estimated 3,325,000 cases of measles would occur as compared to 2,872 actual cases in 1983 with a program. Instead of an expected 1.5 million rubella cases annually, there were only 3,816 actual cases. Mumps cases were lowered from an expected 2.1 million to 32,850 actual cases. Without a vaccination program, disease costs would have been almost $1.4 billion. Expenditures for immunization, including vaccine administration costs and the costs associated with vaccine reactions, totaled $96 million. The resulting benefit-cost ratio for the MMR immunization program was approximately 14:1. The savings realized due to the use of the combination vaccine rather than single antigen vaccines totaled nearly $60 million (White et al., 1985). 3.1.1.1
Measles Vaccine
Despite the lack of formal cost-effectiveness studies, measles vaccination is cited to be among the most cost-effective public health interventions implemented world-wide. (WHO, 2008) A few studies have been conducted on the cost-effectiveness of additional vaccination in the case of a measles outbreak, or on measles eradication strategies. Stover et al. (1994) assessed the cost-effectiveness of a program to identify and immunize susceptible hospital employees during a measles outbreak. In three US hospitals, they compared blind MMR vaccination with targeted MMR vaccination to only those at risk (being those born from 1957 onwards). Their analysis showed that a directed MMR immunization program was projected to be costeffective compared to universal MMR vaccination of all hospital employees. However, no formal transmission model was used, making it difficult to assess the effects of vaccinating hospital workers on the transmission of measles, mumps or rubella to the patient population. Sellick et al. (1992), also showed targeted measles immunization for susceptible workers to be more cost-effective than universal vaccination. A few studies were found that assessed the cost-effectiveness of improved measles control or measles eradication within the general population. Pelletier et al. (1998) studied the benefitcost ratio of two-dose measles vaccination of infants in Canada, using a transmission dynamic model. Zwanziger et al. (2001) used a static approach to evaluate the economic impact of increasing measles immunization rates in the United States. They examined the relationship between measles incidence and the immunization rate and converted it to a linear model, which was linked to a decision analysis model to assess the cost-effectiveness. The decision analysis study by Shiell et al. (1998) used similar methodology to assess the cost-effectiveness of measles vaccination to prevent school-based outbreaks in Australia, finding favorable costeffectiveness. A study by Gay et al., used a transmission dynamic model to assess the impact of adding a second booster dose of measles vaccine for infants aged 18 months or 5 years to the national vaccination program in Canada (Gay et al., 1998). Their model analyzed the transmission between five different age groups over time. The results of this modeling study showed that a combination of a catch-up campaign and a booster vaccine for infants would have an immediate impact in reducing the transmission of measles, whereas adding only a routine second dose would still allow endemic transmission between older infants for at least 10–15 years. Beutels and Gay (2003) analyzed the costs and benefits of the eradication of measles, using a transmission dynamic model that simulated ten different measles vaccination strategies for a hypothetical west-European country. The conclusion of this analysis was that very high (>95%) coverage two-dose vaccination is optimal, irrespective of past vaccination coverage. Additionally, the addition of a catch-up campaign to this two-dose vaccination strategy would
Economics and Vaccines
77
be cost-saving in the case of low historical coverage ( BCG vaccine against TB is still being used for TB proxylaxis. However, since there is still considerable discussion on the efficacy and duration of protection of BCG vaccine in the prevention of TB, we decided not to discuss vaccines against TB in this section.
3.2.1
Hemophilus influenzae Type B
Vaccination against Hemophilus influenzae type b (Hib) has been introduced in most Western countries in the early nineties. The choice to introduce the vaccine in the routine childhood vaccination programs was justified by it’s high incidence of between 20 and 69 per 100,000 infants 12 months only need a single dose of vaccine in order to build immunity against the pathogen, and younger infants need three doses to boost their immune system, vaccination of the first group will be more affordable. The cost-effectiveness of a single dose intervention ranged from €19,000 per QALY in Switzerland to €2,600 per QALY in the Netherlands, with the main differences found in the incidence of disease. The three dose scenario was analyzed to be less cost-effective, with cost-effectiveness ratios ranging from €55,000 per QALY in Switzerland, to €21,700 per QALY in the Netherlands.
Economics and Vaccines
3.2.4
77
Influenza Vaccines
Influenza has been frequently referred to as ‘‘the last great uncontrolled plague of mankind,’’ Elderly persons are at increased risk of developing influenza-related complications which require complex healthcare and might lead to mortality. By 1997, all but three countries in the European Union had universal vaccination programs for citizens aged ≥65 years. In a review by Postma et al. (2002), an overview is given of health economic studies on influenza vaccination of the elderly in developed countries. The studies provided remarkably similar results. Benefit to cost ratios ranging from 0.7 to 50 were found, indicating that in most instances the benefits of influenza vaccination of the elderly outweigh the costs associated with the program. Even in the study were a benefit to cost ratio of 0.7 was found, the costeffectiveness ratio was still below the acceptable threshold for cost-effectiveness. In particular, influenza vaccination among elderly people at higher risk, such as the chronically ill elderly, is generally found to be cost saving. A number of studies have been performed on the cost-effectiveness of influenza vaccination of healthy working adults. The results of these studies are not homogeneous; some studies report cost savings, whereas other studies report no economic benefits at all. Most differences in outcome are related to whether indirect costs due to work loss are included in the analysis, the efficacy of the vaccine in relation to the dominant influenza strains in the year of analysis, and the severity of the influenza epidemic (Postma et al., 2002). A review of 11 studies reported in western countries found that eight of the studies reported cost-savings against three studies that reported no economic benefit (Postma et al., 2002). Another review study, by Wood et al. (2000) also noted the disparity in results between the various economic studies, but also concluded that the published studies seem to suggest that influenza vaccination in the healthy, working adult would be a cost-effective health intervention, at least from the perspective of an employer. Some studies have analyzed whether vaccination of healthcare workers to protect high-risk patients would be a cost-effective strategy. A review by Burls et al. (2006) found that most costeffectiveness studies did not take the effects on patient transmission into account and therefore underestimated the cost-effectiveness of the intervention. The study by Burls et al. estimated that vaccination of healthcare workers to protect high-risk patients would be a highly costeffective intervention, with a CER between cost-saving and 405 GBP per LYG.
3.3
Vaccination Campaigns Against Sexually Transmitted Infections and Other Vaccine-Preventable Viruses
3.3.1
Hepatitis B
Hepatitis B virus (HBV) infection is still an important public health problem, despite the availability of an effective vaccine for several decades now. According to the WHO recommendation for universal routine vaccination many countries of high and intermediate endemicity have implemented routine vaccination. Implementation of such programs is often done at the infant age or young adolescent age, with a combined catch-up program. Favorable cost-effectiveness of implementing these programs has been evidenced broadly (Beutels, 2001), (Beutels et al., 2002). For example, the specific Italian implementation of vaccination
1347
1348
77
Economics and Vaccines
in 1992 of a combined universal infant schedule and catch-up of 12-years olds was chosen after careful examination of economic data (Bonanni et al., 2003). Discussions on introducing universal vaccination for infants and/or adolescents are currently going on in countries that currently only vaccinate selected target populations (such as the UK and the Netherlands) and in developing countries for which vaccination becomes feasible due to recent price reductions.
3.3.2
Human Papilloma Virus (HPV)
Recently, two Human papilloma virus vaccines have been registered for prophylactic use. Trials have proved efficacy up to 5 years; long-term effectiveness has yet to be demonstrated in follow-up of the initial trails and post-marketing research. Both vaccines have proven to be highly effective against HPV-types 16 and 18, that are associated with 70% of cervical cancers worldwide (Paavonen et al., 2007; Parkin and Bray, 2006). Slight differences seem to exist between both vaccines, potentially influencing exact cost-effectiveness profiles of both. In particular, one is quadrivalent, also protecting against genital warts caused by types 6 and 11 (Gardasil), whereas the other is bivalent only, but claiming higher likelihood for long-term protection (Frazer, 2007). Various cost-effectiveness analyses have already been performed, generally building on cost-effectiveness analyses for cervical cancer screening (Dasbach et al., 2006). Notably, vaccination is assumed to be implemented on top of the existing screening programs. All cost-effectiveness models for the HPV-vaccines generally assume no deterioration in the coverage of the screening. Additional vaccination of young teenage boys might be considered, but is less likely to be cost-effective (Dasbach et al., 2006). Approximately 20 publications are now available on the cost-effectiveness of HPVvaccination. For effectiveness, most of these models build on the observations in the trials for HPV-infections of the vaccine-types, rather than ‘‘hard endpoints’’ such as pre-cancerous stages and cervical cancer. Generally, acceptable cost-effectiveness of around US50,000 per QALY or lower are estimated for vaccinating young teenage girls. One concern generally expressed from the cost-effectiveness analyses is that a slight deterioration in the coverage of screening could easily offset the savings and health benefits of vaccination. So, implementation of vaccination should be combined with campaign to sustain screening coverage rates, and such costs should be included in the cost-effectiveness analyses of vaccination.
3.3.3
Hepatitis A
Hepatitis A vaccines have been available for over a decade. With the disease impact primarily affecting developing rather than developed countries, the picture of universal vaccination differs strongly from Hepatitis B. In particular, many developed countries have strategies implemented on vaccinating risk groups, such as ethnic minorities and travelers to Hepatitis A endemic regions. In a recent review, none of the cost-effectiveness studies yet being performed for the Hepatitis A vaccine were done for a developing country (Anonychuk et al., 2008). Costeffectiveness studies have up to now primarily evaluated targeted vaccination, such as vaccinating antibody-negatives only, prison inmates and the military (Jefferson et al., 1994). Additionally, some studies have been done specifically addressing the cost-effectiveness of vaccinating chronic Hepatitis C patients with Hepatitis A vaccine (Jacobs et al., 2002).
Economics and Vaccines
77
Findings indicate that favorable cost-effectiveness is highly unlikely the older the patient is and if antibody screening is performed prior to vaccination. More cost-effectiveness studies are needed and standardization of the approach should be enhanced.
3.3.4
Rotavirus
After initial drawbacks on the use of rotavirus vaccines due to suspected intussusception, recently two vaccines are being marketed worldwide (Rotarix and RotaTeq). Safety and efficacy have been shown in various clinical trials for the respective 2- and 3-dose schedules (Vesikari et al., 2006). Rotavirus infection is considered to be the most prevalent cause of acute gastroenteritis globally. Currently, the vaccine is initially considered for the developed world, where mortality and serious morbidity is relatively low. A small number of cost-effectiveness studies have been done on rotavirus vaccination, however these publications either lack comparability by using QALYs or do not include the full scale of savings and health gains (Widdowson et al., 2007). In particular, due to lack of valid data, rotavirus episodes without hospitalization are not yet factored into the models, leading to underestimation due to the omission of potential work loss of parents. In the absence of more robust information on these aspects, models are yet very sensitive to mortality assumed for rotavirus. Typically, mortality is assumed at approximately 1 per 100,000 infants annually, however valid registrations of rotavirus deaths may be lacking in many countries. Generally, from the cost-effectiveness analyses it appears that rotavirus vaccination may not be highly cost-effective, with cost-effectiveness ratios varying from €50,000 to 100,000 per QALY from the health-care payer perspective (Jit and Edmunds, 2007). Only marginally better ratios were found from the societal perspective for England and Wales (Jit and Edmunds, 2007). If however from the societal perspective parental work and QALY losses are included related to those cases of disease without health-care services visits involved, cost-effectiveness may improve dramatically. For the Belgium situation cost-saving potentials were reported (Bilcke et al., 2007). Cost-effectiveness ratios are highly sensitive to the QALY impact assumed. Up to now only one study is available on the specific quality-of-life aspects of rotavirus infection (Se´ne´cal et al., 2006). In particular, a prospective study was designed in Canada to estimate QALY-impacts for the infants involved and the caregivers using the Health Utilities Index and the Visual Analogue Scale. However, more research is needed to further underpin this crucial parameter in rotavirus cost-effectiveness analysis. Also, we note that further discussions on including QALY impacts for caregivers are required and that current standards for cost-effectiveness analysis do not include QALY loss with caregivers.
4
Discussion and Conclusion
In this chapter, we have provided a brief overview of the main challenges that are associated with modeling the economic impact of vaccines. Often, the vaccines are given as prophylaxis, and concern a very large part of the population. The latter implies that the budget impact is a prominent issue to consider prior to the introduction of new vaccination programs. Moreover, it is difficult to stop a program once it has been started, due to reasons of equity. Therefore, with assessing these interventions, budget impact and allocation of the total budget forms an important aspect of the final decision.
1349
1350
77
Economics and Vaccines
Most vaccines yield benefits not only for the direct beneficiary, but also for the nonvaccinated population. This phenomenon, called herd immunity, is an important aspect of vaccines. Lastly, due to the preventative character of vaccine interventions, benefits might occur later in life, which should be valued according to the preferences of the population. Vaccine interventions are a public health decision, affecting many citizens directly, and have long term budgetary consequences. Therefore, decision regarding new vaccination programs should consider all aspects of the intervention carefully. To assess the economic and health impact of the intervention, health economic studies such as cost-effectiveness analysis are being used. By using cost-effectiveness analysis, decision makers gain an understanding of those interventions that provide best value for money and that should be prioritized in order to maximize the utility gains in the population. In this chapter, we have provided an overview of health economic evaluations of most vaccines that have been introduced in the developed world. Most of the vaccines that have been introduced all exhibit favorable cost-effectiveness ratios, ranging from cost saving or highly costeffective for childhood cluster vaccines, Hemophilus influenzae type b vaccine, influenza vaccination for the elderly, meningococcal C vaccine and hepatitis A and B vaccine for infants. For some vaccination programs, the cost-effectiveness is very sensitive to changes in certain assumptions, such as the relation with zoster for the varicella vaccine, the occurrence of herd immunity with the pneumococcal conjugate vaccine for infants, and whether or nor screening rates for cervical cancer will decline as a result of HPV vaccination. Each of these assumptions has the potential to change an intervention from cost-effective towards not costeffective or vice versa. Although the cost-effectiveness studies are extremely sensitive to changes in these assumptions they serve a clear purpose in showing decision makers the impact of such changes. It is worth noting that most of the vaccine interventions prior to the 1990s have been implemented with only limited guidance of health economic studies. However, there are signs that economic evaluations are increasingly being used for guidance on whether or not to implement a new vaccination program. Welte et al. (2005) investigated whether or not health economic data played a role in the decision to start national vaccination programs against meningococcal C infections, and found a clear relationship between increasing incidence rates for meningococcal C infections and the commissioning of an health economic study on vaccination. This indicates that health economic studies are starting to play a more important role in decision making on the introduction of new vaccination programs. However, it needs to be kept in mind that, health economic data are only a single consideration next to for instance equity, or ethical reasons.
Summary Points Infectious diseases are an important cause of mortality and morbidity, causing approximately 27% of the total disease burden in DALY. A large part of DALY lost due to infectious diseases could be prevented by improving existing vaccination programs for the population. Diseases such as childhood cluster diseases, hepatitis A and B, respiratory infections caused by influenza, pneumococcal and meningococcal infections, and Hemophilus influenzae type B, are for a large part preventable by current vaccines. The implementation of new vaccine programs or/and strategies is a costly process with long term consequences. To gain a better understanding of the potential impact on health
Economics and Vaccines
77
benefits and costs of a vaccine intervention, health-economic evaluations are frequently used, which estimate the future impact on health gains and costs. In economic evaluations of vaccination campaigns, models are frequently used to assess the population effects of the vaccination campaign. Vaccine trials are small in relation to population based vaccination programs, they can only provide an estimate of the individual efficacy of the vaccine and do not give a good estimate for the overall effectiveness of a mass vaccination campaign in the population. By vaccinating a large group in the population, the circulation of the bacteria or virus will diminish, leading to a reduction of disease beyond the direct effects of vaccination. Vaccination campaigns, especially those against childhood cluster diseases, and influenza vaccination in the elderly are amongst the most cost-effective interventions. Vaccine interventions prior to the 1990s have been implemented with only limited guidance of health economic studies. However, health economic studies are starting to play a more important role in decision making on the introduction of new vaccination programs.
References Akumu A, English M, et al.(2007). Bull World Health Organ. 85(7): 511–518. Anderson RM, May RM. (1991). Infectious Diseases of Humans; Dynamics and Control. Oxford University Press, New York. Anonychuk A, Tricco A, et al.(2008). Pharmacoeconomics. 26: 17–32. Austrian R. (1981). Proc Am Philos Soc. 125(1): 46–51. Baker CJ. (2007). N Engl J Med. 357: 1757–1759. Banz K, Wagenpfeil S, et al.(2003). Vaccine. 21: 1256–1267. Bart K, Foulds J, et al.(1996). Bull World Health Organ. 74: 35–45. Beutels P. (2001). Health Econ. 10: 751–774. Beutels P, Bonanni P, et al.(1999). Vaccine. 17(19): 2400–2409. Beutels P, Edmunds WJ, et al.(2002). Pharmacoeconomics. 20: 1–7. Beutels P, Gay NJ. (2003). Epidemiol Infect. 130(2): 273–283. Beutels P, Thiry N, van Damme P. (2007). Vaccine. 25(8): 1355–1367. Bilcke J, Beutels P, et al. (2007). Cost-Effectiveness of Rotavirus Vaccination of Infants in Belgium. KCE, Brussels. Bonanni P, Pesavento G, et al.(2003). Vaccine. 21: 685–691. Bos JM, van Alphen L, et al.(2008). ‘‘The use of modelling in the economic evaluation of vaccines. Expert Rev Pharmacoeconomics Outcomes Res. 2(5): 443–455.
Brisson M, Edmunds WJ. (2002). Vaccine. 20: 1113–1125. Brisson M, Edmunds WJ, et al.(2000a). Vaccine. 18: 2775–2778. Brisson M, Edmunds WJ, et al.(2000b). Epidemiol Infect. 125: 651–669. Broughton E. (2007). J Public Health (Oxf). 29(4): 441–448. Burls A, Jordan R, et al.(2006). Vaccine. 24(19): 4212–4221. Caro J, Getsios D, et al.(2005). Pediatr Infect Dis J. 24(Suppl): S75–S82. Caro JJ, Getsios D, et al.(2005). Pediatr Infect Dis J. 24(5 Suppl): S48–S54. Clemens J, Brenner R, et al.(1996). JAMA 275: 390–397. Cookson R, Mc Daid D. (2003). Health Policy 63: 133–139. Dasbach E, Elbasha E, et al.(2006). Epidemiol Rev 28: 88–100. Edmunds WJ, Brisson M, et al.(2001). Vaccine. 19: 3076–3090. Edmunds WJ, Brisson M, et al.(2002). Vaccine. 20: 1316–1330. Edmunds WJ, Medley GF, et al.(1999). Stat Med. 18: 3262–3282. Ekwueme DU, Strebel PM, et al.(2000). Arch Pediatr Adolesc Med. 154: 797–803. Evers S, Ament A, et al.(2007). Eur J Clin Microbiol. 26: 531–540. Frazer I. (2007). Int J Infect Dis. 11: S10–S16. Gay NJ, Pelletier L, et al.(1998). Vaccine. 16(8): 794–801.
1351
1352
77
Economics and Vaccines
Gwatkin DR, Guillot M. (1998). The Burden of Tropical Diseases Among the Poorest and Richest 20% of the Global Population. TDR/ER/RD/96.1, World Health Organization. Hay J, Daum R, et al.(1987). Pediatrics 80(3): 319–329. Hinman AR, Irons B, et al.(2002). Bull World Health Organ 80(4): 264–270. Iskedjian M, Walter JH, et al.(2004). Vaccine. 22: 4215–4227. Jacobs R, Koff R, et al. (2002). Am J Gastroenterol. 97: 427–434. Jefferson T, Behrens R, et al. (1994). Vaccine. 12: 1379–1383. Jit M, Edmunds W. (2007). Vaccine. 25: 3971–3979 Johannesson M, Meltzer D. (1998). Health Econ. 7(1): 1–7. Kahn MM, Ehreth J. (2003). Vaccine. 21: 702–705. Martens L, ten Velden G, et al. (1991). Ned Tijdschr Geneeskd. 135(1): 16–20. Miller E, Marshall R, et al.(1993). Rev Med Microbiol. 4: 222–230. Miller MA, Redd S, et al. (1998). Vaccine. 16(20): 1917–1922. Paavonen J, Jenkins D, et al.(2007). Lancet. 369: 2161–2170. Parkin D, Bray F. (2006). Vaccine. 24: S11–S25. Pelletier L, Chung P, et al. (1998). Vaccine. 16(9): 989–996. Platonov A, Griffiths U, et al. (2006). Vaccine. 24(13): 2367–2376. Plotkin SA Mortimer EA. (1994). Vaccines, 2nd ed. W.B. Saunders Company, New York. Pokorn M, Kopac S, et al. (2001). Vaccine. 19(25–26): 3600–3605. Postma M, Jansema P, et al. (2002). Drugs. 62(7): 1013–1024. Postma MJ, Bos JM, et al. (2003). Eur J Health Econ. Purdy KW, Hay JW, et al. (2004). Clin Infect Dis. 39: 20–28. Sellick JA, Jr., Longbine D, et al. (1992). Ann Intern Med. 116(12 Pt 1): 982–984.
Se´ne´cal M, Brisson M, et al. (2006). Quality of Life Lost Associated with Rotavirus Gastroenteritis in Canadian Families: A Prospective Community Based Study. 7th Canadian Immunization Conference, Winnipeg. Seward JF, Jumaan AO, et al. (2002). JAMA. 287: 2211–2222. Shiell A, Jorm LR, et al.(1998). Aust N Z J Public Health. 22(1): 126–132. Sisk J, Whang W, et al. (2003). Ann Intern Med. 138. Stevenson M, Beard S, et al.(2002). Vaccine. 20: 1778–1786. Stover BH, Adams G, et al. (1994). Infect Control Hosp Epidemiol. 15(1): 18–21. Thiry N, Beutels P, et al. (2003). Pharmacoeconomics. 21: 13–18. Thompson KM, Duintjer Tebbens RJ. (2006). Risk Anal. 26: 1423–1440. Thompson KM, Duintjer Tebbens RJ. (2007). Lancet. 369: 1363–1371. Tormans G, Van Doorslaer E, et al. (1998). Eur J Pediatr. 157: 395–401. Vesikari T, Matson D, et al. (2006). N Engl J Med. 354: 23–33. Welte R, Trotter CL, et al. (2005). Pharmacoeconomics. 23(9): 855–874. White CC, Koplan JP, et al. (1985). Am J Public Health. 75: 739–744. Whitney G, Farley M, et al. (2003). N Engl J Med. 348: 1737–1746. WHO. (2007). The World Health Report 2007 – A safer future: global public health security in the 21st century. WHO. (2008). Immunisation fact sheet Nr 288. Widdowson M, Meltzer M, et al. (2007). Pediatrics 119: 684–697. Wood S, Nguyen V, et al. (2000). Pharmacoeconomics. 18(2): 173–183. Zwanziger J, Szilagyi PG, et al. (2001). Health Serv Res. 36(5): 885–909.
78 Economic Costs and Disability-Adjusted Life Years in Polio Eradication: A Long-Run Global Perspective M. M. Khan 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1354
2 2.1 2.1.1 2.1.2 2.2
How Do We Calculate the Costs and the Effectiveness? . . . . . . . . . . . . . . . . . . . . . . . 1356 Estimating Polio Cases With and Without Vaccination . . . . . . . . . . . . . . . . . . . . . . . . . . 1357 Choosing a Flexible Equation for Empirical Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1357 Using the Estimated Equation for Prediction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1358 Estimating the Disability Adjusted Life Years . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1359
3 3.1 3.1.1 3.2
Deriving the Cost Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1360 Cost Parameters for Routine Polio Immunization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1360 Price of Vaccine and Cost per Dose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1361 Cost Parameters for Polio Eradication Program . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1363
4 4.1 4.1.1 4.2 4.3
Global Cost and Effectiveness of Polio Vaccination . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1363 Number of Polio Cases Averted and Dalys Saved . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1363 Calculating the Effectiveness of Polio Vaccination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1366 Net Cost of Routine Polio Immunization and the Cost-Effectiveness Ratios . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1368 Cost-Effectiveness of Polio Eradication Activities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1369
5
Conclusions and Policy Implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1370 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1371
#
Springer Science+Business Media LLC 2010 (USA)
1354
78
Economic Costs and Disability-Adjusted Life Years in Polio Eradication
Abstract: This paper estimates the costs and effectiveness of polio immunization and eradication activities from long-term global perspective. The effectiveness of immunization has been expressed in terms of Disability Adjusted Life Years (> DALYs) averted over the years 1970–2050 and the aggregate effectiveness was calculated using 2007 present values. The projected annual numbers of polio cases without vaccination were obtained from polio caseloads in pre-vaccination era in the USA and in Europe. Costs of routine polio immunization and eradication activities were also expressed in terms of 2007 dollars. The results indicate that polio immunization will prevent more than 42 million cases of polio, four million paralysis and 855 thousand deaths over the years 1970–2050. In terms of 2007 dollars, the aggregate global cost of polio immunization over 1970–2050 will exceed $123 billion. Despite the high cost of the program, direct medical care cost savings associated with polio prevention are so high that the program becomes cost-saving at the global level. Excepting for few low income countries in Africa and South-east Asia, routine polio immunization remains a > highly costeffective intervention for all countries of the world. Even for poorer countries, polio immunization and eradication activities remain cost-effective. However, very poor countries will require continued financial assistance to help achieve the full economic benefit of polio immunization and eradication. List of Abbreviations: AFP, acute flaccid paralysis; AFR, Africa region of WHO; AMR, region of the Americas; DALY, disability adjusted life years; EMR, eastern mediterranean region; > EPI, expanded program for immunization; EURO, European region; GNI, > gross national income; ICD10, International Classification of Disease version 10; > IPV, inactivated polio vaccine; NIDs, National Immunization Days; > OPV, oral polio vaccine; SEAR, south-east Asia region; WPR, western pacific region; YLD, years lost due to disability for the incident cases of health conditions; YLL, years of life lost due to premature death
1
Introduction
The World Health Assembly in 1988 adopted a set of strategies to eradicate polio by the year 2000. The eradication program consisted of a number of specific interventions and targets: immunization of 90% or more children against polio with four doses of Oral Polio Vaccine (OPV) by 12 month of age as part of the Expanded Program on Immunization (EPI); conducting National Immunization Days (NIDs) to provide two doses of OPV to all children less than 5 years and strengthening surveillance and laboratory investigation of suspected polio cases. It was suggested that the NIDs be carried out for at least three consecutive years and all countries should conduct surveillance and laboratory investigation of any case of Acute Flaccid Paralysis (AFP) for children less than 15 years and any case of polio at any age. If the surveillance system identified transmission of wild poliovirus in a geographic area, the plan required immediate adoption of mop-up campaigns to vaccinate children with two extra doses of OPV. The eradication activities adopted have shown significant accomplishments since 1990. > Figure 78‐1 shows the time trend of total global cases reported from 1980 to 2007 and over the years the number of cases declined at a very rapid rate. The reported case-load of confirmed polio exceeded 50,000 per year in 1980 (the actual case count, including the ones without any long-term symptoms, could be as much as ten times the reported numbers) and expansion of regular polio immunization helped reduce the case-load by more than 50% to
Economic Costs and Disability-Adjusted Life Years in Polio Eradication
78
. Figure 78‐1 Global total of polio cases by year: 1980–2007
about 23,000 by 1990 (WHO, 2000). Adoption of polio eradication strategies in early 1990s further reduced the number of reported cases and by the year 2000 polio cases dipped below the three thousand level (WHO, 2002). Although the plan was to eradicate polio by the year 2000 (CDC, 1998), the average number of polio cases remained at around 1,600 over the years 2000–2007. This is despite the fact that the number of polio endemic countries, i.e., countries with indigenous wild poliovirus presence, reduced from 50 in 1999 to 20 in 2001. By early 2008, only four countries, Nigeria, India, Pakistan and Afghanistan remained polio endemic (WHO, 2008). According to the World Health Organization (WHO) calculations, polio cases in 2005 were responsible for loss of 13,261 Disability Adjusted Life Years (DALYs) compared to about 5.7 million DALYs lost due to pertussis and 9.8 million due to measles (Mathers and Loncar, 2006). The low burden of polio in 2005 clearly shows dramatic success of polio immunization and eradication activities. It is important to point out that a significant part of polio related loss of DALYs in 2005 was due to premature deaths. For example, the burden of polio in Americas was 2,361 DALYs in 2005 and all of it was due to premature deaths. Most of these cases were in the age group 45 years old or older indicating that many of the polio-related deaths in 2005 were the result of polio transmission that happened prior to early 1960s. It became clear by mid-1990s that polio eradication cannot be achieved by 2000 as planned and a new target date of 2005 was set. Polio outbreaks during 2003–2005 again frustrated the polio eradication target-date. In November 2007, high-level polio consultation agreed to intensify immunization activities in the remaining endemic countries to stop poliovirus transmission for good (WHO, 2008). To complicate the polio eradication efforts further, poliovirus importation affected another ten countries during 2006–07 and clearly global polio eradication certification date must be pushed back by few additional years and is not expected before 2010 at the earliest. Despite the fact that polio eradication has remained elusive, significant progress has been made in controlling polio through regular and intensive immunizations. At this stage
1355
1356
78
Economic Costs and Disability-Adjusted Life Years in Polio Eradication
of polio control, focusing the discussion narrowly on eradication strategies is probably counter-productive. Preventing polio require continued vigilance in terms of immunization coverage and outbreak responses. Given the uncertainty associated with eradication and high levels of health benefits derived from regular polio immunization, policy makers should emphasize continued polio immunization in general. This paper demonstrates that polio immunization program itself has been one of the greatest success stories of modern public health interventions. If we consider the > economic costs and health benefits of regular polio immunization activities alone, the benefits clearly are high enough to justify continuation of polio vaccination in the longer run, irrespective of polio eradication status. A separate analysis can be conducted to examine the benefits and costs of polio eradication. To better understand the success of polio control strategies, analyses should subdivide the interventions into two discrete sets of activities: control of polio through regular immunization and supplemental and/or intensive polio vaccination for eradication. After the successful global eradication, regular polio immunization will not be required and the world will save money by stopping polio vaccination altogether. Complete cessation of polio vaccination, although feasible in post-eradication era, may not be a practical option any more. Fear of accidental and intentional reintroduction of poliovirus in the environment made many countries, especially the developed world, unwilling to discontinue polio immunization. In this sense, financial benefit of eradication may not be fully realized and from policy perspective it is much more important to discuss the long-run > financial costs and benefits of continued polio immunization rather than eradication. This paper is an attempt to estimate the long-run global economic costs and effectiveness of polio vaccination program. Since the polio control strategies have been divided into control and eradication activities, the paper will also discuss potential incremental costs and benefits of polio eradication. This involves estimating the economic costs of intensive supplementary immunization, establishment of surveillance system and laboratories and final certification cost of polio eradication. The estimation of benefits in terms of health outcomes require estimation of deaths and paralysis prevented due to the immunization program. The WHO suggested method (Murray and Lopez, 1996) will be used to convert pre-mature deaths and cases of paralysis prevented into Disability Adjusted Life Years (DALYs). On the cost side, vaccine and vaccine delivery costs will be considered. Since the prevention of polio cases saves medical resources, net cost of immunization will be derived by subtracting the medical care costs from the cost of the immunization program.
2
How Do We Calculate the Costs and the Effectiveness?
This paper is a > cost-effectiveness analysis of polio vaccination and eradication. The effectiveness of polio vaccination is defined as the polio cases averted (paralysis and deaths averted). To obtain an estimate of the cases averted, we need to derive the number of cases that would have occurred if there were no vaccinations. Cases of polio without vaccination were estimated from the historical data on polio cases in the USA and Italy. Cost-effectiveness of eradication was estimated by assuming that eradication will allow discontinuation of vaccination. So, the future vaccination cost saved becomes the benefit of polio eradication activities. It is also assumed that a small number of hard-to-prevent cases can be prevented only through eradication activities.
Economic Costs and Disability-Adjusted Life Years in Polio Eradication
78
To estimate the global health outcomes and costs related to polio immunization and eradication, this study will use the WHO region-specific estimates for the analysis. The WHO divided the world into six regions: African region (AFR), Region of the Americas (AMR), Eastern Mediterranean Region (EMR), European Region (EUR), South-East Asia Region (SEAR) and Western Pacific Region (WPR). For each of the regions, the study estimates the number of cases one would have expected if there were no polio vaccination. The predicted case-load allows estimation of number of deaths and disability averted due to polio vaccination. The medical care costs and other associated costs will also be derived by using regionspecific average cost information.
2.1
Estimating Polio Cases With and Without Vaccination
To estimate the number of cases averted since the introduction of vaccination program, time series modeling was carried out to predict the number of polio cases with and without vaccination. The prevalence of polio prior to the introduction of polio vaccination was used to derive the yearly cases if there was no vaccination. For each of the WHO regions, efforts were made to collect incidence rate of polio prior to the introduction of vaccines. For the Region of the Americas (AMR), information was collected on population, population under five and the polio cases from 1912 to 1954 for USA. From the polio cases reported, rate per 100,000 children were derived so that we can use the rates to project polio cases in the whole of AMR in absence of vaccination. For the European Region, time series data for pre-vaccination period were obtained for Italy. The Chinese rates are also available for about 25 years prior to 1970, but the incidence rates reported for China are clearly too low. It is likely that polio cases were underreported by a wide margin for China and for other developing countries of the world. Even in South America the rates were underreported by as much as 80% (Musgrove 1989). Therefore, to predict the cyclical changes in polio incidence pattern, it will be better to use the information from a country with relatively more reliable data. For most of the regions of the world, we have used Italy’s time pattern of polio rates over the pre-vaccination years. The reported cases for each of the regions were inflated to make the reported rates equal to predicted rates on the average for the pre-immunization years in each of the regions.
2.1.1
Choosing a Flexible Equation for Empirical Analysis
One problem with statistical prediction of polio incidence rates is its relatively high year-toyear variability. The incidence rate per 100,000 children varied from less than 30 to more than 300 in the USA over the 43 pre-vaccination years. Therefore, to project the incidence rates in the future, we need a flexible time series modeling technique, which will allow frequent ups and downs within a short time frame. If we use the average incidence rate of pre-vaccination period, it will distort the cost-effectiveness ratios because in an analysis with long time horizon, both costs and effectiveness are affected significantly by the timing of the costs and outcomes. Another problem with time series modeling is that predicted values tend to become either explosive or asymptotic when the period of prediction is long. For our analysis, it is important
1357
1358
78
Economic Costs and Disability-Adjusted Life Years in Polio Eradication
to identify a model that will allow cyclical ups and downs over the years, for at least 30–40 years in the future, in a predictive model. Polio incidence numbers are such that after a high incidence rate in a year it tends to go down for sometime before increasing again. After examining the nature of polio incidences during the pre-vaccination years, it was decided to use a > Fourier series model for projecting polio cases. The advantage of the Fourier series is that since it is based on sine and cosine functions, it can actually show the ups and downs expected in the future. Moreover, the series can indicate abrupt changes in the function. The model applied for the time series analysis of polio incidence rates is described below. Assume that the time series of incidence rates is represented by a vector [R]. At the first stage, we need to standardize the elements of the vector by finding the mean and standard deviation of the elements. The new vector defined after standardization becomes [Y]. Therefore, Yi ¼
Ri m s
where m = mean value of Ri, and s = standard deviation of Ri The time series analysis is then carried out on Y values. The functional form we have used is:Yi ¼ a Cosðw1 tÞ þ b Sinðw1 tÞ þ e1 In this equation, w is the splicing factor and a and b are the parameters to be estimated. We start the analysis by choosing a very low value for w, such as 0.00001. Given the small w, we find a and b so that the sum of squared residuals are minimized, i.e., using the standard regression approach, we want to minimize X ½Yi a Cosðw1 tÞ b Sinðw1 tÞ2 After obtaining the minimizing values of a and b, both the predicted values of Y, say F1, and the error terms (E1) were retrieved for further analysis. At the second stage, the residual values were used as the dependent variable and the earlier stage w value was multiplied by about 10. With the new value of w, again the sum of squared errors was minimized by finding a new set of values of a and b. The error terms (E2) and the predicted values (F2) from the second stage are retained if the standard deviation of the error term after the second stage is lower than the standard deviation of the error term from the first stage. The iterations are continued till the reduction in the standard deviation of the error term is not significantly different from that in the previous stage. If we stop our iterations at the nth round, the predicted Y value becomes ^ ¼ F 1 þ F 2 þ F 3 þ ::::::: þ Fn Y Since the addition of sine and cosine values can represent many different types of irregular functions, the predicted values should be very close to the actual values of polio incidence rates. Nonlinear regression module of standard statistical package programs can be used to find the estimates of the parameters. Once the parameters are derived, the error terms are calculated and another nonlinear regression is run with the error terms as the dependent variable. The SPSS non-linear regression module was used to estimate the above function.
2.1.2
Using the Estimated Equation for Prediction
The estimated Fourier function can now be used to predict the > standardized values of polio incidence rates expected without vaccination. The problem with too long prediction
Economic Costs and Disability-Adjusted Life Years in Polio Eradication
78
period is that the estimates become less and less reliable as we move further and further away from the actual data points. This is especially a problem for polio because the pre-vaccination rates are available for the years prior to 1960s and so we need to predict the rates for about 100 years. To reduce errors due to long time frame, polio incidence rates were predicted for no more than 30 years and the actual reported points plus 30 predicted points were used repeatedly to find the incidence rates beyond the 30-year limit. For the estimation of costs and effectiveness of polio control, we have used projected incidence rate of polio till 2050. The actual number of polio cases by WHO regions is available since early 1970s but the reported incidence rates are quite unreliable. For example, the incidence rates reported for China in 1970s was only about 10% of Italy’s rate in pre-vaccination era. In fact, polio cases were underreported in all regions of the world including Americas and Europe but the degree of underreporting is likely to be low for the later two regions, where more active surveillance systems were in place before 1970. For estimating the number of cases averted, we have assumed that for all WHO regions expecting Americas and Europe the number of cases averted should be close to zero for the years 1970–1975. The reported average number of polio cases for these years were multiplied by a fixed factor to make the average equal to the average of predicted values. The same fixed factor was used to adjust the whole series of reported numbers. A comparison of the number of cases expected (derived by using Fourier series) since 1970 with the corrected actual number of cases has been used as the number of acute polio cases prevented through immunization.
2.2
Estimating the Disability Adjusted Life Years
We have used WHO proposed method for estimating the Disability Adjusted Life Years (DALYs) for polio cases (ICD10 code A80). The WHO method estimates the years lost due to premature deaths and the years lost due to lameness or paralysis. Therefore, by using the number of acute polio cases averted, we need to derive the number of paralysis and deaths averted per year. In WHO methodology, surveillance data on laboratory confirmed cases of wild virus infection in symptomatic persons were used to estimate global polio cases. The infections lead to lameness in some children and the Years Lived with Disability (YLDs) were calculated for the paralysis cases by assuming that the age at onset is 2.5 years for all regions except Africa (2.4 years for Africa). Since we are using the WHO method of estimating DALYs, we will simply use the parameters WHO has used in their estimation of Global Burden of Disease. The disability associated with polio is assumed to continue over the lifetime of the individual. To calculate the YLDs due to lameness, WHO used a > disability weight of 0.369. The methodology of obtaining disability weights has been discussed in Murrey and Lopez (1996). In our study, we do not have information on paralysis cases by year, especially for the years prior to 1970s. To get an estimate of paralysis cases without vaccination in pre-vaccination era, we have used one study that reported the probabilities of paralysis and central nervous system problems for North America and Western Europe in 1960s. Over the years 1964–1967, WHO received information on the isolation of viruses from different countries of the world. The data for North America and Europe are relatively complete and can be used to define the probability of paralysis and death. The reported numbers from these two regions indicate that 20.8% of all polio cases create disease of central nervous system and 10.27% cases lead to paralysis (Cockburn and Drozdov, 1970).
1359
1360
78
Economic Costs and Disability-Adjusted Life Years in Polio Eradication
The WHO also assumes that the number of deaths from polio is equal to 10% of the lameness and other nervous system attack cases. We have used the same assumption to derive the number of deaths due to polio (WHO, 2006). Given the estimates of paralysis and nervous system attacks cases, the number of deaths can be projected using the above parameters. Years of life lost due to pre-mature death is based on DALY assumption on life expectancy at age 3 years, which is 77.8 years or 30.9 discounted years without ageweighting. Since the YLDs and YLLs are distributed over a long period of time, all health outcomes and costs were converted into values specific to the year 2007. Again, using the method followed by the WHO, discount rate chosen for the analysis was 3%. For each of the years, the estimated numbers of deaths were multiplied by the number of years lost due to death to get the total YLLs. Although the standard life expectancy at age three was 77.8 years, the duration of disability was assumed to be about 10 years shorter on the average. Therefore, discounted value of years of disability of a polio case becomes about 29.6 years. For the years of life saved prior to 2007, a multiplicative factor was used and after 2007, the discount factor was applied.
3
Deriving the Cost Parameters
3.1
Cost Parameters for Routine Polio Immunization
The aggregate economic cost of routine immunization will be derived for all years starting from 1970 to 2050. In early 1970s, polio immunization was being strengthened and implemented around the world. In routine immunization programs, polio is not the only vaccine being delivered and so the cost of polio immunization can be estimated by appropriately apportioning the routine immunization cost to polio component. The economic savings associated with complete cessation of polio vaccination can also be approximated by using the average cost per dose for non-vaccine cost items plus the cost of procuring and handing the vaccine. > Table 78‐1 below lists the cost of polio immunization reported by various country level studies. One striking aspect of the numbers in the table is the wide variability of polio immunization cost across different countries of the world. Cost per dose in New York City immunization program was about $114 in 1993 prices while the number was less than $1.50 for many poor developing countries. Since the costing studies used different approaches of costing, it is important to identify cost parameters for country-groups (based on income per capita of countries) when all resources used in the delivery of immunization services are accounted for. One cost component is the vaccine cost and although the cif price of vaccines should not vary significantly across countries, in-country handling and processing costs differ widely based on its economic status. Although Bart et al. (1996) made an attempt to define the vaccine and non-vaccine cost differences between developed and developing countries, global cost estimates can be further improved by defining the average cost by economic status of countries and by geographic regions. For estimating the polio immunization cost for a geographic region, a weighted average of country-group costs will be used. For each of the regions of the world, proportion of population living in income-based country categories (low income, middle income, upper-middle income and high-income) were used as weights to derive the region-specific cost parameters (World Bank, 2006).
Economic Costs and Disability-Adjusted Life Years in Polio Eradication
78
. Table 78‐1 Cost per dose for polio immunization activities as reported by costing studies from different countries of the world
Country
Year of study
Vaccine cost per dose
Other cost per dose
Morocco
1999
$0.10
$1.12
1.245
$1.394
Kaddar et al. (1999)
Bangladesh
1999
$0.13
$1.04
1.245
$1.295
Kaddar et al. (2000); Khan and Yoder (1998)
Philippines
1990
$0.15a
$1.61
1.586
$2.553
Brenzel and Claquin (1994)
Tanzania
1990
$0.11a
$0.78
1.586
$1.237
Brenzel and Claquin (1994)
Mauritania
1990
$0.12a
$0.87
1.586
$1.380
Brenzel and Claquin (1994)
Turkey
1990
$0.15a
$2.36
1.586
$3.743
Brenzel and Claquin (1994)
Burkina Faso
1990
$0.10a
$1.65
1.586
$2.617
Brenzel and Claquin (1994)
USA (NY City)
1993
$5.00a
Developing countries
1993
$0.11b
$1.51
1.435
$2.167
Bart et al. (1996)
Developed countries
1993
$4.58b
$5.09
1.435
$7.304
Bart et al. (1996)
$109
Inflation factor, 2007 prices
1.435
Other cost/ dose in 2007 prices
$156
Source of cost information
Fairbrother and DuMont (1995)
a
These studies did not report vaccine cost separately. Assumed vaccine cost, including in-country handling, has been used b Cost per dose includes wastage related costs. Wastage is assumed to be 33% of vaccine doses delivered for developing countries and 10% for developed countries This table shows polio vaccine prices and other costs associated with administering polio vaccination in different countries of the world. Since the costing studies were conducted at different points in time, reported costs were adjusted to express all numbers in terms of 2007 prices. Annual inflation rates for the years 1990–2007 were obtained for the USA and the inflation rates were used for the adjustment
3.1.1
Price of Vaccine and Cost per Dose
Procurement price of OPV in recent years has increased quite significantly due to low production capacity and uncertainty about OPV demand in the near future and using this level of international price will distort the costing exercise. To ensure that we are considering the economic cost of resources used in costing the polio programs, OPV procurement price in 2001 may be considered the opportunity cost of the vaccine. The UNICEF procurement office reported that a dose of OPV in 2001 was about $0.09, equivalent to inflation adjusted price of $0.11 per dose in 2007. With this international price of vaccines, we have to add resource costs related to handling, transportation and wastage of the vaccines (see costing methodology used by Liu et al. 2002).
1361
1362
78
Economic Costs and Disability-Adjusted Life Years in Polio Eradication
For the low-income countries of the world, the minimum vaccine cost should be about $0.15 after adding all these additional cost items. From > Table 78‐1, we can say that the average cost of administering vaccine per dose (other costs) is about $1.30 for low-income countries and so the total cost per dose becomes $1.45 including the vaccine cost. Allowing proportionate increase in costs based on the average income of the group of countries, the cost of polio vaccine per dose becomes $3.46 for middle-income countries, $9.22 for upper-middle income countries and $15.15 for high-income countries of the world. The cost for high-income countries already includes the cost of using IPV in the regular immunization program and no further adjustments in the cost numbers will be needed for this country-category. > Table 78‐2 shows the country-group specific weights used for the derivation of average cost per dose of polio-vaccine administered by geographic regions (the WHO regions). . Table 78-2 Cost of administering polio vaccination, including the vaccine costs, by WHO regions for the year 2007 Percent of population living in countries categorized as WHO regions
Low income
African region
88
6
6
0
$2.04
2
20
40
38
$10.17
60
34
5
1
$2.66
Americas Eastern Mediterranean
Middle income
Upper-middle income
Highincome
Average cost per dose
European region
4
48
4
44
$8.75
South-east Asia
80
20
0
0
$1.85
Western Pacific
81
5
5
9
$3.17
Cost of administering or delivering a dose of polio depends on the average income levels of countries. In lowincome countries, cost of delivering vaccines becomes low mainly due to low labor costs. To find the average cost per dose of polio vaccine delivered for a geographic region, costs per dose by income category of countries were multiplied by the percent of population in the geographic region living in that category of countries. Average cost of delivering a dose of polio vaccine for low-income, middle income, upper-middle income and high income countries were multiplied by the population proportions mentioned in columns 2–5. Weighted averages of the costs are shown in the final column of the table
The calculated average costs per dose are also reported in the table. Note that cost per dose administered includes vaccine wastage, which is relatively high for developing countries of the world. The cost parameters derived and reported in the table are used in estimating the global cost of polio vaccination through regular immunization programs. To find the total cost of polio vaccination, we also need information on > vaccine coverage rate by country groups or by regions. It is assumed that the vaccine coverage in developed countries increase from about 60% in 1970 to about 90% by the year 2000. In many developed countries, vaccine coverage (crude) remained around 60% in early 1970s and due to low coverage in some countries. The average coverage rate among high income countries after 2000 is assumed to be at 95%. For developing countries of the world, the coverage of three doses of polio vaccine was only 15% in 1970, which increased to 60% in 1985 and 80% in 1995. After that the coverage is assumed to remain at 90% level.
Economic Costs and Disability-Adjusted Life Years in Polio Eradication
3.2
78
Cost Parameters for Polio Eradication Program
The eradication program has five distinct components: the supplementary immunization activities (> NIDs and SNIDs), AFP Surveillance, AFP Surveillance laboratory costs, mopping up immunization activities, polio eradication certification activities. These activities are organized in different countries and regions at different points in time depending upon the progress in routine immunization coverage and success in reducing polio incidence rates. For example, polio eradication campaign was introduced in the Americas in 1986, when a 5-year intensive immunization program was adopted. The eradication activities in South Asia and Africa did not start before 1995. Since the cost of eradication occurred in different years, appropriate > discounting factors must be used to derive the eradication costs in terms of 2007 prices. Bart et al. (1996) reported some cost parameters related to intensive polio immunization. They assumed a cost of $0.18 per dose for developing countries and $5.64 per dose for developed countries (in 1993 prices). In this paper, cost parameters are defined for four categories of countries using the costs reported in the literature (Bart et al., 1996; Jian et al., 1998; Levin et al., 2000a,b). The estimated cost per dose delivered through intensive immunization programs for low-income, middle-income, upper-middle income and high-income countries were found to be $0.30, $1.14, $2.94 and $6.77 respectively. Using the same population weights in > Table 78‐2, weighted cost per dose of polio through intensive programs becomes: $0.51 for Africa, $3.98 for Americas, $0.78 for Eastern Mediterranean, $3.66 for Europe, $0.47 for South East Asia and $1.06 for Western pacific. Excepting for the South East Asian Region (SEARO), no other WHO regions report the cost of eradication activities. For SEARO, we have used the costs reported by the WHO and for other regions, eradication costs were estimated using the assumptions used by Bart et al. (1996) on start date of the campaign, coverage of vaccination, etc. For example, we have assumed that in Eastern Mediterranean Region, eradication activities began in 1992 targeting 10% of children and in South East Asia it began in 1994 covering 90% of children.
4
Global Cost and Effectiveness of Polio Vaccination
The parameters described in sections II and III above are presented in > Table 78‐3 below. These parameters will be used to calculate the global cost of polio immunization.
4.1
Number of Polio Cases Averted and Dalys Saved
Using the parameters reported in > Table 78‐3 and the estimates of number of polio cases in all the six WHO regions, we can calculate the number of polio cases averted and the number of deaths prevented through routine immunization and eradication programs. > Figure 78‐2 shows the actual and predicted values of polio incidence rates (per 100,000 children below 5 years of age) for USA during the pre-vaccination period. Note that the predicted values almost perfectly predict the actual observations. The estimated polio incidence rate function was used to project the number of cases without vaccination for the Americas. Similarly, incidence rates of polio in Italy for pre-vaccination period (1924–1963) were used to predict the polio incidence rates for European and other regions of the world
1363
1364
78
Economic Costs and Disability-Adjusted Life Years in Polio Eradication
. Table 78-3 Parameters used for the estimation of costs and effectiveness of polio vaccination and eradication activities Parameters for measuring cost and effectiveness Proportion of polio cases likely to become paralyzed Proportion of polio cases with other nervous system attacks Duration of paralysis (discounted years) Disability weight used for paralysis Case fatality rate (per 100 polio cases) Years of life lost due to premature death (discounted years)
Parameter values 0.1027 0.1053 68 years (29.6 years) 0.369 2 78 years (30.9 years)
Routine immunization cost parameters (in 2007 prices) for Low-income countries
$1.45
Middle-income countries
$3.46
Upper-middle income countries High income countries
$9.22 $15.15
Eradication program cost parameters (in 2007 prices) for Low-income countries
$0.30
Middle-income countries
$1.14
Upper-middle income countries
$2.94
High income countries
$6.77
Medical care cost per year per case and percent of cases utilizing medical care for (% utilization rates are in parentheses) Low-income countries
$504 (20%)
Middle-income countries
$2,364 (40%)
Upper-middle income countries
$5,400 (65%)
High income countries
$30,000 (80%)
This table shows the parameters used in this paper to calculate the costs and effectiveness of polio immunization program. First seven rows report the parameters needed to calculate the health outcomes related to polio. Almost 10% of polio cases lead to long-term paralysis and another 10% are afflicted with other types of nervous system attacks. Those who are paralyzed, the duration of paralysis is considered to be 68 years. Disability weight indicates the loss of health associated with a disease. The WHO assumed that the disability weight for polio-related paralysis is 0.369, i.e., living with paralysis for 1 year is equivalent to living in perfect health for (1–0.369) = 0.631 year. Cost of immunization per dose and medical care cost of polio cases are also reported. In this paper, medical care cost is defined by the median income of countries in the category. For high-income countries in Americas, medical care cost per year is considered $30,000 (equivalent to Bart et al. (1996) medical cost numbers for developed countries). For other high income countries, medical care cost per case of polio paralysis or nervous system attack is considered $18,000 per year
(AFR, SEAR, WPR and EMR). The Fourier series estimation predicted the observed incidence rates for Italy almost perfectly. As mentioned in the method section, the estimated Fourier series were used to predict polio incidence rates for all the years from 1912 to 2050. > Figure 78‐3 shows the predicted incidence rates per 100,000 children based on US incidence data for pre-vaccination years. This predicted series indicates the incidence of polio that would have occurred without polio immunization in the Americas.
Economic Costs and Disability-Adjusted Life Years in Polio Eradication
78
. Figure 78-2 Actual and predicted polio incidence rates in the USA, 1912–1955
. Figure 78‐3 Observed and predicted polio incidence rates without vaccination in the Americas: 1912–2050
1365
1366
78 4.1.1
Economic Costs and Disability-Adjusted Life Years in Polio Eradication
Calculating the Effectiveness of Polio Vaccination
The effectiveness of polio vaccination is simply the difference between the projected polio cases without vaccination and actual polio cases with vaccination. The population of the world was about 6,055 million in 2000 and it is expected to increase to 8,560 million in 2050. If the polio cases prevented are calculated for each of the six regions of the world, global total of polio cases prevented over the years 1970–2050 (without discounting) becomes about 42 million. More than four million paralysis cases will be potentially averted by 2050 and immunization will also avert another four million cases of other nervous system attacks. The undiscounted number of deaths prevented due to successful vaccination and eradication over the years should be about 855 thousands. Given the number of deaths and disabilities, we are now in a position to convert these health outcomes into DALYs. It should be noted that all the deaths and disabilities are not prevented in the same year and therefore, we need to express the health outcomes in terms of 2007 values. In other words, all the health outcomes prior to 2007 will be multiplied by the discount factor and all the health outcomes after 2007 will be divided by the discount factor. As mentioned earlier, we will use a discount rate of 3%, the rate used by the WHO in their DALY calculations. To illustrate the DALY calculation steps followed, let us look at the data set used for Africa Region. > Table 78‐4 shows data on actual incidence and projected polio incidence without
. Table 78‐4 Adjusted incidence and projected incidence without polio immunization with DALY calculations for Africa region (1981–1990) Cases Paralysis Polio prevented, prevented cases* (2) = [pred- (3) = [(2) Year (1) (1)] *0.1027]
Deaths prevented (4 ) = [2% of (2)]
YLDs prevented^ (5) = [(3)* 0.369*29.6]
YLLs prevented DALYs (6) = [(4)* prevented 30.9] (7) = [(5) + (6)]
1981 62,865
27,698
2,845
576
31,071
17,803
1982 58,455
45,021
4,624
936
50,502
28,936
48,874 79,438
1983 45,990
90,773
9,322
1,888
101,823
58,342
160,165
1984 44,520
96,675
9,929
2,011
108,443
62,135
170,578
1985 59,025
73,835
7,583
1,536
82,823
47,455
130,278
1986 51,375
80,788
8,297
1,680
90,623
51,925
142,548
1987 41,730
100,118
10,282
2,082
112,306
64,348
176,654
1988 68,145
72,749
7,471
1,513
81,605
46,758
128,363
1989 46,440
131,840
13,540
2,742
147,890
84,737
232,626
1990 63,420
98,752
10,142
2,054
110,774
63,470
174,244
*Adjusted for underreporting of polio cases This table shows the total number adjusted polio cases in column 2 and the number of cases averted in column 3. Cases averted have been estimated by projecting the number of cases that would have occurred in absence of polio immunization and subtracting the adjusted incidence number from the estimate. The numbers of averted cases were used to find the number of paralysis and deaths averted. Years lost due to premature mortality (YLL) has been derived from the estimates of deaths averted and Years lost due to disability (YLD) has been derived from paralysis cases averted. The calculations in the table ignored the loss of health associated with other nervous system attack cases
Economic Costs and Disability-Adjusted Life Years in Polio Eradication
78
vaccination for the years 1981–2000. Prior to 1981, it is assumed that polio vaccination was not effective enough to avert any case. This actually underestimates the effectiveness of polio vaccination as the immunization program obviously had some impact on preventing polio before 1980. Also, if a particular year shows high incidence of polio (adjusted number due to underestimation of cases), higher than or equal to the projected incidence without vaccination, number of cases prevented in that year also becomes zero. The DALY numbers are by year by geographic region and to get an aggregate measure of effectiveness of polio immunization, we need to add the DALY values together. Since the DALYs prevented are distributed over a long period of time, we have used 2007 as the base year and all DALY values were adjusted so that the aggregate number becomes same as the equivalent DALYs saved in 2007 alone. As before, 3% discount factor was used for the conversions. For each of the WHO regions, similar calculations were carried out for the years 1970–2050 and for each of the years number of cases prevented and other health outcomes were derived. The discounted values of DALYs were calculated from the health outcome information and the numbers are presented in > Table 78‐5. Note from > Table 78‐5 that DALYs averted are relatively high for the Americas and the reason for this is mainly due to two factors: first, the reported incidence of polio were
. Table 78‐5 Estimated health impacts of routine polio immunization in the WHO regions, 1970–2050 (in thousands of years)
WHO regions Africa Americas Eastern Mediterranean Europe South-east Asia Western Pacific Global total
YLDs prevented YLLs prevented (w/o (w/o discounting) discounting)
Discounted YLDs (discount rate 3%)
Discounted YLLs (discount rate 3%)
Discounted DALYs
7,003.91
4,013.04
5,517.02
4,006.52
9,523.54
10,409.08
5,964.11
10,263.97
5,880.97
16,144.94
4,812.82
2,757.61
4,387.16
2,513.72
6,900.88
4,687.57
2,686.99
6,349.54
3,638.11
9,987.65
10,045.09
5,755.55
8,147.97
4,668.56
12,816.53
8,798.11
5,041.07
8,138.98
4,663.41
12,802.39
45,756.58
26,218.37
42,804.64
25,371.29
68,175.93
This table summarizes the estimates of health outcomes expressed in terms of DALYs due to routine polio immunization by the six WHO regions of the world. Both discounted and undiscounted values are presented
higher in the Americas than in other regions of the world during the pre-vaccination period and second, the prevention of deaths and disability happened in the Americas much earlier than any other regions of the world. Since the discounted values are expressed in terms of 2007, all deaths and disabilities averted prior to 2007 are inflated by the discount factor. It should also be mentioned that in the calculations above we have used the number of symptomatic polio cases even though almost all children are likely to be infected with poliovirus
1367
1368
78
Economic Costs and Disability-Adjusted Life Years in Polio Eradication
if not immunized. Since the number of deaths has been estimated from the number of paralytic polio cases, the case fatality rate becomes relatively low, about 2% of all symptomatic polio cases.
4.2
Net Cost of Routine Polio Immunization and the Cost-Effectiveness Ratios
> Table 78-6 reports the costs of vaccination and medical care cost savings achieved through routine immunization due to the prevention of polio cases and polio related medical care costs in six regions of the world. All costs are expressed in terms of 2007 prices and 2007 values.
. Table 78‐6 Aggregate net cost of routine polio immunization and cost per DALY averted for all years over 1970–2050 expressed in terms of 2007 values
WHO regions
Cost of routine Medical care immunization costs averted (millions) (millions)
Net cost of routine polio immunization (millions)
Discounted DALYs averted (thousands)
Net cost per DALY averted
$8,729
$4,039
$4,690
9523.54
$44,959
$194,478
$149,519
16,144.94
$7,776
$6,691
$1,085
6,900.88
Europe
$30,412
$107,719
$77,307
9,987.65
South-east Asia
$13,038
$4,502
$8,536
12,816.53
Western Pacific
$18,677
$28,882
$10,205
12,802.39
Negative
$123,591
$346,311
$222,720
68,175.93
Negative
Africa Americas Eastern Mediterranean
Global
$492.5 Negative $157.2 Negative $666.0
In this table, net costs of routine polio immunization have been derived by WHO regions. Net cost is defined as the cost of routine polio immunization minus the medical care costs averted due to the prevention of polio cases. In the final column of the table, cost per DALY is shown. Note that the cost per DALY is negative for the whole world taken together and for three regions of the world
To convert past and future costs in terms of 2007 values (time preference adjustment), discount rate of 3% was used. The table also shows the cost-effectiveness ratios for the six regions as well as for the whole world using DALYs averted as the effectiveness measure. Estimating the cost of routine immunization is straightforward; we can simply multiply the cost per dose with the total number of polio doses administered. Number of doses administered is obtained by multiplying the number of infants (less than a year) in a region with vaccine coverage rate and then with four (four doses of vaccine). Calculation of medical care cost savings is much more involved. Paralysis cases need medical care over the life time and not all cases seek medical attention. Given the assumption on medical care utilization rates for paralysis cases by country-categories (see > Table 78‐3), number of cases which would
Economic Costs and Disability-Adjusted Life Years in Polio Eradication
78
otherwise seek medical care in any year can be derived by obtaining the cumulative number of averted paralysis cases over the previous 67 years. The aggregate global economic cost of routine immunization over the years 1970–2050 would be about $123.6 billion in terms of 2007 values while the medical care cost averted on the global scale will be about $346 billion. If we consider a very long-term perspective, continuation of routine polio immunization is actually cost saving for three regions of the world: Americas, Europe and Western Pacific. Therefore, if risk of polio exists, it will be irrational for these regional countries not to continue with polio immunization. Even for poorer regions of the world, Africa, Eastern Mediterranean and South-east Asia, net cost per DALY averted are relatively low. The cost-effectiveness ratio for South-east Asia was $666 per DALY averted, the highest cost-effectiveness ratio among all the six regions of the world. Therefore, even if eradication activities were not in place (or were not feasible), it makes perfect economic sense to continue with the routine immunization in the long run. Medical care cost savings on the global scale will be more than enough to pay for the immunization cost. For the whole world taken together, the net cost of polio immunization becomes negative. Therefore, polio immunization remains a highly desirable intervention if we consider the whole world as one unit. In South-East Asia, the intervention appears least cost-effective among all the six regions. One of the reasons for this relatively high cost-effectiveness ratio is the lower medical care cost associated with polio cases and lower probability of seeking care for polio. Over time, health-seeking behavior of population will change and may make the polio immunization relatively more cost-effective.
4.3
Cost-Effectiveness of Polio Eradication Activities
To calculate the cost-effectiveness of polio eradication, we assume that the eradication campaign conducted in different regions of the world can successfully eradicate wild polio transmission and vaccination will not be needed to protect population from wild poliovirus. In this sense, the benefit of eradication activities is the cost-savings achieved by discontinuation of polio vaccination as well as prevention of few hard-to-protect cases through supplemental vaccination activities. > Table 78‐7 reports the cost-effectiveness polio eradication activities by the WHO regions. The method of calculation of the costs and effectiveness is similar to the approach followed for routine program but successful eradication has one additional component, the cost savings due to discontinuation of routine immunization. It is assumed that countries will not stop polio immunization in all areas and 50% of routine program will continue after eradication certification. The year of polio eradication certification is assumed to be 2015. > Table 78‐7 indicates that polio eradication activities will be cost savings for the whole world taken together, even if 50% of regular polio immunization continues after the certification date. However, the estimates show wide regional variability. For example, in South-east Asia, polio eradication activities show very high cost-effectiveness ratio, almost $8,000 per DALY averted, if polio immunization cannot be stopped altogether after certification. If routine polio immunization is completely discontinued after 2015, the net cost becomes negative for all the regions of the world. Therefore, polio eradication activities will be cost-saving in the long-run if complete cessation of polio immunization is feasible after
1369
1370
78
Economic Costs and Disability-Adjusted Life Years in Polio Eradication
. Table 78-7 Cost-effectiveness ratios for polio eradication activities expressed in 2007 values (assumed date of eradication 2015)
WHO regions Africa Americas Eastern Mediterranean
Eradication cost (millions)
Vaccine cost savings (millions)
Medical care cost savings (millions)
Net cost of DALYs averted eradication by eradication (millions) (thousands)
Net cost/ DALY
$1,826
$1,572
$8
$246
$73
$5,169
$1,002
$6,098
70.11
Negative
$1,227
$1,221
$86
$80
51.98
Negative
112.6
$2185
Europe
$2,195
$2,640
$418
$863
44.66
South-east Asia
$2,887
$1,952
$37
$898
112.64
$7,972
Western Pacific
$5,766
$5,130
$198
$438
97.71
$4,482
$13,974
$17,684
$1,749
$5,459
489.70
Global
Negative
Negative
This table shows the cost-effectiveness ratios of eradication activities by WHO regions. Column 2 reports the costs associated with eradication activities. Polio eradication will potentially allow discontinuation of routine polio vaccination and vaccine costs will be saved in the post-eradication years. This savings are shown in column 3. Column 4 shows the estimates of medical care cost savings for few hard-to-reach cases, which can be reached and prevented by eradication activities. Net cost of eradication is shown in column 5 (costs minus the savings). Net costs per DALY averted by polio eradication activities are reported in the final column
certification. Without complete cessation of polio immunization, the eradication cost incurred may be better used elsewhere to improve the health status of population in a more effective manner.
5
Conclusions and Policy Implications
Polio eradication program has shown significant progress towards polio free world since 1985. At this point in time, it is important to examine the costs and effectiveness of polio vaccination and eradication programs. This will be helpful to policy makers to design and advocate future eradication activities for other infectious diseases. In this study, estimates of costs and health effects of polio vaccination and eradication are presented by considering a very long time frame, from 1970 to 2050. Economic evaluation of control and eradication of infectious diseases should consider long time horizon as eradication activities continue to provide health benefits long after the complete cessation of eradication and control activities. In this paper, health outcomes are measured in terms of deaths averted, years of life saved due to the prevention of premature death and years of disability averted due to the prevention of paralysis. Cost parameters were also derived by country-categories defined by economic
Economic Costs and Disability-Adjusted Life Years in Polio Eradication
78
status of countries. In discussing policy alternatives at the global level, economic status based country categorization provides more realistic policy directions. The analysis of the paper indicates a number of important results and policy implications.
Summary Points Routine polio immunization will prevent more than 42 million polio cases over the years
1970–2050. Total number of paralysis and deaths averted through the program should be about four million and 855 thousand respectively. Polio vaccination strategy is clearly one of the most effective public health interventions the world has ever seen. The number of polio cases and polio endemic countries has declined very rapidly over the last 20 years due to the successful implementation of polio vaccination and eradication strategies. Although the global cost of polio immunization is quite high, more than $123 billion (in 2007 values) for the period 1970–2050, the benefits and health outcomes are also high. In fact, at the global scale, routine polio immunization is cost-saving. The WHO defines an intervention as highly cost-effective if the cost per DALY averted is less than the Gross National Income (GNI) per capita of the country or the region. The cost-effectiveness ratios of routine polio immunization suggest that the program will be highly cost-effective in all countries in Africa with GNI per capita of $492 or more. The average GNI per capita of African countries was about $850 in 2006 but 22 countries in Africa had GNI per capita below $492 in 2006. For these very poor countries, international financial assistance will be needed to ensure continued and effective implementation of polio immunization and eradication activities. For all countries in the Eastern Mediterranean, routine polio immunization is a > highly cost-effective intervention. Cost per DALY averted becomes only $157 for this region and all countries in the region had GNI per capita of more than $157. Like Africa, the cost-effectiveness ratio of routine polio immunization in South-east Asia was also relatively high, about $666 per DALY averted. Again, some of the countries in this region will require international assistance in implementing the polio immunization program. However, the relatively high cost-effectiveness ratio in the region is due to low utilization of medical care services by polio and paralysis cases. With rapid economic progress in India, utilization rate of medical care services is likely to expand, lowering the net cost per DALY averted for the whole region. Polio eradication has remained elusive and the health outcomes of eradication activities, over and above the routine polio immunization, are relatively low. Despite the low health outcomes of eradication, eradication activities were cost-saving in three regions of the world, namely, the Americas, Eastern Mediterranean and Europe. If countries stop routine polio immunization completely after eradication, the interventions will be cost savings for all regions of the world. If polio eradication cannot be achieved within the next 5 years or so, emphasis should shift towards strengthening routine polio immunization as the program remains highly costeffective with or without eradication. From the developed countries’ point of view, it is economically justifiable to help the poor countries to adopt and continue with effective polio immunization programs. Polio control provides such a high level of cost savings for the high-income countries, less
1371
1372
78
Economic Costs and Disability-Adjusted Life Years in Polio Eradication
than half of the cost savings can fully fund polio vaccination in the developing world. Allocating only 10% of the net savings of high-income countries will help to fund 20% of polio vaccination costs in the developing world. High-income countries, for protecting their own net savings generated through the eradication of polio, should become actively involved in the strengthening the global immunization program.
References Bart KJ, Foulds J, Patriarca P. (1996). Bull World Health Organ. 74(1): 35–45. Brenzel L, Claquin P. (1994). Soc Sci Med. 39(4): 527–536. CDC. (1998). JAMA. 279(2): 103–104. Cockburn WC, Drozdov SG. (1970). Bull World Health Organ. 42: 405–417. Fairbrother G, DuMont KA. (1995). Am J Public Health. 85(12): 1662–1665. Jian Z, Jing-jin Y, Rong-zhen Z, et al. (1998). Int J Health Plan Manage. 13: 5–25. Kaddar M, Levin A, Dougherty L, Maceira D. (2000). Costs and Financing of Immunization Programs: Findings of Four Case Studies. Partnerships for Health Reform Project, Abt Associates Inc., May. Kaddar M, Mookherji S, DeRoeck D, Antona D. (1999). Case Study on the Costs and Financing of Immunization Services in Morocco. Special Initiatives Report number 18, Partnerships for Health Reform Project, Abt Associates Inc., September. Khan M, Yoder R. (1998). Expanded Program on Immunization in Bangladesh: Cost, Cost-effectiveness, and Financing Estimates. Technical Report number 24, Parnerships for Health Reform Project, Abt Associates Inc., September. Levin A, Ram S, Afsar A, et al. (2000a). The Costeffectiveness of Mixes of Operational Approaches to Polio Eradication: Findings of Two Case Studies. Special Initiatives Report Number 32, Partnerships for Health Reform Project, Abt Associates Inc., October. Levin A, Ram S, Kaddar M. (2000b). The Impact of the Polio Eradication Campaign on the Financing of Routine EPI: Finding of three Case Studies. Special Initiative Report number 27, Partnerships for Health Reform Project, Abt Associates Inc., March.
Liu X, Levin A, Makinen M, Day J. (2002). OPV vs IPV: Past and Future Choice of Vaccine in the Global Polio Eradication Program. Partnerships for Health Reformplus Project, February. Mathers C, Loncar D. (2006). Updated projections of global mortality and burden of disease, 2002–2030: Data sources, methods and results. Evidence and information for policy working paper, World Health Organization, Geneva, November. Murray CJL, Lopez AD, (eds.). (1996). The Global Burden of Disease: A Comprehensive Assessment of Mortality and Disability from Diseases, Injuries and Risk Factors in 1990 and Projected to 2020. Harvard University Press, Cambridge. Musgrove P. (1989). Is Polio Eradication in the Americas Economically Justified? in Health Economics: Latin American Perspectives, Pan American Health Organization, Scientific Publication no. 517. World Bank. (2006). World Development Report 2007. Washington DC. World Health Organization. (2000). Global Polio Eradication Initiative: Strategic Plan 2001–2005. Geneva. World Health Organization. (2002). Endgame issues for the global polio eradication initiative. Technical Consultative Group to Global Eradication of Polio, Clinical Infectious Diseases. 34: 1 January. World Health Organization. (2006). Global Program on Evidence for Health Policy (GPE), “Global burden of Poliomyelitis in the year 2000”, Global Burden of Disease 2000 Notes, Draft 15-08-06. World Health Organization. (2008). Conclusions and Recommendations of the Advisory Committee on Poliomyelitis Eradication, Geneva, 27–28 November 2007. Weekly Epidemiological Record. Year 83, number 3, 18 January, pp. 25–35.
79 Financial Burdens and Disability-Adjusted Life Years in Echinococcosis P. R. Torgerson 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1374
2
Human Clinical Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1376
3
Economic Effects of Echinococcosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1378
4 4.1 4.2 4.3
Calculating the DALY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1381 Disability Weight . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1381 Years Lived with Disability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1381 Survival Analysis and AE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1383
5
Modeling Uncertainty . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1384
6
The Global Burden of Echinococcosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1386
7
Concluding Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1387 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1387 Key Facts About Echinococcosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1388
#
Springer Science+Business Media LLC 2010 (USA)
1374
79
Financial Burdens and Disability-Adjusted Life Years in Echinococcosis
Abstract: Human echinococcosis is a severe zoonotic parasitic disease which can be fatal if left untreated. Two forms of the disease exist, namely > cystic echinococcosis (CE) caused by Echinococcus granulosus and > alveolar echinococcosis (AE) caused by E. multilocularis have a widespread distribution. Both forms of the disease are characterized by slow-growing, spaceoccupying lesions which most commonly occur in the liver with clinical features similar to those of a tumor. The disease results in financial burdens due to the direct costs of treatment and indirect costs due to morbidity effects such as decreased employment. Although this review primarily targets human disease, CE also has significant animal health costs which can add to the overall burden of disease in society. Most of the burden of disease occurs in lower income countries. A number of studies have demonstrated that both the financial costs of the disease and the burden of disease in terms of DALYs are substantial in endemic areas and in a few communities, such as pastoralists in Tibet, echinococcosis can be amongst the highest contribution to total burden of disease. On a global scale the disease burden may rival that of many better known diseases such as African trypanosomiaisis and onchocerciasis. CE is also largely preventable and often highly cost effective to do so. Despite this, echinococcosis remains a neglected disease. List of Abbreviations: AE, alveolar echinococcosis; CE, cystic echinococcosis; DALY, disability adjusted life year; HCC, hepatocellular carcinoma; OIE, World Organization for Animal Health (Office International des Epizooties); PAIR, puncture-aspiration-injection-respiration; YLD, years of life lived with disability (a component of the DALY); YLL, years of life lost (a component of the DALY)
1
Introduction
Human echinococcosis is caused by infection of the > larval stage of a number of tapeworm species (Eckert and Deplazes, 2004). Cystic echinococcosis (CE) caused by the larval stage of Echinococcus granulosus is by far the most important on a global scale. Some pathogenic genotypes are now being recognized as separate species such as E. ortleppi (McManus and Thompson, 2003). Alveolar echinococcosis (AE) is caused by the larval stage of E. multilocularis can be important locally. Human infections with E. vogeli and E. oligathus results in polycystic echinococcosis. Relatively few cases have been described and they were all in Latin America. CE is a condition of livestock and humans, arising from the ingestion of infective eggs of the cestode Echinococcus granulosus. Dogs are the primary > definitive hosts for this parasite, with livestock acting as > intermediate hosts and humans as aberrant intermediate hosts (> Figure 79-1). The outcome of infection in livestock and humans is cyst development in the liver, lungs, or other organ system. The distribution of E. granulosus is considered worldwide, with only a few areas such as Iceland, Ireland, and Greenland believed to be free of autochthonous human CE (Torgerson and Budke, 2003). However, CE is not evenly distributed geographically (> Figure 79-2). The United States has very few cases with most human cases being in immigrants. The same is true for several countries of Western and Central Europe, although in some countries such as Italy and Spain there is still significant transmission to humans. In many parts of the world, however, CE is considered an > emerging disease. For example, in the former Soviet Union and Eastern Europe, there have been dramatic increases in the number of observed cases in recent years (Torgerson et al., 2002, 2003, 2006). Additionally,
Financial Burdens and Disability-Adjusted Life Years in Echinococcosis
79
. Figure 79-1 Life cycle of Echinococcus granulosus. This demonstrates how E. granulosus is transmitted. It is a cestode parasite which naturally transmits between dogs (the small, none pathogenic adult is in the intestine) and farm animals (where larval cysts grow in the liver and other organs). Dogs are infected through eating offal containing the larval cysts (or hydatid cysts). Humans and livestock are infected through the ingestion eggs passed in the faces of dogs. Image supplied by the Institute of Parasitology, Zurich
in other regions of the world, such as parts of China, the geographical distribution and extent of CE is greater than previously believed (Chai, 1995). There are also substantial numbers of cases in much of Latin America and the parasite is endemic to Australia. The parasite is widespread in the Middle East and much of sub-Saharan Africa. CE not only causes severe disease and possible death in human patients, but also results in economic losses due to treatment costs, lost wages, and livestock associated production losses (Torgerson, 2003). Echinococcus multilocularis is found throughout the northern hemisphere (> Figure 79-3). The larval stage of this parasite causes AE (Eckert and Deplazes, 2004). Naturally the parasite cycles between the definitive hosts foxes and small rodents. Dogs are also suitable definitive hosts. Humans can be aberrant hosts (> Figure 79-4) and human infection arises
1375
1376
79
Financial Burdens and Disability-Adjusted Life Years in Echinococcosis
. Figure 79-2 Global distribution of zoonotic forms of Echinococcus granulosus. This shows the world-wide distribution of Echinococcus granulosus with very few areas being free of the parasite. Image supplied by the Institute of Parasitology, Zurich
from ingestion of eggs. Human AE consists of a space occupying lesion and the primary lesion is nearly always found in the liver. The metacestode is slow growing but locally is highly invasive. In late stage cases the metacestode will metastasize to distant organs and, untreated, the disease has a 90% or greater fatality rate (Ammann and Eckert, 1996). In Europe the disease has a widespread distribution in central and Eastern Europe. AE can be found throughout much of Turkey and the disease is endemic in much of the former Soviet Union. In western China in certain communities very high human prevalences are seen (Budke et al., 2004). The disease also has a sporadic occurrence in Hokkaido Japan (Kimura et al., 1999). E. multilocularis is widespread in North America although human AE cases are rare and have been largely confined, for example, to particular communities in Alaska (Rausch et al., 1990; Stehr-Green et al., 1988).
2
Human Clinical Disease
The initial phase of CE is asymptomatic with small, well-encapsulated cysts. After an undefined period of several months to years, the infection may become symptomatic as a spaceoccupying lesion. However, 60% of infections will remain asymptomatic (Eckert and Deplazes, 2004; Kern, 2003). The liver is the most common organ involved, usually with over two thirds of cysts (> Figure 79-5). The lungs are infected in about 20% of cases, with other organ involvement accounting for less than 10% of cases (> Figure 79-6). The treatment options for CE include surgical removal of the lesions and in many parts of the world CE is the most common reason for abdominal surgery. Surgery has a success
Financial Burdens and Disability-Adjusted Life Years in Echinococcosis
79
. Figure 79-3 Life cycle of Echinococcus multilocularis. This demonstrates how E. multilocularis is transmitted. It is a cestode parasite which naturally transmits between dogs or foxes (the small, none pathogenic adult is in the intestine) and small rodents (where larval cysts grow in the liver and other organs). Dogs or foxes are infected through predating. Image supplied by the Institute of Parasitology, Zurich
rate of up to 90% (Eckert and Deplazes, 2004; Kern, 2003). An alternative to surgery is the PAIR technique, (WHO, 1996). Chemotherapy, using benzimidazoles, has also been used successfully (Kern, 2003). In calcified cysts, there is an indication for a wait and see approach to treatment. AE, due to the metacestode stage of E. multilocularis, is an often-fatal condition if untreated. The cyst is multivesicular and highly infiltrative locally. The primary site of metacestode development is almost exclusively the liver. Secondary metastasis may form in a variety of adjacent or distant organs in longer standing cases, making surgical management difficult. Patients present with cholestatic jaundice and/or epigastric pain, fatigue, weight loss, hepatomegally or abnormal routine laboratory findings (Ammann and Eckert, 1996; Eckert and Deplazes, 2004). Treatment options include partial and radical surgical resection for localized lesions in combination with long-term chemotherapy using benzimidazoles. Recent studies have suggested that modern treatment methods have substantially improved the
1377
1378
79
Financial Burdens and Disability-Adjusted Life Years in Echinococcosis
. Figure 79-4 Global distribution of Echinococcus multilocularis. This shows that Echinococcus multilocularis is confined to the northern hemisphere. This is significant concern in Europe that the parasite is on the increase as the fox population has increased substantially in the last 20 years, possibly as a result of the successful rabies vaccination program. Europe is now seeing increases in the numbers of cases of human AE. Image supplied by the Institute of Parasitology, Zurich
prognosis for this disease (Ammann and Eckert, 1996; Bresson-Hadni et al., 2000; Torgerson et al., 2008). Human infections with E. vogeli and E. oligathus results in polycystic echinococcosis. Relatively few cases have been described and they were all in Latin America. In 80% of cases the lesions involved the liver; the rest were located in the lung or single organ sites (D’Alessandro, 1997). The most common clinical presentation includes liver masses, enlarged abdomen, abdominal pain, weight loss and fever. In about 25% of cases there are signs of biliary hypertension and biliary obstruction. From the limited numbers of cases that have been reported the fatality rate is at least 26%.
3
Economic Effects of Echinococcosis
In the human population it can be seen that the burden of disease will be due to the cost of sickness and disability due to the disease and the cost of treatment. > Direct financial costs can be estimated from the hospital costs of treatment. Indirect financial costs will include the loss of income due to disability and other long term convalescent costs. Lost income due to premature mortality is also possible (Torgerson, 2003). Early studies on disease burden for echinococcosis were mainly targeted at financial losses due to the disease (Majorowski et al., 2005; Torgerson and Dowling, 2001; Torgerson et al., 2000, 2001). This has some advantages as the disease also has an economic effect on livestock. For humans, the financial costs
Financial Burdens and Disability-Adjusted Life Years in Echinococcosis
79
. Figure 79-5 MRI images of cystic echinococcosis. Hydatid cysts can be clearly seen as space occupying lesions in the liver. Images from (Aliev et al., 2004) with permission
of echinococcosis not only depend on the incidence of the disease but also on the cost of treatment. The cost of treatment can have huge variations depending on where treatment is undertaken. In Jordan for example, a lower middle income country the cost of treating a case of CE is on average approximately $524 (Torgerson et al., 2001). This compares to the cost of treating a case of CE in the UK of approximately $15,000 (Torgerson and Dowling, 2001). In Switzerland the direct costs of treating a case of AE is on average $150,000 (Torgerson et al., 2008) (based on the approximate exchange rates of April 2008). Other indirect costs have similar huge variations depending on the relative economic output of the country where the disease occurs. Later studies, including the global burden of CE (Budke et al., 2006) have also incorporated DALY estimates for disease burden. An important part of the societal burden of disease is that of animal health costs. CE is a zoonosis and many species of agricultural animals, but especially sheep are afflicted by CE. In animals CE will result in loss of edible liver and can also result in loss of productivity such as lowered milk production, loss of food conversion or impaired fertility (Torgerson, 2003). In terms of financial costs on a global scale loss of animal productivity is perhaps half of the total burden of disease. The total global burden of animal health losses is believed to be at least
1379
1380
79
Financial Burdens and Disability-Adjusted Life Years in Echinococcosis
. Figure 79-6 CAT scan of cerebral echinococcosis. Although CNS involvement of echinococcosis is rare, when it does occur it can be a devastating condition. Image supplied by the Government Paediatric Hospital, Bishkek, Kyrgyzstan
$141 million and possibly as high as $2.1 billion (Budke et al., 2006). In some communities the loss of animal productivity may be the major financial burden on society (Torgerson et al., 2001). The financial losses caused to the livestock industry is a strong argument for calculating disease burden in financial terms (Carabin et al., 2005) particularly as the disease often occurs in low income countries where livestock production may make up a substantial part of rural income. In addition, including livestock costs in the disease burden estimates can also substantially improve the potential cost effectiveness of disease control, especially if cost sharing between agricultural and health sectors is undertaken. Indeed at least one study has suggested that the financial return from disease control may be more than the cost of implementing the program (Budke et al., 2005). Also of importance to consider is the financial burden of controlling the disease. This can be considerable in terms of actual expenditure. This includes the meat inspection costs and the routine treating of dogs with anthelmintics for example. However, these costs must also be put in the context of the costs of the disease prevented. The costs of control will also be dependent on the government policy and stage of the control program. In New Zealand the financial costs of control of CE were substantive, but the program was successful in eliminating the parasite from livestock and preventing transmission. The present day costs are now relatively small and consist of surveillance to maintain the parasite free status and the costs of treating a small number of human cases who may have been infected some time ago when there was active transmission. A review of the New Zealand and other programs can be found in (Craig and Larrieu, 2006). Likewise there is a potential cost of controlling AE by periodically distributing baits containing anthelmintics to reduce E. multilocularis in the fox population. Preliminary programs, for example in Zurich in Switzerland (Hegglin et al., 2003) have proved to be successful, but the long term effects of reducing transmission to humans still remains unknown.
Financial Burdens and Disability-Adjusted Life Years in Echinococcosis
4
Calculating the DALY
4.1
Disability Weight
79
Central to calculating a DALY for echinococcosis is to assign a disability weight to the condition. The majority of cases of both AE and CE are a primary space occupying disease of the liver. Therefore the disability weights, derived for AE and CE, are based on values for liver cancer obtained from the original Global Burden of Disease Study as well as from the Dutch Disability Weight (DDW) Group, which produced a set of disability weights for use in a Western European context (Stouthard et al., 2000). Liver cancer was chosen it often results in similar clinical symptoms (> Tables 79-1 and > 79-2). One of the main differences between the conditions is that echinococcosis has a more chronic disease course. This is, however, accounted for by having echinococcosis patients remain at their assigned disability weight for a prolonged period of time. In burden studies on echinococcosis the disability weight assigned has depended on the stage of the disease. Thus a disability weight of 0.2 (Dutch weight for clinically disease free cancer) for patients was assigned to patients with regressed or calcified lesions or those with mild disease. A weight of 0.239 was assigned for patients with stable lesions. This is the Global Burden of Disease weight for pre-terminal liver cancer – (Murray, 1994). A weight of 0.809 was assigned for patients with severe and deteriorating disease (the Global Burden of Disease weight for terminal liver cancer). Disability weights have been assumed to be age independent because there is no evidence to suggest disparity in clinical presentation dependent on age of onset. Standard age weighting and correction parameters together with discount rates can be used. The age structure of the population can be calculated from available data. For example in Switzerland the mean age of onset of AE patients is 54 years.
4.2
Years Lived with Disability
This is dependent on the progression of the disease and is the time from the onset of morbidity to cure or death. To calculate this modeling techniques and assumptions have been made based on the natural history of the disease. For CE this varied between 5 years for mild lesions to their expected lifespan for more severe lesions. The latter is justifiable even when there is surgical cure of the disease because quality of life studies have suggested that patients never fully recover from the disease, have a lower quality of life even if they are un aware they have the disease and suffer substantially higher levels of unemployment (Budke et al., 2004; Torgerson and Dowling, 2001; Torgerson et al., 2003). For AE, if treatment is available, then chemotherapy may be required for long periods of time. Thus in Switzerland, for example therapy is required for an average of 7 years to affect a cure and this would be the minimal time lived with the disability (Torgerson et al., 2008). Without treatment modeling has suggested that life expectancy of AE is between 5 and 10 years and hence in this situation it would require a weight of 0.239 for the early part of the disease progressing to 0.809 for the terminal stage (Torgerson et al., 2008). Estimated life expectancy of an individual has been based on the Japanese estimated life span in order to standardize DALYs lost in accordance with the Global Burden of Disease Study (Murray, 1994). A model > life-table, West Level 26, has been used to estimate expected longevity for each age. In the Tibetan study (Budke et al., 2004) a Chinese life-table
1381
43
25
21
7
AE (n = 30)
AE (n = 76)
AE (n = 33)
CE (n = 59)
5
76
14
23
83.9
Hepatomegaly (%)
42
60
25
20
56
Mass-related pain (%)
9
9
7
3
3.2
Lung involvement (%)
36
–
14
7
2.1
Asymptomatica (%)
Schaefer and Khan (1991)
Wilson and Rausch (1980)
Vuitton et al. (1996)
Vuitton et al. (1996)
Sithinamsuwan et al. (2000)
Reference
This table compares the symptoms of HCC with that of the two common forms of echinococcosis. From this it can be seen that echinococcosis can present with very similar clinical problems as does HCC a These cases were found incidentally in patients without clinical signs (diagnosed by chance at necropsy, laparotomy, or during ultrasound examination for other reasons such as pregnancy), other categories were diagnosed clinically and confirmed radiologically
42.6
Jaundice (%)
HCC (n = 336)
Presenting clinical signs
79
. Table 79-1 Comparison of the presenting clinical signs of alveolar echinococcosis (AE) and cystic echinococcosis (CE) of the liver with hepatocellular carcinoma (HCC)
1382 Financial Burdens and Disability-Adjusted Life Years in Echinococcosis
Financial Burdens and Disability-Adjusted Life Years in Echinococcosis
79
. Table 79-2 Other presenting clinical signs of hepatocellular carcinoma (HCC) with comparison to alveolar echinococcosis (AE) and cystic echinococcosis (CE) Clinical signs of HCC (Sithinamsuwan et al., 2000)
AE
CE
GI bleeding
√
√
Edema
√
√
Abdominal enlargement
√
√
Encephalopathy
√
√
Liver abscess-like symptoms
√
√
Non-specific symptoms (anorexia, nausea, vomiting, malaise, wt. loss)
√
√
Fever
√
√
Anemia
√
√
Ascites
√
√
Cachexia
√
√
Chronic liver stigmata
√
√
Splenomegaly
√
√
This table also illustrates why it is possible to use the symptoms of HCC to derive a disability weight for echinococcosis. (√) indicates the clinical sign has been associated with this condition
(Lopez et al., 2000) was also utilized for comparison. The general DALY formula was employed in the construction of DALYs specific for alveolar echinococcosis (AE) and cystic echinococcosis (CE). In the Tibetan study, DALYs were constructed on the premise of solely chemotherapeutic therapy, as this is the most common treatment modality for the region and in nearly all cases the only treatment currently available.
4.3
Survival Analysis and AE
A recent manuscript reported the DALY due to AE in Switzerland. In this study a large data set of 329 cases of AE spanning the last 40 years was examined. Relative > survival analysis was undertaken and it demonstrated a substantially improved prognosis for AE for patients being treated now compared to 40 years ago (Torgerson et al., 2008). This is due to the development of new treatment methods. Compared to the normal Swiss population the study demonstrated that for an average 54-year-old patients diagnosed with AE in 1970 the life expectancy was estimated to be reduced by 18.2 and 21.3 years for men and women, respectively. By 2005 life expectancy of patients with AE was reduced by approximately 3.5 and 2.6 years, respectively (> Figure 79-7). This reduced life expectancy can be used as the basis for calculating the YLL. The YLD was also calculated from examining the hospital records to determine the mean time that affected individuals were under treatment. YLDs were calculated according to standard methods. Standard age weightings were used and a discount rate of 3% applied. For the scenario that no treatment would be available, the YLLs were estimated from the survival curve from 1970 where the survival analysis indicated a considerably shortened life expectancy. This will be a means of giving better estimates of YLLs in endemic areas where
1383
1384
79
Financial Burdens and Disability-Adjusted Life Years in Echinococcosis
. Figure 79-7 Comparative life expectancy at age 54 years for AE patients (--- males, females) compared to Swiss population norms (. . .males, ----females). This figure demonstrates that the life expectancy of AE patients has improved considerably in Switzerland in the past 30 years. This is probably due to better surgical techniques but more importantly the use of benzimadazole chemotherapy introduced in the mid 1970s. However in remote rural communities where poverty is widespread such as in Tibet such treatment is not available and the prognosis remains bleak
such treatment options are not available. An example would be remote rural areas of the Tibetan plateau where prevalences of 5% have been recorded. In this Swiss study it was also possible to estimate the cost effectiveness of treatment. To do this present incidence rates were used but the life expectancy of AE in the 1970s used. In this scenario, YLDs were also calculated based on the last 2–3 years of life having a disability weight of pre-terminal liver cancer, and the rest of life living with the disease a weight of terminal liver cancer (0.239 and 0.809) respectively. This indicated that, although treatment costs per patient were of the order of $150,000, because the treatment prolongs life expectancy substantially, the cost per DALYs saved is $9,000. If this is compared to the mean annual GDP per head of Switzerland of approximately $32,000 it is easy to argue that treatment, although expensive in Switzerland, is nevertheless cost effective.
5
Modeling Uncertainty
> Modeling
uncertainty for economic effects of echinococcosis were used initially in studies in Tunisia and Jordan (Majorowski et al., 2005; Torgerson et al., 2001). This has many advantages, most notably surveillance data from which the costs of the disease can be calculated are not deterministic and depend, interalia upon sampling error, errors due to diagnosis and problems with bias and underreporting. For financial costs there will be uncertainty in the mean cost of treatment due to large variations in the treatment costs of individual patients. In developing the DALY there is also uncertainty in terms of morbidity and mortality. In the Tibetan study (Budke et al., 2004), the DALY for AE was developed, with disability outcomes
Financial Burdens and Disability-Adjusted Life Years in Echinococcosis
79
divided into five components (cured, improved, stable, worse, or death) based on the health survey as well as findings from past studies where albendazole was utilized as the sole treatment for human AE (> Table 79-3). To model uncertainty, Monte-Carlo techniques were employed. From published data (> Table 79-3), the results of chemotherapeutic treatment of 103 AE . Table 79-3 Outcomes due to treatment of AE with benzimadazoles. Data used in stochastic analysis of (Budke et al., 2004) Number in study
Cured
Improved
Stable
Worse
Death
5
0
11
2 (18%)
35 37 15
Reference
1 (20%)
2 (40%)
1 (20%)
1 (20%)
Wen et al. (1994)
0
5 (46%)
3 (27%)
1 (9%)
Liang et al. (1997)
2 (6%)
4 (11%)
25 (72%)
4 (11%)
0
Horton (1989)
0
11 (30%)
10 (27%)
12 (32%)
4 (11%)
Ammann et al. (1994)
1 (7%)
12 (80%)
0
2 (13%)
0
Liu et al. (1991)
This table reviews the clinical outcome of patients with AE treated with benzimadazoles. Benzimadazoles are drugs that can be used to kill helminth parasites. In the case of echinococcosis they can arrest the development of the cyst and are useful as an alternative to surgical treatment
patients were used to construct a multinomial distribution for the likely outcome of treatment. Of these 103 subjects, there was an approximate probability of 4% of cure resulting from calcification and regression of the lesions. Patients in this category were assigned a disability weight of 0.200 for 5 years. A probability of approximately 27% was given for having mild disease with disability weight 0.200, a probability of approximately 41% was given for having disease equated to a disability weight of 0.239, and a probability of approximately 22% was given for severe disease equating to a disability weight 0.809. Patients assigned to these three later disease states were provided with a disability weight until the end of their expected lifespan, based on a trinomial distribution. In addition, approximately 6% of patients were assigned the outcome of eventual death, which equates to a disability weight of 0.809 for 10 years followed by death. Disability weights for cystic echinococcosis were assigned in a similar manner based on the results of albendazole treatment of 547 patients from past studies (> Table 79-4). A significant challenge in estimating the burden of echinococcosis is that of missing data. In wealthy countries there are often good data sets from public health statistics or from hospital records. Thus relatively accurate estimates of the incidence of disease can be made. However, most of the burden of echinococcosis is in developing nations with 40% being in China alone (Budke et al., 2006). In many areas even if treatment is available reporting systems are inaccurate and hence there is likely to be significant underestimates of the true numbers affected by the disease. This is illustrated by the estimated 75% underreporting in Chile and Uzbekistan of cystic echinococcosis (Nazirov et al., 2002; Serra et al., 1999). For communities where treatment is largely unavailable such as remote herdsmen communities on the Tibetan plateau most of the data must come from carefully planned surveys which will give an indication of point prevalences based on, for example, cross sectional ultrasound surveillance studies. Such data can be used to model the likely number of cases in society and the disease burden. A further problem is that even utilizing such data, ultrasound will not diagnose
1385
1386
79
Financial Burdens and Disability-Adjusted Life Years in Echinococcosis
. Table 79-4 Outcomes due to treatment of CE with benzimadazoles. Data used in stochastic analysis of (Budke et al., 2004) Number in study
Cured
Improved
Stable
Worse
Death
Reference
58
14 (24%)
29 (50%)
15 (26%)
0
0
253
72 (28%)
129 (51%)
46 (18%)
6 (3%)
0
Horton (1989)
59
50 (85%)
5 (8%)
1 (2%)
3 (5%)
0
Chai et al. (2002)
118
97 (82%)
6 (5%)
0
15 (13%)
0
Chai et al. (2002)
59
25 (42%)
25 (4%)
9 (16%)
0
0
Nahmias et al. (1991)
Wen et al. (1994)
Like > Table 3 this reviews the outcomes of patients with CE treated with benzimadazoles. This data was used to estimate the probability of patients responding to treatment in a particular way and hence calculate uncertainty estimates for the DALYs for echinococcosis
pulmonary echinococcosis and therefore this must be accounted for as this could be an additional 10% or more cases. Estimates of the prevalence of undiagnosed cases need to be undertaken as there is increasing evidence that ‘‘sub clinical’’ cases, i.e., those individuals who unknowingly have echinococcosis nevertheless suffer a significantly decreased quality of life and income reduction (Budke et al., 2004). Therefore such individuals will need to be given a disability weight to estimate the total DALYs for the society. The uncertainty of missing data can also be modeled using Monte-Carlo techniques. In the global burden disease study for CE it was assumed that approximately 10% of annual cases are not officially diagnosed and do not receive medical attention due to the socio-economic status of the patient or the seemingly subclinical nature of the illness (Budke et al., 2006). Based on past studies, this is a most likely a very conservative estimate (Budke et al., 2004; Torgerson et al., 2000, 2001). For example in China, mass ultrasound screening in remote areas has revealed high prevalences of CE (Budke et al., 2004), A number of these have advanced clinical disease but would not normally have access to treatment due to poverty and remoteness from medical facilities. It has also been widely acknowledged that human cases of CE are significantly and systematically underreported by the healthcare establishment even though they are recorded in the clinical records or patient admissions. Therefore, adjustments were made to account for the substantial underreporting of treated cases known to occur.
6
The Global Burden of Echinococcosis
A preliminary estimate of the global burden of cystic echinococcosis has now been made (Budke et al., 2006). Data on country-specific annual reported human CE cases were obtained from the OIE-HANDISTATUS 2 database for the years 1996–2003 and from literature searches. Type and quality of incidence data varied by country or region. However, the majority of data consisted of annual numbers of detected cases per susceptible population or was converted into this form for analysis purposes. If both an OIE-reported and a literature based value were available, the larger of the two was utilized. However, if the higher value appeared to be from a survey evaluating a highly endemic region and was, therefore, not applicable to the entire country, a corresponding adjustment was made. Adjustments
Financial Burdens and Disability-Adjusted Life Years in Echinococcosis
79
were also made to account for the substantial underreporting of treated cases known to occur. For financial costs a few studies had been made and modeling demonstrated that there was a significant linear correlation between the cost of treatment and the GDP per head of economic output. This was used as the basis for estimating the treatment costs in countries where such studies had not been made. In total the global burden of disease for CE was estimated at 285,782 DALYs, with financial burden of treatment and morbidity of close to $200 million. When underreporting is taken into account this rises to as much as 1 million DALYs or more and costs over $700 million. Locally the burden of echinococcosis can be substantial. For example, in Shiqu county, located on the Tibetan plateau of western China, there an extremely high prevalence of both AE and CE. The burden of disease is approximately 51,000 DALYs in a population of 63,000 or 0.81 DALY per person. This represents an extremely high burden of disease in this population and is amongst the highest burdens of any disease in this community (Budke et al., 2004). This burden consisted of approximately 18,000 DALYs for CE and 33,000 for AE with financial costs of at least $200,000 per annum (Budke et al., 2005). The global burden of alveolar echinococcosis has not been completed but in addition to Shiqu county (detailed above) the burden of AE in Switzerland has been estimated to be approximately 77 DALYs at a cost of at least $3 million per annum (Torgerson et al., 2008). The huge differences between Switzerland and Shiqu county are of course due to the much high incidence in Shiqu and the very limited treatment options. In Switzerland, although it is an emerging disease (Schweiger et al., 2007) with an increasing incidence, the incidence is still low and modern but expensive treatment methods, not available in Shiqu, have dramatically improved the life expectancy of AE and hence the YLL is quite small. However, the financial costs in Switzerland are higher even though the burden of DALYs is much lower. This is due to the very high treatment costs and high salaries enjoyed by Swiss residents and illustrates the complexities and problems of reliance on a purely financial approach to burden analysis.
7
Concluding Remarks
Echinococcosis is an emerging disease in many parts of the world and this is true of both AE and CE. It is a neglected zoonosis although the global burden of disease may be as much as that of diseases such as Dengue, Onchocerciasis, African Trypanosomiasis and Schistosomiasis (Budke et al., 2006). Certainly CE is preventable by instituting the appropriate control programs such as veterinary control of animal slaughter and regular periodic anthelmintic treatment of dogs. Mathematical modeling has suggested that it can be highly cost effective to control in terms of $ per DALY avoided. In addition there is a positive financial return due to decrease costs of treating cases and increase livestock productivity (Budke et al., 2005). Therefore control of this disease should be set as a priority. The disease has been eliminated in a few countries but still remains a challenge elsewhere.
Summary Points Echinococcosis is a global disease with economic effects in both humans and domestic animals.
The global burden of disease is high with most of the burden in lower income countries.
1387
1388
79
Financial Burdens and Disability-Adjusted Life Years in Echinococcosis
CE and AE are emerging diseases in many areas. Cystic echinococcosis is a preventable disease and is often highly cost effective, both financial terms and in terms of costs per DALY averted.
Disease burden estimates both in terms of DALYs and financial estimates will indicate the relative priority of echinococcosis control.
Financial estimates including animal health losses as well as DALY estimates means cost sharing between agricultural and health sectors can be implemented even though control is essentially reliant on veterinary services. This can improve apparent cost effectiveness.
Key Facts About Echinococcosis Echinococcosis is a parasitic disease caused by parasites of the genus Echinococcus. The disease is caused by the larval stage of the parasite. The adult stage is found in various
carnivore species. This is usually the domestic dog for Echinococcus granulosus (which causes cystic echinococcosis) and can be dogs or foxes for E. multilocularis (which causes alveolar echinococcosis). The adult stage is not pathogenic. The larval stage is a space occupying cyst which is found in the intermediate host. This is most commonly sheep or other farm animals (with Echinococcus granulosus) or small rodents (E. multilocularis). Dogs and foxes are infected by consumption of these cysts in, for example, discarded offal. Intermediate hosts are infected by consuming eggs which have been excreted by carnivores in their feces. When humans ingest eggs, for example through contaminated food or eggs transferred by fingers from dogs, a hydatid cyst develops. This occurs most commonly in the liver but can also occur in the lung. The disease is very pathogenic and without treatment can be fatal. Treatment usually consists of surgical removal of the cyst although chemotherapy is also possible. The disease occurs in communities where there is interaction between dogs and livestock or where there is substantial infection of the wild fox population. The parasite has a global distribution but is less common in wealthy countries. In some areas it is an important emerging disease. In Europe increasing numbers of cases of human alveolar echinococcosis are being seen and this is associated with an increase in the fox population. In central Asia there has been a substantial increase in cystic echinococcosis following the collapse of the Soviet Union.
References Aliev MA, Baimahanov BB, Axmetov EA. (2004). In: Torgerson PR, Shaikenov B (eds.) Echinococcosis in Central Asia: Problems and Solutions. Dauir, Almaty, Kazakhstan, pp. 176–184.
Ammann RW, Eckert J. (1996). Gastroenterol Clin North Am. 25: 655–689. Ammann RW, Ilitsch N, Marincek B, Freiburghaus AU. (1994). Hepatology. 19: 735–742.
Financial Burdens and Disability-Adjusted Life Years in Echinococcosis Bresson-Hadni S, Vuitton DA, Bartholomot B, Heyd B, Godart D, Meyer JP, Hrusovsky S, Becker MC, Mantion G, Lenys D, Miguet JP. (2000). Eur J Gastroenterol Hepatol. 12: 327–336. Budke CM, Deplazes P, Torgerson PR. (2006). Emerg Inf Dis. 12: 296–303. Budke CM, Jiamin Q, Qian W, Torgerson PR. (2005). Am J Trop Med Hyg. 73: 2–10. Budke CM, Jiamin Q, Zinsstag J, Qian W, Torgerson PR. (2004). Am J Trop Med Hyg. 71: 56–64. Carabin H, Budke CM, Cowan LD, Willingham AL, Torgerson PR. (2005). Trends Parasitol. 21: 327–333. Chai J, Menghebat JW, Sun D, Liang B, Shi J, Fu C, Li X, Mao Y, Wang X, Dolikun G, Wang Y, Gao F, Xiao S. (2002). Chin Med J. 115: 1809–1813. Chai JJ. (1995). Biomed Env Sci. 8: 122–136. Craig PS, Larrieu E. (2006). Adv Parasitol. 61: 443–508. D’Alessandro A. (1997). Acta Trop. 67: 43–65. Eckert J, Deplazes P. (2004). Clin Microbiol Rev. 17: 107–135. Hegglin D, Ward PI, Deplazes P. (2003). Emerg Inf Dis. 9: 1266–1272. Horton RJ. (1989). Trans R Soc Trop Med Hyg. 83: 97–102. Kern P. (2003). Lang Arch Surg. 388: 413–420. Kimura H, Furuya K, Kawase S, Sato C, Yamano K, Takahashi K, Uraguchi K, Ito T, Yagi K, Sato N. (1999). Jpn J Infect Dis. 52: 117–120. Liang M, Ito A, Liu YH, Wang XG, Yao YQ, Yu DG, Chen YT. (1997). Trans R Soc Trop Med Hyg. 91: 476–478. Liu YH, Wang XG, Chen YT. (1991). Chin Med J. 104: 930–933. Lopez AD, Salomon J, Ahmad O, Murray CJL, Mafat D. (2000). Life Tables for 191 Countries: Data, Methods, and Results. World Health Organization, Geneva. Majorowski MM, Carabin H, Kilani M, Bendsalah A. (2005). Trans R Soc Trop Med Hyg. 99: 268–278. McManus DP, Thompson RCA. (2003). Parasitology. 127: S37–S51. Murray CJL. (1994). Bull WHO. 72: 429–445. Nahmias J, Goldsmith R, Schantz P, Siman M, el-On J. (1991). Acta Trop. 50: 1–10. Nazirov FG, Ilkhamov IL, Ambekov NC. (2002). Med J Uzbekistan. 2/3: 2–5. Rausch RL, Wilson JF, Schantz PM. (1990). Ann Trop Med Parasitol. 84: 239–250.
79
Schaefer JW, Khan MY. (1991). Rev Inf Dis. 13: 243–247. Schweiger A, Ammann RW, Candinas D, Clavien P-A, Eckert J, Goostein B, Halkic N, Meuellhaupt B, Prinz BM, Reichen J, Tarr PE, Torgerson PR, Deplazes P. (2007). Emerg Inf Dis. 13: 878–882. Serra I, Garcia V, Pizzaro A, Luzoro A, Cavada G, Lopez J. (1999). Rev Med Chil. 127: 485–492. Sithinamsuwan P, Piratvisuth T, Tanomkiat W, Apakupakul N, Tongyoo S. (2000). World J Gastroenterol. 6: 339–343. Stehr-Green JK, Stehr-Green PA, Schantz PM, Wilson JF, Lanier A. (1988). Am J Trop Med Hyg. 38: 380–385. Stouthard MEA, Essink-Bot ML, Bonsel GJ. (2000). Eur J Pub Health. 10: 24–30. Torgerson PR. (2003). Acta Trop. 85: 113–118. Torgerson PR, Budke CM. (2003). Res Vet Sci. 74: 191–202. Torgerson PR, Carmona C, Bonifacino R. (2000). Ann Trop Med Parasito. 94: 703–713. Torgerson PR, Dowling PM. (2001). Ann Trop Med Parasitol. 95: 177–185. Torgerson PR, Dowling PM, Abo-Shehada MN. (2001). Ann Trop Med Parasitol. 95: 595–603. Torgerson PR, Karaeva RR, Corkeri N, Abdyjaparov TA, Kuttubaev OT, Shaikenov BS. (2003). Acta Trop. 85: 51–61. Torgerson PR, Oguljahan B, Muminov AE, Karaeva RR, Kuttubaev OT, Aminjanov M, Shaikenov B. (2006). Parasitol Int. 55: S207–S212. Torgerson PR, Schweiger A, Deplazes P, Pohar M, Reichen J, Ammann RW, Tarr PE, Halkik N, Mullhaupt B. (2008). J Hepatol 49: 72–77. Torgerson PR, Shaikenov BS, Baitursinov KK, Abdybekova AM. (2002). Trans R Society Trop Med Hyg. 96: 124–128. Vuitton DA, Bresson-Hadni S, Bartholomot B, Mantion G, Miguet JP. (1996). In: Uchino J, Sato N (eds.) Alveolar Echinococcosis. Strategy for Eradication of Alveolar Echinococcosis of the Liver. Fuji Shoin, Sapporo, pp. 243–251. Wen H, Zou PF, Yang WG, Jian L, Wang YH, Zhang JH, Roger R, New C, Craig PS. (1994). Trans R Society Trop Med Hyg. 88: 340–343. WHO. (1996). Bull WHO. 74: 231–242. Wilson JF, Rausch RL. (1980). Am J Trop Med Hyg. 29: 1340–1355.
1389
80 Measuring Japanese Encephalitis (JE) Disease Burden in Asia W. Liu . D. Ding . J. D. Clemens . N. T. Yen . V. Porpit . Z.‐Y. Xu 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1392 2 Methodological Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1394 3 Overall Burden of Disease Due to JE in China, Vietnam, Thailand and Indonesia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1395 4 DALYs Lost by Age and Contribution of YLL and YLD to the Lost DALYs . . . . . . . 1396 5 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1397 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1398
#
Springer Science+Business Media LLC 2010 (USA)
1392
80
Measuring Japanese Encephalitis (JE) Disease Burden in Asia
Abstract: Japanese > encephalitis is spread in virtually all Asian countries where pig rearing and rice cultivation are common agricultural practices. Disease burden due to JE was assessed in Shanghai, China (temperate region), Bali, Indonesia (tropical region), Northern Provinces of Thailand (subtropical region) and Northern Provinces of Vietnam (subtropical region) by a model of cumulative JE incidence for hypothetical cohorts in the four sites, each consisting of 100,000 newborn babies from birth through 15 years of age. The cumulative JE incidence rates were measured by age-specific, JE incidence rates averaged from an epidemic cycle during the pre-JE immunization eras. The model predicted 271 JE cases, 36 deaths, 81 disabilities, and 3,022 disability adjusted life years (> DALYs) lost associated with JE for the cohort in China; it predicted 110 JE cases, 11 deaths, 30 disabilities and 1,012 lost DALYs for the cohort of Vietnam. In Thailand, 192 cases, 33 deaths, 92 disabilities, and 3,161 lost DALYs were estimated. In Indonesia, 63 JE cases, 6 JE deaths, 23 disabilities would occur and 1,412 DALYs would be lost due to JE. On average, disability contributes about 58–70% to the DALYs lost in these four countries. Both years lived with disability (YLD) and years of life lost (YLL) were higher in the 0–9 year age group. The overall disease burden due to JE is high for all four countries. However, the burden was highest in Shanghai of the temperate region and lowest on the tropical island of Bali. List of Abbreviations: CSF, cerebrospinal fluid; DALYs, disability adjusted life years; IgM, immunoglubin M; JE, Japanese encephalitis; YLD, years lived with disability; YLL, years of life lost
1
Introduction
Japanese encephalitis (JE), a mosquito-borne viral infection, is the most important cause of viral encephalitis in Asia. Following the successful control of poliomyelitis, JE has become a major cause of neurological disability of children in Asia. Over 50,000 cases of JE were reported with 10,000–15,000 deaths annually (Burke and Leake, 1988; Endy and Nisalak, 2002; Halstead and Jacobson, 2003; Vaughn and Hoke, 1992), a likely underestimate given that JE is not a notifiable disease in most Asian countries. JE virus belongs to the family of Flaviviridae which includes West Nile, Murray Valley, St. Louis encephalitis, dengue fever and other viruses (Solomon et al., 2000, 2004). It is transmitted by mosquito vectors, most commonly by the species Culex tritaeniorhynchus which lays eggs in still water pools such as rice paddy fields or drainage ditches. Pigs are found to be the primary amplifying host of the virus, displaying very high levels of viremia. Hence in the vast rural regions of Asia where the farmers make a livelihood through rice cultivation and pig rearing, JE virus finds an optimal habitat for circulation between mosquito vector and mammalian hosts. Herons, ducks, as well as other wild and domestic birds have also been implicated in the transmission of JE virus (Burke and Leake, 1988; Solomon et al., 2000). Most JE infections in humans are asymptomatic (Chakraborty et al., 1981; Grossman et al., 1973; Halstead and Grosz, 1962). Seroprevalence studies indicate an annual risk of infection in 10% of children and nearly universal exposure by adulthood in JE endemic regions (Halstead and Jacobson, 2003). The clinical manifestations are complicated and range from simple febrile seizure to the most severe form of encephalitis. Patients infected by JE virus initially present with non-specific symptoms which include high fever, malaise, anorexia, nausea and vomiting (Innis, 1996; Solomon et al., 2000). During the acute phase which usually
Measuring Japanese Encephalitis (JE) Disease Burden in Asia
80
lasts 3–4 days, the patient develops a change in consciousness of different severity: from mild clouding to stupor and coma. It is during this phase that patients frequently present for health care. Seizures and stiff neck are frequently observed. Less commonly observed are tremor, abnormal movements and cranial nerve involvement. Because the clinical presentation of JE cannot be differentiated from meningoencephalitis of other etiologies, confirmation requires either isolation of the causal virus or specific serological test to demonstrate the presence of anti-JE virus Immunoglubin M (IgM) in the cerebrospinal fluid (CSF) or serum. Currently, in many Asian countries, laboratory diagnosis for JE is not routinely conducted particularly at rural hospitals, resulting in underestimation of JE disease burden in most endemic countries. Case fatality rates also vary but on average 10–30% of JE patients die. For those that survive a JE infection, more than a third will have a JE-related disability including intellectual, behavioral, and neurological sequelae. Late sequelae after a long latent period of recovery have also been observed (Ding et al., 2007). Thus, a major fraction of the disease burden of JE is due to permanent disability among JE survivors, and the mere counting of JE cases and deaths is inadequate to define the burden of JE for a sound decision-making such as introduction of JE vaccine into public health programs. Children between the ages of 1 and 15 years are most commonly affected (Vaughn and Hoke, 1992). Adults can also be infected in areas where the virus is newly introduced (Tsai, 2000). Approximately 3 billion people live in JE endemic areas with 700 million children Figure 80-1 It is clear that the observed JE risk was highest in Shanghai, China, which has a temperate climate; it was the lowest in Bali, Indonesia, a tropical island. The risk in Vietnam and Thailand were intermediate, and both countries belong to subtropical region.
. Figure 80-1 Cumulative incidence rates of Japanese encephalitis (JE) from birth to 15 years of age (Shanghai, China, eight Northern Provinces, Vietnam, and eight Northern Provinces, Thailand) and from birth to 11 years of age (Bali, Indonesia). This figure showed cumulative incidence rates of JE from birth to 15 years of age in a hypothetical cohort of 100,000 healthy newborns in China (Shanghai), Vietnam (eight Northern Provinces) and Thailand (eight Northern Provinces); and the cumulative incidence of JE in a hypothetical cohort of 100,000 new healthy newborns from birth to 11 years of age in Indonesia (Bali)
1395
1396
80
Measuring Japanese Encephalitis (JE) Disease Burden in Asia
. Table 80-1 JE cases, deaths, disability adjusted life year (DALY) in a hypothetical cohort of 100,000 children in China (Shanghai), Northern Provinces of Vietnam, Northern Provinces of Thailand, and Indonesia (Bali) Outcomes (events per 105 persons) No. of JE cases
China
Vietnam
Thailand
Indonesia
271
110
192
No. of JE deaths
36
11
33
63 6
No. of JE disability
81
30
92
23
No. of YLLs
1,259
380
1,154
430
No. of YLDs
1,763
632
2,007
981
Total DALYs
3,022
1,012
3,161
1,412
This table showed JE cases, disabilities, deaths, and disability adjusted life year (DALY) in a hypothetical cohort of 100,000 children in the four study sites in China, Vietnam, Thailand, and Indonesia YLLs: years of life lost; YLDs: years lived with disability; DALYs: disability adjusted life years a In China (Shanghai), Northern Provinces of Vietnam, Northern Provinces of Thailand, a hypothetical cohort of 100,000 children was model from birth till 15 years of age; in Indonesia (Bali), the hypothetical cohort of 100,000 children was model from birth to 11 years old b The age-specific, annual incidence rates were obtained from routine reports from three sites (China, Vietnam, and Thailand). In Bali, Indonesia, we obtained all estimates from a published study of a hospital-based surveillance in 2001–2003 (Kari et al., 2006) c All data used to calculate cumulative incidence rates, deaths, and DALYs in the four countries represent the disease burden estimates in pre-immunization era d All health outcomes are discounted at 3% per annual
In the cohort of 100,000 babies, the model predicted 271 cases, 36 deaths, 81 disabilities, and 3,022 DALYs lost associated with JE from birth through 15 years of age (> Table 80-1) in Shanghai, China; 110 cases, 11 deaths, 30 disabilities and 1,012 lost DALYs were estimated in Vietnam (> Table 80-1); and 192 cases, 33 deaths, 92 disabilities, and loss of 3,161 DALYs associated with JE were assessed for the cohort of 100,000 babies within their first 15 years of life (> Table 80-1) in Thailand. In Bali, Indonesia, 63 JE cases, 6 JE deaths, 23 disabilities, and 1,412 DALYs would be lost due to JE among 100,000 children during their first 11 years of life (> Table 80-1). On average, disability contributes about 58–70% of DALYs lost in these four countries. The overall disease burden is high for all four study sites.
4
DALYs Lost by Age and Contribution of YLL and YLD to the Lost DALYs
Years of life lost due to JE-associated disability (YLD) contributed more to the total DALY loss across all age groups in all study sites. In Vietnam and Indonesia, the contribution to DALYs by YLD was higher in the 0–4 year age groups, and lower in 10–14 year age groups. On the contrary, in Thailand greater contributions to DALYs were made by YLD in the 5–9 and 10–14 year age groups than in the 0–4 year age groups. In China, the highest DALY loss due to YLD was found in 5–9 year age group, which was followed by the 0–4 and 10–14 year age
Measuring Japanese Encephalitis (JE) Disease Burden in Asia
80
groups. In Vietnam, Thailand and Indonesia, loss of YLD accounted for 60–70% of total lost DALYs across all age groups, whereas in China, YLD only accounted for about 58% of total DALYs lost.
5
Conclusions
Disease burden due to JE was the highest in Shanghai, China located in temperate zone, and it was lowest in tropical Bali where mosquitoes are active around the year, and other mosquito-borne diseases (dengue fever, malaria and filariasis) are more prevalent (Ding et al., 2007). The estimate of JE risk in Bali is more accurate because special surveillance studies were carried out with laboratory support (Kari et al., 2006). It is unlikely that the number of DALYs lost due to JE in Bali in 2001–2003 was under-estimated because after 2003 the number of hospitalizations due to JE in a major hospital for sentinel surveillance declined (data not shown). The four study sites may not be representative of the respective four countries, nor of other countries in the climatic zones in which these Asian countries are located. Further studies are needed to explain the distribution of the disease in different geographic regions. Despite the fact that JE disease burden is high across the continent of Asia and JE immunization is cost-saving (Ding et al., 2003) to public health perspectives, only a few countries in Asia have adopted these vaccines in public health programs. Therefore JE continues to be the most important cause of viral encephalitis among children in most Asian countries. There are many reasons for this gap; one is the failure of insufficient knowledge of disease burden to provide the practical information needed by policymakers for JE vaccine introduction. Our study has several limitations. In three of the four sites the estimation of the disease burden relied on routine reports, which were based on statistics of the number of hospitalizations with acute encephalitis syndrome. Laboratory diagnosis of JE was done only for a sample of suspected JE cases. Estimation of true JE risk on the basis of laboratory JE diagnosis for a non-randomly collected sample could be biased. In Bali, Indonesia, we conducted hospital-based and laboratory-supported surveillance at all hospitals with pediatric inpatient service, to capture all hospitalizations due to meningo-encephalitic syndromes. As few patients go out of the island for care, the catchment population was clearly defined; thus the incidence rate of hospitalized JE could be estimated accurately. This kind of surveillance, however, is difficult, expensive and un-sustainable. Another limitation of the study is that the estimation of disease burden is always retrospective, and the disease trend in the future is usually unknown. We assessed JE risk for an epidemic cycle in 15 year for China and in 10 years for Vietnam. However, it could be changed with a change in socio-economic and the natural environment. The secular change of the disease needs continuous investigations in the future. Finally, in most of our calculations, we applied the JE disability index suggested by Murray and Lopez (0.61) (Murray and Lopez, 1996). The arbitrary decision to use a single index for a summary estimate of disease burden may cause bias because the case-fatality rate and fraction of disability varies in different populations and different health care systems. Moreover, it has been argued that the weights provided by Murray and Lopez are based on subjective estimates and are not sufficiently evidenced-based.
1397
1398
80
Measuring Japanese Encephalitis (JE) Disease Burden in Asia
Summary Points As for all infectious diseases, the risk of JE varies by age; lifetime risk should be estimated for defining the disease burden due to JE.
As for all infectious diseases, the risk of JE varies with secular time, and cyclic epidemics are
observed; cumulative age-specific incidence and mortality rates averaged from entire epidemic cycle should be measured for modeling the life-time risk. We used the rates under 15 years of age because almost all JE cases in endemic populations occur in persons under 15 years. Clinically JE is not distinguishable from acute encephalitis of other etiologies. Laboratory assays are necessary for JE diagnosis and thus for estimation of the disease burden. The estimated disease burden could be biased due to poor access to health care, incomplete JE reports, unreliable JE diagnoses, short-term surveillance, poorly defined catchment populations and under-ascertainment of post-JE disability. The estimated disease burden due to JE was high in all four study sites; it was highest in Shanghai, in a temperate region, and lowest on the tropical island of Bali. Years of life with disability (YLDs) contributed 58–70% of the total lost DALYs due to JE in all four study sites. Post-JE disability is the major component of the disease burden. Consideration of acute illness alone does not capture the major part of the disease burden. The disease burden due to JE is high for all four study sites covering temperate, subtropical and tropical regions of Asia. JE immunization is cost-effective or cost-saving. Based on the disease burden in the four study sites, JE immunization should be considered for all JE endemic countries of Asia.
80.1. Japanese encephalitis fact sheet Japanese encephalitis (JE) is the main cause of viral encephalitis throughout Southeast Asia, the Indian subcontinent, Nepal, far eastern Russia, China, Pakistan, Indonesia, the Philippines and Papua New Guinea. It has spread across the Torres Straits into northern Australia and Pacific islands including Guam and Saipan. The JE virus is a member of the flavivirus group and antigenically related to West Nile, St. Louis, Murray Valley and, to a lesser extent, Yellow Fever and Dengue agents. JE is mosquito born, mostly by Culex tritaeniorhynchus which feeds in the evening and at night. Pigs and aquatic birds have a high viremia for 4–6 days after being infected and are the most important reservoir. Children aged less than 15 years have the highest incidence of JE. Victims with encephalitis face a mortality rate of 10–25%. Up to half of the survivors experience significant neurological disabilities.
Acknowledgments This project was supported by grants from Bill and Melinda Gates Children’s Vaccine Program (CVP), Program for Appropriate Technology for Health (PATH), Seattle, WA USA (contract number: 00-GAT.770–790–01149-LPS) and the Korean International Cooperation Agency (KOICA), South Korea (contract number: 2003–138).
Measuring Japanese Encephalitis (JE) Disease Burden in Asia
80
References Burke DS, Leake CJ. (1988). In: Monath TP (ed.) The Arboviruses: Ecology and Epidemiology. CRC, Boca Raton, pp. 63–92. Chakraborty MS, Chakravarti SK, Mukherjee KK, Mitra AC. (1981). Inapparent infection by Japanese encephalitis (JE) virus in West Bengal. Indian J Public Health 24(3): 121–7. Ding D, Kilgore PE, Clemens JD, Wei L, Zhi-Yi X. (2003). Cost-effectiveness of routine immunization to control Japanese encephalitis in Shanghai, China. Bull World Health Organ 81(5): 334–42. Ding D, Hong Z, Zhao SJ, et al. (2007). Long-term disability from acute childhood Japanese encephalitis in Shanghai, China. Am J Trop Med Hyg 77(3): 528–33. Endy TP, Nisalak A. (2002). Japanese encephalitis virus: ecology and epidemiology. Curr Top Microbiol Immunol 267: 11–48. Grossman RA, Edelman R, Willhight M, Pantuwatana S, Udomsakdi S. (1973). Study of Japanese encephalitis virus in Chiangmai Valley, Thailand. 3. Human seroepidemiology and inapparent infections. Am J Epidemiol 98(2): 133–49. Halstead SB, Grosz CR. (1962). Subclinical Japanese encephalitis. I. Infection of Americans with limited residence in Korea. Am J Hyg 75: 190–201. Halstead SB, Jacobson J. (2003). Japanese encephalitis. Adv Virus Res 61: 103–38. Hanna JN, Ritchie SA, Phillips DA, et al. (1996). An outbreak of Japanese encephalitis in the Torres Strait, Australia, 1995. Med J Aust 165(5): 256–60. Igarashi A. (1992). Epidemiology and control of Japanese encephalitis. World Health Stat Q 45(2–3): 299–305. Innis BL. (1996). Japanese encephalitis. In: Porterfield JS, ed. Exotic viral infections. London: Chapman and Hall; 147–73.
Innis BL, Nisalak A, Nimmannitya S, et al. (1989). An enzyme-linked immunosorbent assay to characterize dengue infections where dengue and Japanese encephalitis co-circulate. Am J Trop Med Hyg 40(4): 418–27. Kari K, Liu W, Gautama K, et al. (2006). A hospital-based surveillance for Japanese encephalitis in Bali, Indonesia. BMC Med 4: 8. Lopez AD, Salomon J, Ahmad O, Murray CJL, Mafat D. (2001). Life tables for 191 countries: data, methods and results. Geneva: World Health Organization. Global Programme on Evidence for Health Policy Discussion Paper No.9. 2001. Lowry PW, Truong DH, Hinh LD, et al. (1998). Japanese encephalitis among hospitalized pediatric and adult patients with acute encephalitis syndrome in Hanoi, Vietnam 1995. Am J Trop Med Hyg 58(3): 324–9. Murray CJL, Lopez AD. (1996). The global burden of disease: a comprehensive assessment of mortality and disability from diseases, injuries and risk factors in 1990 and projected to 2020. Cambridge (MA): Harvard School of Public Health on behalf of the World Health Organization and the World Bank. Solomon T, Dung NM, Kneen R, Gainsborough M, Vaughn DW, Khanh VT. (2000). Japanese encephalitis. J Neurol Neurosurg Psychiatry 68(4): 405–15. Solomon T. (2004). Flavivirus encephalitis. N Engl J Med 351(4): 370–8. Tsai TF. (2000). New initiatives for the control of Japanese encephalitis by vaccination: minutes of a WHO/CVI meeting, Bangkok, Thailand, 13–15 October 1998. Vaccine 18(Suppl 2): 1–25. Vaughn DW, Hoke CH, Jr. (1992). Epidemiol Rev 14: 197–221. Zhang YZ, Siriayayapon P, Zhang HL, et al. (2004). Situational analysis of Japanese encephalitis in Dali prefecture, Yunnan Province, China from 1992 to 2001. Endemic Diseases Bulletin (China) (4): 31–5.
1399
81 Pandemic Influenza: Potential Contribution to Disease Burden M. Nun˜o 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1402
2
The Genetic Basis of Influenza Pandemics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1403
3 3.1 3.2 3.3 3.4 3.5
The Burden of Disease Attributable to Influenza . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1404 Quantifying the Burden of Influenza . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1405 Mortality and Morbidity Burden . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1405 Evaluating the Burden in Non-Temperate Regions of the World . . . . . . . . . . . . . . . . . . 1407 Pediatric Burden . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1408 Socioeconomic Burden . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1411
4 Global/National Models for Evaluating the Burden . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1411 4.1 The Basic Reproduction Number . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1412 4.2 Modeling/Quantifying the Burden of Influenza . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1412 5 Learned Lessons and New Challenges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1413 5.1 Control Measures to Combat a Pandemic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1413 6
Concluding Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1414 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1415 Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1416
#
Springer Science+Business Media LLC 2010 (USA)
1402
81
Pandemic Influenza: Potential Contribution to Disease Burden
Abstract: Records of disease outbreaks resembling influenza date to the writings of Hippocrates (fifth century BPE). Since then, influenza has afflicted humans around the globe. The most severe (‘‘Spanish Flu’’ > pandemic) of three major outbreaks of the twentieth century killed approximately 20–50 million people worldwide. More recently, the global spread of highly pathogenic bird-adapted strain H5N1 is considered a significant pandemic threat. Since 2003, a total of 379 cases and 239 deaths have been reported. This chapter provides an overview of the genetic characteristics of the virus that elucidate its ability to continuously evade a host’s immune system; it describes some of the approaches used to quantify the burden of influenza and discusses their implications for the prevention and containment of future pandemics. The preliminary findings of the studies discussed here suggest that influenza-related burden is highly underestimated in tropical and subtropical regions of the world. This implicates that proper assessment of influenza-related morbidity and mortality worldwide is essential in planning and allocating resources to protect against what could be one of mankind’s most devastating challenges. A summary of learned lessons from past influenza pandemics are described and new intervention strategies aim at curtailing a future pandemic are discussed. More importantly, however, is the discussion of today’s challenges such as antiviral resistance, limited resources in a world that is globally connected and the imminent gap between the capacity (resources available) of developed and developing parts of the world to respond to a pandemic. List of Abbreviations: CDC, Center for Disease Control and Prevention; FF, DM, Currency Code for France and Germany; Flu, Influenza; HA, Hemmaglutinin; H1N1, > H2N2, H3N2, H5N1, Influenza A Virus Subtypes; ILI, Influenza-Like-Illness; NA, Neuraminidase; NB, Influenza B Virus Glycoprotein; NPI’s, Non-Pharmaceutical Interventions; PB1-F2, Proapoptotic Influenza A Virus Protein; PI’s, Pharmaceutical Interventions; P&I, Pneumonia & Influenza; R0, Basic Reproduction Number; S-I-R Model, Susceptible-Infected-Recovered Model
1
Introduction
Influenza (flu) is among the most ancient of pathogens of man and the most thoroughly studied of viruses, yet their rapid evolution makes control of > epidemics and the prevention or even mitigation of pandemics a persistent public health challenge. The replication of the flu virus is noisy, i.e., offspring produced are remarkably variable. Flu viruses mutate continuously in so-called > ‘‘antigenic drift,’’ a strategy that allows the virus to evade the least experienced and adaptable immune systems in the human host population, the young and the very old. This chapter is organized as follows:
Section 1 provides a brief overview of the genetic basis for pandemics. Section 2 Provides a historical overview of the worldwide impact of flu epidemics and pandemics,
Describes approaches used in assessing flu related disease burden, Highlights studies that have quantified morbidity, mortality and socioeconomic costs in countries around the world, and
Pandemic Influenza: Potential Contribution to Disease Burden
81
Especially considers the burden in children, the most infected and infectious population group.
Section 3 presents global and national modeling frameworks for evaluating the burden. Section 4 discusses the learned lessons from previous flu epidemics and pandemics, highlights new challenges facing humans today, and discusses preparation strategies for curtailing future pandemics. Section 5 concludes with thoughts on the future burden of flu.
2
The Genetic Basis of Influenza Pandemics
Epidemics caused by influenza viruses occur every winter in temperate regions of the globe. The size of these epidemics varies from year to year and from place to place. Very large epidemics associated with the widespread circulation of genetically novel influenza viruses, socalled pandemics, are caused only by influenza viruses of a single type (type A). Influenza (flu) virus, a member of the family Orthomyxoviridae is a segmented single-stranded RNA virus which exists in three types, A, B and C. Only types A and B cause disease of any consequence in humans. The genomes of both A and B viruses are comprised of eight gene segments which code for eleven viral proteins. Ten of these are functionally similar. The NB protein is unique to B viruses and the recently discovered PB1-F2 protein is found only in A viruses (Chen et al., 2001; McCullers et al., 2004; Steinhauer and Skehel, 2002). Two of these proteins, hemaglutinnin (HA) coded for by the gene segment conventionally labeled #4, and neuraminidase (NA), coded for by the segment labeled #6, appear on the surface of the influenza virion, and are responsible for the greater portion of the immune response in humans. All B-type influenza viruses share variants of the same HA and NA proteins. The > clinical attack rates for A and B type viruses are similar in children (McCullers et al., 2004); but the overall clinical attack rate is much lower in epidemics of B-type than A-type viruses. In addition, it has been recently demonstrated that mortality associated with influenza B epidemics has declined about linearly over the most recent 40 years (Reichert et al., 2007). This difference in epidemiology suggests that the circulation of influenza B viruses in human populations is determined by population-based immunity. Reassortment(> genetic reassortment) among human influenza B viruses is believed to be the genetic basis for sporadic increases in observed attack rates for B-viruses (McCullers et al., 2004). A-type viruses, however, exhibit another dimension of genetic complexity. A-type viruses circulate not only in humans but in many non-human reservoirs, most notably aquatic birds. The sudden appearance in A-type viruses responsible for human influenza infections of one or more gene segments from influenza viruses that previously circulated only in non-human reservoirs produces chimeric influenza viruses that are sufficiently immunologically novel that very widespread illness and injury occurs among humans. It is also possible that a virus that previously circulated only in a non-human reservoir could mutate sufficiently that it becomes de novo capable both of producing disease in humans and being widely transmissible among humans. Such a virus is maximally different from previous human immunological experience (all gene segments are novel) and potentially could produce the most severe pandemics. Those newly emergent viruses whose genomes are comprised of a substitution of only some of the gene segments of viruses that previously circulated in humans are thought to have arisen via reassortment during dual infections involving both a non-human and human influenza virus.
1403
1404
81
Pandemic Influenza: Potential Contribution to Disease Burden
The appearance of an influenza virus with a substituted gene segment is called > ‘‘antigenic shift’’ (Hall et al., 1973; Monto and Kioumehr, 1975; Reichert et al., 2007). Sixteen immunologically distinguishable forms of HA (H1-H16) have been identified and nine of NA (N1-N9). Because these two proteins are found on the surface of the virion, influenza has long been typed by reference to these two antigens. Only three HA subtypes and two versions of NA have been found in viruses that have circulated widely in humans. Pandemics of the twentieth century and their dates of emergence in the northern/southern hemispheres were caused by viruses of type H1N1 (1918/1918), H2N2 (1957/1958), and H3N2 (1968/1969). The 1918–1919 pandemic virus emerged en bloc as an avian virus the mutation and adaptation of which permitted replication in and transmission among humans (Taubenberger et al., 2005; Tumpey et al., 2005; Webster et al., 1992). The 1957 pandemic H2N2 virus emerged with substitutions of the HA, NA and PB1 gene segments into the evolved H1N1 viruses from an avian virus. The 1968 pandemic (H3N2) resulted from reassortments of the same avian sequences into multiple lineages of the evolved H2N2 viruses resulting in substitutions of HA and PB1 genes (Lindstrom et al., 2004). In each of the twentieth century pandemics, therefore, both the HA and the PB1 genes were substituted. The nomenclature for distinguishing immunologically and epidemiologically important influenza A-type viruses is, therefore, incomplete. H1N1 viruses re-emerged in 1977 in what is widely believed to have been an accidental release from a research laboratory, probably in China. Currently the subtypes H1N1 and H3N2 are circulating widely in humans. A small number of isolates of H1N2 viruses have also been reported. Therefore, the only A-type influenza viruses not circulating at present are those with hemagglutinin of type 2, H2. In this context, we are all spectators to the brush with emergence of influenza A viruses of type H5N1 including a unique-to-human-experience PB1 gene, which is widely extant in the avian reservoir, causing massive disease in domestic poultry with a sporadic history of infecting small numbers of heavily exposed humans. With a > case fatality rate of approximately 50%, these viruses pose the potential for a redux of the Great Pandemic of 1918 in which 1.5–3% of humanity perished. Records of disease outbreaks resembling influenza date from the writings of Hippocrates (fifth century BPE). It is likely, however, that humans have suffered from influenza shortly, in evolutionary time, after the domestication of animals in which we now recognize the circulation of pathogenic influenza viruses, and in epidemic proportions since the rise of town-size urbanizations. The pandemic of 1580 is thought to be the first ‘‘confidently identified as (associated with) influenza’’ (Pyle, 1986); but ‘‘Clear’’ descriptions are attributed to authors in the tenth century (Langmuir and Farr, 1976). The first epidemiological level description of an influenza pandemic is due to William Farr who is credited with conceptualizing the burden of influenza as the mortality in excess that expected in the pandemic of that year in London. Detailed records begin with the pandemic of 1889–1892, with age-specific mortality, but serologic data became available only in the 1930s.
3
The Burden of Disease Attributable to Influenza
In 1847 William Farr described the burden of flu and developed methods to quantify its contribution to mortality (Farr, 1847; Langmuir and Farr, 1976). Since then, numerous studies have been proposed to quantify the burden of flu in humans. Flu outbreaks in temperate climates occur from November through April in Northern and from May through September in Southern hemispheres. While seasonal periods characterize
Pandemic Influenza: Potential Contribution to Disease Burden
81
flu outbreaks in Northern and Southern hemispheres, outbreaks in tropical climates are highly irregular and therefore more difficult to predict. Flu pandemics are worldwide epidemics that involve high morbidity and mortality resulting from major mutations in viral genome (antigenic shift). Unlike flu epidemics, pandemics do not necessarily occur in seasonal time. Some criteria for defining a major pandemic include the occurrence of a new emerging virus subtype for which humans have no immunity to and effective person-to-person contact. Three major pandemics have been observed in the last century. The first and foremost severe, the ‘‘Spanish Flu’’ (H1N1), occurred during 1918–1919. It is estimated that 20–50 million individuals succumbed to the disease worldwide (approximately 2% of a world’s population), more than 25% of the US population became ill and 2.5% of those infected in the US died. The next major outbreaks, the ‘‘Asian Flu’’ (H2N2) and ‘‘Hong Kong Flu’’ (H3N2) emerged in 1957 and 1968, respectively. While the impact of these latter pandemics was considerably mild when compared to 1918–1919 pandemic, 2 and 1 million worldwide deaths were estimated to have come from the 1957 and 1968 pandemics, respectively (Simonsen et al., 1998).
3.1
Quantifying the Burden of Influenza
The concept of > excess mortality was first introduced by William Farr (Farr, 1847; Langmuir and Farr, 1976) to describe influenza epidemics in London. This approach of excess mortality was later applied by Serfling in 1963 and various modifications of this approach are now widely used (Serfling, 1963). Based on the excess criteria, an epidemic occurs when the number of disease cases (mortality) exceeds the number expected. However, defining the term epidemic may differ mildly. For instance, the US denotes an epidemic threshold for mortality based on estimated of the expected number of deaths given in a particular week accompanied by 95% confidence intervals around the projected estimate. With this approach, an epidemic is reported when the upper 95% confidence interval is exceeded.
3.2
Mortality and Morbidity Burden
Prior to 1937, more than 90% of excess deaths were properly attributable to pneumonia. However, over the following decade total excess mortality has dropped to 1/3 that level and the fraction of the remainder attributable P&I dropped by 30%. Currently, this fraction has dropped to below 10% making the evaluation of flu-related burden a more challenging task. The burden associated with flu has been typically evaluated by a criterion that uses deaths recorded by physicians as caused by pneumonia and influenza (P&I) along with other measures such as influenza-like-illness (ILI). Excess mortality due to flu is calculated by estimating a baseline of deaths that would be expected in the absence of flu virus activity, and the number of deaths actually observed. This criterion, however, is not very sensitive as it fails to evaluate flu-related deaths that could have occurred outside the specified flu season or that could have been missed since several other diseases exhibit flu-like symptoms. Although P&I data has been widely used to evaluate excess mortality, it seems to account for about 25% of all flu related deaths; this suggests that estimates based on P&I may not be appropriate to measure the total impact of influenza on mortality (Simonsen et al., 1997). Further efforts aimed to improve excess mortality estimates due to flu have used > all-cause mortality data instead.
1405
1406
81
Pandemic Influenza: Potential Contribution to Disease Burden
The Serfling methodology (and various modifications of it) has been used to parameterize a baseline model based on statistical expectations (95% confidence interval of the baseline) by training data from non-epidemic years. A surveillance system based on the Serfling approach signals an epidemic whenever the observed time series data (e.g., P&I, all-cause mortality data) exceeds the established threshold. The model assumes an average mortality described by b0, a linear trend denoted by b1*t, 52-week cyclical period (or 52.1667 for adjusted leap years) denoted by b2cos(2pt/52) + b3sin(2pt/52). This model can be denoted by the following equation: YðtÞ ¼ b0 þ b1 t þ b2 cosð2pt=52Þ þ b3 sinð2pt=52Þ þ error where Y(t) denotes the estimated mortality for week t. Variations of this model that account for potential deviations from a linear trend approach include a quadratic term b1*t2. We illustrate the proportion of deaths in the United States attributed to pneumonia and influenza (P&I) reported by the 122 cities mortality reporting system. > Figure 81‐1 illustrates that the seasonal peaks observed for outbreaks in 1985–1989 seem to have exceed the epidemic threshold. The sinusoidal function across denotes the seasonal baseline that is established based on periods outside the flu season. US estimates proposed by the Center of Disease Control (CDC) show that each year 5–20% individuals get infected with the virus, some 200,000 individuals get hospitalized and 36,000 die from the disease. Simonsen et al., modified this model to assess the burden of flu on mortality based on P&I and all-cause mortality weekly data for 20 years
. Figure 81‐1 Pneumonia and influenza mortality for 122 US cities during 1985–1989
Pandemic Influenza: Potential Contribution to Disease Burden
81
(United States: 1972–1992). This study showed that excess mortality estimates were comparable to previous estimates that were based on monthly data (Simonsen et al., 1997). More significantly however, this study showed that P&I excess mortality data only captured 25% of the all-cause excess mortality estimates. In a separate study, Simonsen et al. (Simonsen et al., 2000) estimated an average seasonal rate of excess P&I hospitalization of 49 cases in 100,000. This study further showed that the average risk of flu-related P&I hospitalization was twice as high during A (H3N2) seasons than those observed during seasons dominated by subtypes H1N1 and B. Individuals of age less than 65 had a 57% of all-flu-related hospitalizations, however the average seasonal risk of flu-related P&I hospitalization was significantly higher for this risk group. Further modifications of the Serfling model have been implemented. Thomson et al. proposed a Poisson regression model to predict excess cases of P&I during influenza epidemics (1979–2001). This study reported significant influenza-related hospitalizations among elderly (50–64 years) and children younger than 5 years old (Thompson et al., 2003). Their findings showed that a yearly average of 133,900 cases listed as P&I hospitalizations were actually associated with the flu virus. A more recent study implemented a Poisson extension of the seasonal type model to predict excess cases of P&I during 16 seasons of flu epidemics in France (Denoeud et al., 2007). This study showed that morbidity data (flu-like-illness and virological data) may be used to predict excess P&I. The implications of this study provide a real possibility of flu-related burden for countries in which specific pneumonia and flu mortality surveillance data is simply not available.
3.3
Evaluating the Burden in Non-Temperate Regions of the World
Most studies have focused on assessing the mortality of flu in regions of the world that exhibit temperate climates, however more recent efforts have demonstrated that burden of flu in the tropics/subtropics is significant. Singapore’s yearly clinical infection from seasonal include are estimated at 20% (Ng et al., 2002). In a recent study, Lee et al. (Lee et al., 2007) implemented a linear regression model to evaluate the excess mortality due to flu pandemics of the last century. The authors estimated an excess > mortality rate of 7.76 per 1,000 people during May–June and October–November of 1918. Using a similar approach (Murray et al., 2006) these estimates reached a rate of 18 per 1,000 people. Estimates of monthly excess mortality for the 1957 pandemic yield 0.47 per 1,000 people (mid May). Although excess mortality during the1968 pandemic seems mild when compared the 1918 and 1957 outbreaks, this pandemic exhibited two waves. The first wave occurred in mid August with estimated monthly excess mortality of 0.27 per 1,000 people (543 deaths in a population of 2,012,000). Excess deaths peaked again in May–June of 1970 and with an estimated excess mortally in the order of 0.15 per 1,000 people (309 deaths in a population of 2,074,500). Most significantly, this study showed that excess mortality estimates in Singapore are comparable, if not, higher than those observed in temperate regions. Their 1918 pandemic estimate of 1.80% (18 deaths in 1,000 people) exceeds global estimates of 1.06% and Taiwan’s rate of 1.44% (Murray et al., 2006). Based on the findings in Lee et al. and studies referenced therein, > Table 81‐1 summarizes the burden of flu in the 1918–1919 pandemic and it illustrates that tropical (sub-tropical) countries were affected more significantly than regions with temperate climates. These findings show that Kenya was the most affected country with an upper bound mortality rate estimate of 57.8%. Furthermore, four of the five highest affected countries corresponded to countries with tropical-subtropical climates.
1407
1408
81
Pandemic Influenza: Potential Contribution to Disease Burden
. Table 81‐1 Estimated deaths and mortality rates attributed to influenza during the pandemic of 1918–1919. Countries are listed in order (highest to lowest) according to the estimated mortality rate Country
Climate
Kenya
Tropical
No. deaths (in Mortality rate (per 1,000) 1,000) % 104–150
Mortality rate ranking (high to low)
40–57.8
1
South Africa
Subtropical
300
44.3
2
India
Tropical
185
6.1–43.9
3
The Philippines
Tropical
81.0–288
8.0–28.4
4
Portugal
Temperate
59.0–159
9.8–26.4
5
Singapore
Tropical
2.87–6.66
7.8–18.0
6
Ceylon (Sri Lanka)
Tropical
51.0–91.6
10.0–17.9
7
Spain
Temperate
257–311
12.3–14.9
8
Taiwan
Subtropical
25.4–52.8
6.9–14.4
9
Japan
Tropical/ Temperate
368–517
6.7–9.4
10
United States
Temperate
402–675
3.9–6.5
11
Canada
Temperate
50.0–51.0
6.1–6.3
12
England
Temperate
116–200
3.4–5.8
13
Argentina
Temperate
10.2–46.0
1.2–5.4
14
British Honduras Subtropical (Belize)
1.01–2.00
2.3–4.6
15
Denmark
Temperate
6.02–12.4
2.0–4.1
16
Australia
Temperate
14.5–15.4
2.7–2.9
17
Trinidad and Tobago
Tropical
0.30–1.00
0.1–0.2
18
This table was adapted from the results presented in Lee et al. (Lee et al., 2007)
3.4
Pediatric Burden
School-age children play a significant role in spreading the flu in a population. Children touch their noses, eyes, mouths, interact closely with other children and have contacts with members of their family. They intensify the spread of the disease since they shed the virus for longer periods of time and at higher virus titers than adults. In addition to their risk to others, the inexperience immune system of children enhances their risk of developing complications from an infection. Annual recommendations from the Center of Disease Control include the vaccination of children aged 6 months–5 years; however more recent discussions recommend vaccination of all-young age children. Several studies have been reported on the burden of flu on morbidity and mortality in children (Neutzil et al., 2002; Reichert et al., 2001; Simonsen et al., 1998). A prospective surveillance study (1974–1999) of 1665 healthy children younger than 5 years showed annual
Pandemic Influenza: Potential Contribution to Disease Burden
81
infection ranging between 15 and 42% (Neutzil et al., 2002). It was further shown that children younger than 2-year of age were more likely than older children to have serious complications such as pneumonia, croup, chonchiolitis and sepsis (Neutzil et al., 2002). A study assessing age-associated burden of flu in epidemic and pandemic periods showed that half of flu-related deaths during the 1968–1969, with larger proportions of those 1957–1958 and 1918–1919 were attributed to people younger than 65 years of age. This study further showed that the high number of cases observed in these pandemics decrease significantly in prospective outbreaks. More particularly, they found that after each pandemic, the absolute risk of flu-related mortality among people younger than 65 decreased from 7- to 28-fold over the following decade during severe epidemics. In contrast, the corresponding risk reductions among those people older than 65 years of age range from 2- to 3-fold or less (Simonsen et al., 1998). A multiple source data study of the UK during 1995–1999 showed that on average 10% of children contract clinical influenza while a further 20% may be asymptomatic for children younger than 5 years old (Watkins, 2004). While vaccination is the optimal choice for protecting young age children, and thereby, reducing the burden of flu in this group, it has also been shown that the risk to other members of the population can be reduced through children vaccination. Ahmed et al. showed that vaccinating young age children improved community level protection (Ahmed et al., 2001). So far, our discussion of the burden of flu in children has been limited to countries in which flu outbreaks arise during the winter. However, recent studies of flu-related mortality in tropical and subtropical regions show that burden to children in these parts of the world could exceed the US estimates. Unlike developed countries, children in developing countries face additional challenges such as malnutrition, higher risks associated with bacterial infections, limited access to pharmaceuticals, limited health care, and poor living conditions (Simonsen, 2001). Studies of Cuba and Singapore reported 3–15% flu-related viral isolates in hospitalized children (Cancio et al., 2000; Chew et al., 1998). Intervention plans aimed at protecting high-risk, as well as, all other members of a population against seasonal influenza have varied from country to country. Japan is the only country that has ever adapted a vaccination program based on children rather than adults. In a consequence to their most destructive pandemic experience in 1957 (approximately 8,000 deaths), Japan legislated a vaccination program that focused on school-age children (7–15 years). Under this program, vaccination levels reached 80%. > Figure 81 ‐2 illustrates decreasing mortality trends from all-cause and pneumonia related mortalities. However, this figure also shows that as soon as vaccination measures were relaxed in 1987, flu-related mortality increased. The rising levels of excess mortality became even more evident after government discontinued the vaccination program in 1997. It has been estimated that the vaccination of children in Japan prevented about 37,000–49,000 deaths per year (Reichert et al., 2001). Other studies have shown that high (50–70%) levels of vaccination among children can provide effective protection to other members in a community (Elveback et al., 1976; Longini et al., 1978; Longini et al., 1988). Chiu et al. showed that excess hospitalization in Hong Kong rates were higher that US estimates (Chiu et al., 2002). It is evident that influenza contributes significantly to hospitalization among children in both temperate and tropical regions. More particularly, these findings illustrate that the rates reported for Hong Kong exceed those of the US and that trends of decreasing hospitalization rates with increasing age are evident in both estimates. > Table 81 ‐2 reports estimates of flu associated hospitalizations per 10,000 for children in various age groups.
1409
1410
81
Pandemic Influenza: Potential Contribution to Disease Burden
. Figure 81‐2 Five-year moving average of excess deaths attributed to both pneumonia and influenza (P&I) and all-cause mortality, for Japan and the United States
. Table 81‐2 Estimated flu associated hospitalization rates per 10,000 for the US and Hong Kong United States
1979–1993
age 82-3), including reduced illness, mortality, hospitalization, absenteeism, costs of healthcare, and antiviral use (Bridges et al., 2000; . Table 82-2 Effect of vaccination on healthcare personnel illness and absenteeism Randomized trial 1 (Saxe´n and Virtanen, 1999)
Vaccinated
Controlled
Reduction (due to vaccination)
Days of lost work (due to respiratory infection)
1
1.4
28%
Days felt unable to work (whether on or off duty)
2.5
3.5
28%
Randomized trial 2 (Wilde et al., 1999)
Vaccinated
Controlled
Days of lost work (due to illness)
9.9 (per 100 persons)
21.1 (per 100 persons)
53%
Total respiratory illnesses
28.7 (per 100 persons)
40.6 (per 100 persons)
29% 87%
Incidence of influenza
1.7%
13.4%
Cross-sectional survey (Lester et al., 2003)
Vaccinated
Unvaccinated
Days of lost work (due to influenza-like 63 (per 100 illnesses) persons)
69 (per 100 persons)
0.08%
Influenza-like illnesses
42 (per 100 persons)
54 (per 100 persons)
22%
Days of illness
272 (per 100 persons)
374 (per 100 persons)
27%
The impact of influenza vaccination on healthcare personnel illness and absenteeism was evaluated in two randomized, placebo-controlled, double-blind trials (Saxe´n and Virtanen, 1999; Wilde et al., 1999), and a crosssectional survey (Lester et al., 2003)
1423
1424
82
Prophylaxis of Healthcare Workers in an Influenza Pandemic
. Table 82-3 Cost-effectiveness of influenza vaccination of adults aged Figure 82-2 summarizes the likely prophylactic use of drugs for target groups of healthcare workers and essential service personnel in national pandemic plans published in the official websites of the Food and Agricultural Organization of the United Nations and the WHO National Influenza Pandemic Plans (FAO United Nations, 2008; WHO, 2008). Although recent studies provide scientific and
1425
. Figure 82-1 Global map of pandemic preparedness. Countries with national plans that indicate prophylaxis of healthcare workers and/or essential service staff as part of their pandemic responses. Countries with no, or not yet determined, prophylaxis strategy in their national plans. Countries with no published/accessible pandemic plans through official websites of the Food and Agricultural Organization of the United Nations, and the WHO National Influenza Pandemic Plans
1426
82 Prophylaxis of Healthcare Workers in an Influenza Pandemic
Prophylaxis of Healthcare Workers in an Influenza Pandemic
82
. Figure 82-2 Prophylactic use of antiviral drugs in national pandemic plans. Among fifty-one published contingency plans for pandemic responses (FAO United Nations, 2008; Straetemans et al., 2007; WHO, 2008), healthcare workers are identified as the priority group for pre-exposure (long-term) prophylaxis in 17 national plans (Australia, Bahrain, Republic of Bulgaria, Chile, Czech Republic, Germany, Greece, Hungary, Lithuania, Montenegro, Nicaragua, Norway, Republic of Serbia, Sweden, Switzerland, Timor Leste, United States), and for post-exposure prophylaxis in eight plans (Austria, Cuba, Ireland, Republic of Korea, Mexico, Philippines, Portugal, Singapore). Essential service personnel are given priority for long-term prophylaxis in seven plans (Republic of Bulgaria, Chile, Czech Republic, Greece, Nicaragua, Norway, Sweden), and for post-exposure prophylaxis in six plans (Australia, Montenegro, Philippines, Portugal, Republic of Serbia, Singapore). A number of these plans indicate that the decision on prophylactic use of drugs may be influenced by the availability of vaccines and the size of antiviral stockpiles, in addition to the severity of the pandemic strain. Twenty-five national plans (Belgium, Brazil, Canada, China, Republic of Estonia, Finland, France, Hong Kong, India, Italy, Japan, Luxemburg, Nauru, Netherlands, New Zealand, Nouvelle-Cale´donie, Palau, Poland, Rwanda, Slovak Republic, South Africa, Spain, Thailand, United Kingdom, West Bank/Gaza) indicate no specific strategy for prophylaxis of healthcare workers or essential service personnel. Most plans suggest that the strategic use of drugs should be re-evaluated during different phases of the pandemic, so that necessary modifications to antiviral strategies can be made in a timely fashion
administrative frameworks for prophylaxis strategies that can effectively protect the healthcare workforce in the face of infectious disease threats (Cinti et al., 2005; Toner and Waldhorn, 2006; Xiong et al., 2007), precise planning for antiviral use is hindered by several unknowns – in particular, the transmissibility of the pandemic strain, effectiveness of drugs, and the impact of other curtailing measures. There is also much uncertainty about how a novel influenza strain would affect different populations with distinctly different mobility patterns, healthcare expenditures, and economic structures. The nature of the next influenza pandemic cannot be predicted with certainty, but the high mortality rates among younger adults in past pandemics compel us to consider maximizing the protection of healthcare workers by using available means of reducing disease transmission.
1427
1428
82 5
Prophylaxis of Healthcare Workers in an Influenza Pandemic
Benefits, Logistical Constraints, and Consequences of Prophylaxis
Considering the limited capacity of vaccine production that is concentrated in a very few countries (Fedson, 2003), antiviral drugs will likely be the primary control measure for several months following the emergence of a pandemic strain. Yet more frightening, antiviral therapy may be the only pharmaceutical option for many countries during the entire course of the pandemic. Stockpiling adequate supplies of drugs has therefore become a critical component of pandemic preparedness strategies. However, a major challenge is to define priority groups and formulate antiviral policies that are most likely to optimize the health of the greatest number of individuals in the face of an influenza pandemic. In allocating antiviral drugs, it will be necessary to rate the relative importance of treatment and prophylaxis in avoiding complete coverage of the population, particularly when considering insufficient quantities of drugs, limited production capacity, and a surge in demand for antiviral therapy with the progression of a pandemic (Democratis et al., 2006). These contrasting factors highlight the importance of prioritizing public health policies that maximize both short-term population-wide benefits and long-term epidemiological effects of antiviral therapy. Published modeling studies suggest that the pandemic can be contained at the source if early treatment of diagnosed cases is combined with targeted blanket prophylaxis and social distancing measures (Ferguson et al., 2005; Longini et al., 2005). Significant assumptions are embedded in the core of such models, most of which are unlikely to be fulfilled in a real world environment, and therefore containment failure should be anticipated in devising effective preparedness countermeasures. Clearly, targeted strategies are needed to optimize the use of available supplies of antiviral drugs, and precise planning should be based on pandemic-specific response goals for mitigating the overall disease burden. Long-term prophylaxis of healthcare workers and emergency service personnel may be beneficial in terms of reducing viral transmission between interconnected subpopulations with high levels of exposure, thereby preventing impending institutional outbreaks and staff demoralization, maintaining the quality and accountability of healthcare, and decreasing mortality of the general population. However, this strategy would require a prohibitively large drug stockpile, and delay the progression of the pandemic by the period for which prophylaxis is provided. Regardless of the level of stockpiles, substantial logistical challenges will arise for practical implementation of this strategy that need to be addressed, let alone its possible long-term epidemiological outcomes that may be disadvantageous to the control of disease. Key issues include, but are not limited to, determining target groups, potential regulatory barriers to dispensing, labor consequences that may arise from targeting specific sectors in healthcare and emergency departments but not others, and monitoring antiviral drug use, efficacy, and safety. Targeted > post-exposure prophylaxis could provide significant benefits, but it would still require much greater stockpiles than currently provided.
5.1
Antiviral Resistance
Although antiviral therapy appears to be crucial in any containment strategy, the emergence of drug-resistance will impose significant threats to the effectiveness of drugs for both treatment and prophylaxis (Moscona, 2005). Early treatment largely inhibits generation of resistant viruses by suppressing viral replication, but results in a longer time for selection in favor of pre-existing resistant mutants (Moghadas et al., 2008). Prophylaxis blocks the
Prophylaxis of Healthcare Workers in an Influenza Pandemic
82
transmission of the drug-sensitive strain more effectively, but contributes to the emergence of resistance in the population by increasing the pressure for selection and spread of resistant viruses (Lipsitch et al., 2007; Regoes and Bonhoeffer, 2006). The interplay between these opposing effects appears particularly important for determining the advantages and disadvantages of population-specific antiviral strategies. Previous studies suggest that, in the absence of drug-resistant viruses, targeted prophylaxis of healthcare workers can substantially limit the community-level transmission of disease, by providing considerable protection to those who contribute to caring for influenza patients (Gardam et al., 2007; Lee and Chen, 2007). However, should resistance emerge during the early stages of a pandemic, this strategy may promote population-wide spread of drug-resistance through intense contacts with patients and colleagues in the healthcare setting and contacts outside of the workplace. As has been evident in the experience with outbreaks of Severe Acute Respiratory Syndrome (SARS) in 2003, hospitals and other healthcare facilities are particularly important resources for disease control, but they may also serve as ‘‘hot spots’’ for disease transmission, with subsequent hospital-to-community spread (Loutfy et al., 2004).
6
Concluding Remarks
The impact of the next influenza pandemic may be catastrophic worldwide, particularly for nations with poor healthcare resources and fragile economies. Ongoing efforts may not be sufficient to combat it or prevent every single death, but the basic tenet of pandemic preparedness is not to miss the unprecedented opportunity provided by the increased awareness of the potential devastation of the next influenza pandemic. To a large degree, it will run its course; however, the healthcare system and society will need to prepare for a rapid and effective response by developing system-wide contingency plans that accommodate the full spectrum of harms related to pandemic influenza and benefits related to control activities. Historical precedent, from both seasonal influenza epidemics and the 1918–1919 influenza pandemic, suggests that an emergent influenza strain with high > pathogenicity would severely tax existing health resources, and force healthcare administrators and providers alike to make difficult decisions that may include rationing of scarce resources (e.g., antiviral drugs, intensive care beds, and ventilators) and decisions that violate the autonomy of the individual (e.g., forced quarantine or isolation of individuals). The ethical framework of such decisions is complex; however, evaluation of the potential benefits, costs, and limitation of competing strategies, as well as the resources and practices required to achieve best outcomes, will allow planners and providers to balance the protection of community health against individuals’ rights and freedoms in the context of influenza infection control.
Summary Points Effective protection of healthcare workforce during an influenza pandemic is an essential priority of pandemic preparedness endeavors, and should be based on scientific knowledge and evidence of disease transmission, prevention, and mitigation. Such protection must address the diversity of institutional activities with varying degrees of exposure to the disease. Considering the nature of interaction between ill individuals and healthcare workers, the lack or poor administration of preventive measures will leave healthcare workers highly vulnerable to the infection, and therefore the surge capacity required to treat influenza patients could be severely impaired.
1429
1430
82
Prophylaxis of Healthcare Workers in an Influenza Pandemic
Prophylaxis of healthcare workers is an essential protective mechanism that can maintain
the capacity to cope with the increased demand for healthcare, prevent nosocomial outbreaks and further spread to the community, and reduce morbidity and mortality of the general population. An effective vaccine may not be available for several months following declaration of a pandemic. Preparedness efforts should therefore include securing antiviral supplies, particularly when prophylactic use of drugs is planned for limiting disease propagation locally as well as globally. Significant logistical challenges lie at the heart of prophylaxis strategies for healthcare workers and essential service personnel that must be addressed. Scarce supplies of antiviral drugs create ethical dilemmas of distribution priorities and target groups. These ethical considerations must be taken into account for rationing of limited stockpiles. Antiviral treatment can induce a massive selective pressure on the drug-sensitive pandemic strain, resulting in the clinical outcome of developing drug-resistance. Prophylactic use of drugs can substantially enhance the spread of drug-resistance in the population due to a significant reduction in susceptibility to the drug-sensitive strain. In order to prevent unintended epidemiological consequences of drug therapy, antiviral strategies should be re-evaluated during the progression of the pandemic and necessary adaptations should be rapidly implemented to prevent subsequent outbreaks of drugresistant infections. In order to identify more tangible antiviral strategies with a global perspective, similarities and differences in pandemic plans with respect to the use of antiviral drugs and priority groups for treatment and prophylaxis should be carefully examined. Development of appropriate mathematical models for disease management can help evaluate the potential impact of various public health intervention strategies, optimize health policy decisions, and maximize population-wide benefits of scarce health resources.
References American College of Physicians. The Health Care Response to Pandemic Influenza. Philadelphia, American College of Physicians, 2006. Baccam P, Beauchemin C, Macken CA, Hayden FG, Perelson AS. (2006). J Virol. 80: 7590–7599. Bridges CB, Kuehnert MJ, Hall CB. (2003). Clin Infect Dis. 37: 1094–1101. Bridges CB, Thompson WW, Meltzer MI, Reeve GR, Talamonti WJ, Cox NJ, Lilac HA, Hall H, Klimov A, Fukuda K. (2000). JAMA. 284: 839–844. Bright RA, Medina MJ, Xu X, Perez-Oronoz G, Wallis TR, Davis XM, Povinelli L, Cox J, Klimov AI. (2005). Lancet. 366: 1175–1181. Cinti S, Chenoweth C, Monto AS. (2005). Infect Control Hosp Epidemiol. 26: 852–854. Cox NJ, Tamblyn SE, Tam T. (2003). Vaccine. 21: 1801–1803.
Democratis J, Pareek M, Stephenson I. (2006). J Antimicrob Chemother. 58: 911–915. FAO United Nations. (2008). http://www.fao.org/avianflu/en/national.html Fedson DS. (2003). Clin Infect Dis 36: 1552–1561. Ferguson NM, Cummings DAT, Cauchemez S, Fraser C, Riley S, Meeyai A, Iamsirithaworn S, Burke DS. (2005). Nature. 437: 209–214. Fraser C, Riley S, Anderson RM, Ferguson NM. (2004). Proc Natl Acad Sci USA. 101: 6146–6151. Gardam M, Liang D, Moghadas SM, Wu J, Zeng Q, Zhu H. (2007). J R Soc Interf. 4: 727–734. Halloran ME, Hayden FG, Yang Y, Longini IM Jr, Monto AS. (2007). Am J Epidemiol. 165: 212–221. Hayden FG. (2004). Pediatr Infect Dis J. 23: S262–S269. Horcajada JP, Pumarola T, Martı´nez JA, Tapias G, Bayasz JM, de la Prada M, Garcı´a F, Codina C,
Prophylaxis of Healthcare Workers in an Influenza Pandemic Gatell JM, Jime´nez de Anta MT. (2003). Eur Respir J. 21: 303–307. Inglesby TV, Nuzzo JB, O’Toole T, Henderson DA. (2006). Biosecur Bioterror. 4: 1–10. Jefferson T, Demicheli V, Rivetti D, Jones M, Di Pietrantonj C, Rivetti A. (2006). Lancet. 367: 303–313. Jennings LC, Peiris M. (2006). Inter Med J. 36: 145–147. Lee VJ, Chen MI. (2007). Emerg Infect Dis. 13: 449–457. Lester RT, McGeer A, Tomlinson G, Detsky AS. (2003). Infect Control Hosp Epidemiol. 24: 839–844. Lipsitch M, Cohen T, Murray M, Levin BR. (2007). PLoS Med. 4: e15. Longini IM Jr, Nizam A, Xu S, Ungchusak K, Hanshaoworakul W, Cummings DAT, Halloran ME. (2005). Science. 309: 1083–1087. Loutfy MR, Wallington T, Rutledge T, Mederski B, Rose K, Kwolek S, McRitchie D, Ali A, Wolff B, White D, Glassman E, Ofner M, Low DE, Berger L, McGeer A, Wong T, Baron D, Berall G. (2004). Emerg Infect Dis. 10: 771–776. Lundstrom T, Pugliese G, Bartley J, Cox J, Guither C. (2002). Am J Infect Control. 30: 93–106. Malavaud S, Malavaud B, Sanders K, Durand D, Marty N, Icart J, Rostaing L. (2001). Transplantation 72: 535–537. Moghadas SM, Bowman CS, Ro¨st G, Wu J. (2008). PLoS ONE. 3: e1839. Monto AS. (2003). Vaccine. 21: 1796–1800. Moscona A. (2005). N Engl J Med. 353: 2633–2636. Munoz FM, Campbell JR, Atmar RL, Garcia-Prats J, Baxter BD, Johnson LE, Englund JA. (1999). Pediatr Infect Dis J. 18: 811–815. Nichol KL, Lind A, Margolis KL, Murdoch M, McFadden R, Hauge M, Magnan S, Drake M. (1995). N Engl J Med. 333: 889–893. Nichol KL, Mallon KP, Mendelman PM. (2003). Vaccine. 21: 2207–2217.
82
Nichol KL, Mendelman P. (2004). Virus Res. 103: 3–8. Potter CW. (2001). J Appl Microbiol. 91: 572–579. Regoes RR, Bonhoeffer S. (2006). Science. 312: 389–391. Salgado CD, Farr BM, Hall KK, Hayden FG. (2002). Lancet Infect Dis. 2: 145–155. Sartor C, Zandotti C, Romain F, Jacomo V, Simon S, Atlan-Gepner C, Sambuc R, Vialettes B, Drancourt M. (2002). Infect Control Hosp Epidemiol. 23: 615–619. Saxe´n H, Virtanen M. (1999). Pediatr Infect Dis J. 18: 779–783. Stott DJ, Kerr G, Garman WF. (2002). Occup Med. 52: 249–253. Straetemans M, Bochhols U, Reiter S, Haas W, Krause G. (2007). BMC Public Health. 7: 236–247. Taubenberger JK, Morens DM. (2006). Emerg Infect Dis. 12: 15–22. Toner E, Waldhorn R. (2006). Biosecur Bioterror. 4: 397–402. Webster RG, Peiris M, Chen H, Guan Y. (2006). Emerg Infect Dis. 12: 3–8. Weingarten S, Riedinger M, Bolton LB, Miles P, Ault M. (1989). Am J Infect Control. 17: 202–207. WHO. (2008). http://www.who.int/csr/disease/influenza/ nationalpandemic/en/index.html WHO Avian Influenza. (2008). http://www.who.int/csr/ disease/avian_influenza. WHO Checklist for Influenza Pandemic Preparedness Planning. (2005). http://www.who.int/csr/ resources/publications/influenza/ WHO_CDS_CSR_GIP_2005_4/en/ WHO Writing Group. (2006). Emerg Infect Dis. 12: 81–87. Wilde JA, McMillan JA, Serwint J, Butta J, O’Riordan MA, Steinhoff MC. (1999). JAMA. 281: 908–913. Xiong W, Hollingsworth E, Muckstadt J, Vorenkamp JVL, Lazar EJ, Cagliuso NV Sr, Hupert N. (2007). Infect Control Hosp Epidemiol. 28: 618–621.
1431
83 The Burden of Human African Trypanosomiasis A. Shaw . J. Robays . E. M. Fe`vre . P. Lutumba . M. Boelaert 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1434 2 Magnitude of the Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1434 3 Estimating the DALYs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1435 4 The Economic Burden on Households and Communities . . . . . . . . . . . . . . . . . . . . . . . . . . 1438 5 The Cost-Effectiveness of Controlling HAT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1438 6 What Lessons can be Drawn for other Diseases? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1440 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1440 Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1441
#
Springer Science+Business Media LLC 2010 (USA)
1434
83
The Burden of Human African Trypanosomiasis
Abstract: > Human African Trypanosomiasis (HAT), a once largely forgotten disease, is back on the agenda. A public-private partnership between the World Health Organization (WHO), pharmaceutical companies and international donors succeeded in curbing the recent epidemic, but active transmission is still ongoing in several countries. The burden of the disease in affected individuals is high. Untreated the disease is always fatal and estimates of the Disability Adjusted Life Years (DALYs) lost per premature death range from 25 to 33 years. Even for successfully treated patients, the burden on households and livelihoods is high – between 1.5 and 10 months’ income, even when diagnostics and HAT drugs are provided for free. Costs ranging from $10–17 per DALY averted for case-finding and treatment places the control of T.b. gambiense HAT firmly in the category of highly cost-effective health interventions. List of Abbreviations: DALY, disability adjusted life year; DRC, democratic republic of Congo; HAT, Human African Trypanosomiasis; NGO, Non-Governmental Organization; T.b., > Trypanosoma brucei; WHO, World Health Organization; YLL, years of life lost
1
Introduction
Human African Trypanosomiasis (HAT), also known as “sleeping sickness” is one of the neglected diseases of this world. It is endemic in 36 countries of sub-Saharan Africa where it affects mainly the rural poor. The disease is caused by protozoa of the species Trypanosoma brucei (T.b.) and is transmitted by > tsetse flies. Infection with such parasites eventually leads to death in the absence of appropriate treatment, but the disease presents differently by region. In East Africa, HAT is a zoonotic disease involving domestic animals and wildlife, caused by T.b. rhodesiense, (> East African sleeping sickness) and cattle are often the main reservoir (> disease reservoir in HAT). In West and Central Africa it is caused by T.b. gambiense and transmitted in a man-fly-man cycle. HAT caused by T.b. gambiense (> West African sleeping sickness) accounts for over 90% of all reported cases and has a much more protracted course than the rhodesiense (or East African) type. The disease evolves in two stages, (1) an early stage (> hematolymphatic stage of Human African Trypanosomiasis) when infection is limited to blood and lymph circulation and that is treated with pentamidine or suramin with over 90% cure rates, and (2) the late meningo-encephalitic stage (> meningo-encephalitic stage of Human African Trypanosomiasis) when the parasites have invaded the central nervous system. An arsenicum-derivative, melarsoprol, was until recently the recommended drug for this late stage, but, because of its toxicity and as it was gradually loosing its efficacy, eflornithine is now promoted in this indication for the treatment of the late stage of the gambiense form of the disease (Priotto et al., 2008). Because of the lack of specific symptoms in the early stage, patients usually consult a health professional when the disease is already well advanced and involves the central nervous system. At that point, the patient may have irreversible brain damage, requires the more expensive and toxic drugs and has posed a risk to other members of the community due to his infectiousness to tsetse flies. Therefore, population screening for HAT is the main control strategy for the gambiense form of HAT.
2
Magnitude of the Problem
During the 1940s, 1950s and 1960s large-scale control programmes based on active case detection had succeeded in reducing the incidence of the gambiense form to some 4,000–6,000 reported
The Burden of Human African Trypanosomiasis
83
cases a year during the 1960s (WHO, 2000). There were still regular outbreaks of the acute, rhodesiense, form, with as many as 8,000 cases being reported during a single year in the epidemic in Uganda which started in the late 1970s, but their severity was to some extent mitigated by vector control activities. However, during the 1980s the number of reported HAT cases increased gradually and towards the end of the 1990s disturbing evidence of a large scale resurgence of the gambiense form of the disease began to emerge, as the number of reported cases rose to 45,000 annually, a more than tenfold increase since the sixties. Although this latter number might seem relatively low, when viewed against the other health problems faced by Africa’s people, the clustering of this disease in localized foci where its burden is very heavy, and its potential for epidemic outbreaks meant that this resurgence of sleeping sickness needed to be treated with great seriousness. In addition the reported cases were mainly found by passive surveillance rather than through active surveillance and were thought to represent only the tip of a large iceberg. By the late 1990s there was a widespread general awareness that this largely forgotten disease had made a frightening comeback and that a hidden epidemic of sleeping sickness had been ravaging African countries. Fortunately, this awareness was also translated into a major investment in the control of HAT in recent years. Bilateral donors stepped up their aid (Lutumba et al., 2005) and a partnership between WHO, private pharmaceutical companies, Non-governmental organizations (NGOs), bilateral donors and the health services of the countries affected led to an important increase of HAT control activities and a curbing of the case load. The recent trend in the Democratic Republic of Congo (DRC) is shown in > Figure 83-1. Nonetheless intensive transmission of the disease is still taking place in several areas of Central Africa. The most recent figures published by WHO (2006) for gambiense HAT were 17,036 reported cases in 2004, which were extrapolated to an estimated worldwide annual caseload of 50,000–70,000. Sixty-one percent of the reported cases (10,369/17,036) were from a single country: DRC. Sudan and Angola each reported more than 1,500 cases, while 50–1,500 cases per year were reported in the Central African Republic, Chad, Coˆte d’Ivoire, Guinea and Uganda. Burkina Faso, Cameroon, Equatorial Guinea, Gabon and Nigeria each reported less than 50 cases (> Figure 83-2).
3
Estimating the DALYs
Published global estimates of the annual burden of this disease have ranged between 1.5 and 2.0 million disability-adjusted life years (DALYs) per annum (WHO, 2004), based on the number of reported cases and using the same weightings for the burden associated with both the chronic and acute forms. A number of recent studies have provided the basic data needed for a more accurate global estimate. For the gambiense form, estimates of the DALY burden in affected individuals have been made by interviewing hospital patients diagnosed with the disease. In untreated individuals in Southern Sudan, the number of years of life lost (YLL) due to premature mortality from HAT were estimated at 33 per death (D. McFarland, personal communication, based on work in Trowbridge et al., 2001), in Angola the figure estimated was 30 years per death (Schmid et al., 2004) and in the Democratic Republic of Congo (DRC) a figure of 27 years per death was obtained (Lutumba et al., 2005). The long term sequelae, which can include neurological impairment, make the morbidity component significant as well, even for treated patients. For rhodesiense, an initial estimate (Politi et al., 1995) based on data from Uganda, estimated the number of DALYs incurred per premature death to be 25,
1435
1436
83
The Burden of Human African Trypanosomiasis
. Figure 83-1 Number of new cases reported by the national control program in Democratic Republic of Congo (1926–2004). NC = new case. This graph shows how the number of new cases of Human African Trypanosomiasis (HAT) gradually declined after a peak in 1930 to reach a historical low in 1960. Then, despite ongoing control activities, the number of cases slowly increased to reach a new peak in 1999, from that moment it was brought down again by an intensified control program
a more recent estimate (Odiit, 2003) puts them at 31. Additionally, the acute nature of T.b. rhodesiense results in a high burden per affected case, making the morbidity component of the T.b. rhodesiense DALY high. These estimates show the number of DALYs caused by both forms of the disease to be substantial, and well above those incurred by other tropical diseases such as dengue and filariasis which receive a great deal of international attention. This reflects, firstly, the fact that HAT is nearly always fatal in untreated individuals and secondly that the majority of the people affected are economically active adults. For example data on rhodesiense from Uganda showed 25% of cases occurring in those aged 20–29 years and 60% of those affected being in the 10–39 year age group (Fe`vre et al., 2008), for gambiense in DRC (Lutumba et al., 2007a) very similar percentages were found (21 and 58%,respectively). These DALY estimates provide us with an overall view of the impact of the disease on affected individuals, but it should be noted that few DALY estimates have been made for HAT and that those reported are deterministic estimates, referring to the average impact per patient. Furthermore, the potential long term impacts of HAT on treated and cured patients have not been investigated for either form of the disease, so the DALY estimates do not include sequelae. Having established what the losses in infected individuals are likely to be, extrapolating these to a national or continental scale is more problematical. So far, only one published study (Odiit et al., 2005) has addressed this directly. Using an epidemiological model, applied to data on the number of patients presenting in first and second stage of the rhodesiense form of disease in south-eastern Uganda, the study concluded that for every reported death from the disease, a further twelve remained unreported; this work has been validated in other parts of Uganda (Fe`vre et al., 2005), demonstrating that in a localized epidemic with 500 reported cases,
The Burden of Human African Trypanosomiasis
83
. Figure 83-2 Annual number of Human African Trypanosomiasis (HAT) cases reported between 2000 and 2005. This map shows how Democratic Republic of Congo, Sudan and Angola each reported more than 1,500 cases, while 50–1,500 cases per year were reported in the Central African Republic, Chad, Coˆte d’Ivoire, Guinea and Uganda. Burkina Faso, Cameroon, Equatorial Guinea, Gabon and Nigeria each reported fewer than 50 cases (Adapted from World Health Organization website: http://www.who.int/trypanosomiasis_african/disease/en/index.html)
approximately 300 additional cases died undiagnosed in the community. For gambiense, where the course of the disease is much more extended and more variable, and where a large proportion of the reported cases is usually found by exhaustive population screening, such approach would not be applicable. Work by Robays et al. (2004) gives insight in the proportion of HAT patients who are likely to missed out in active screening exercises for gambiense HAT. This study estimated that between 40 and 50% of patients in the community are missed by the campaign, because some individuals do not present themselves for the initial screening and because of the poor sensitivity of some of the tests used to confirm the presence of the parasite. Developing an effective methodology for estimating the degree of under-reporting for gambiense disease remains a challenge. WHO estimated the total population at risk to be 60 million in 1998 (WHO, 1998). Further work is being undertaken, remapping the foci of the disease and refining this calculation. At the start of the 1997–2005 epidemic, when most cases were passively detected, WHO estimated that only about one in ten HAT patients had been found and correctly diagnosed – however, after more than 5 years of active case-finding, it is hoped that a far higher proportion of patients have been found and treated. Thus, while work on recalculating the global burden of HAT in terms of DALYs is ongoing, it looks as though, despite the falling number of reported cases, the figure is likely to range between 0.5 and 1.5 million DALYs, reflecting new research showing a higher burden per
1437
1438
83
The Burden of Human African Trypanosomiasis
affected individual and depending on the results of evidence-based estimates of the proportion of unreported cases. Recently, Lutumba et al. (2007a) have pointed to the risks when using DALYs for priority setting in HAT. At a global level, the absolute figures do not reflect the major impact the disease can have on local communities and regions, because the disease has such a clustered occurrence. Secondly, it is vital that declining trends in case numbers or DALYs should not lead to a reduction in resource allocation for HAT control because there are many documented examples of resurgence after abrupt cessation of active screening campaigns (Lutumba et al., 2005; Moore et al., 1999; Moore and Richer, 2001; Van Nieuwenhove et al., 2001) and indeed the increase in cases leading to the current epidemic closely mirrored the decline in active surveillance.
4
The Economic Burden on Households and Communities
The key to the high burden this disease places on affected households and communities lies not just in the difficulties individuals face in obtaining a correct diagnosis and thus the treatment required to prevent a fatal outcome, but also in its focal nature. Within these foci the prevalence can be high, often around 1% in the absence of active screening. If intense transmission goes unnoticed in a focus, the prevalence rates can rise relatively rapidly, sometimes over half the population of certain villages have been found affected. Thus the burden of this disease falls very heavily on a few locations, as argued in Lutumba et al. (2007a). As well as often being the active adults in a household, HAT patients also require very high levels of care, placing a further burden on the household. There have been few attempts to quantify the full costs borne by households with HAT patients which include their care at home and during hospital treatment, seeking a diagnosis, lost income, medical fees, transport, etc. In the Republic of Central Africa, the cost to households containing HAT patients who were correctly diagnosed and treated came to an amount equivalent to just under 1.5 months’ household income from agriculture (Gouteux et al., 1987). A recent study in DRC (Lutumba et al., 2007a), where the direct costs of screening and treatment are fully subsidized by the HAT program, undertook a detailed calculation of household costs for diagnosed and treated patients; these came to 5 month’s household income on average, but rose to over 10 months’ income for patients with complications. Just seeking a diagnosis can be both time-consuming and costly: a study in a rhodesiense area of Uganda showed that over 70% of patients had to make three or more visits to a health unit before being correctly diagnosed and just over half had to sell agricultural produce to pay for health care costs (Odiit et al., 2004). Lastly, HAT is overwhelmingly a disease of isolated rural populations and within these populations it is the poorer families who have the most difficulty in obtaining a correct diagnosis for their affected family members and on whom the burden of care and the expenditures needed to deal with the disease weigh most heavily.
5
The Cost-Effectiveness of Controlling HAT
A number of approaches can be used alone or in various combinations to control HAT – finding and treating affected individuals, thus reducing the human reservoir of the disease, dealing with the animal reservoir and controlling the vector.
The Burden of Human African Trypanosomiasis
83
For gambiense disease, active case-finding and treatment is essential to reduce transmission and to find and treat affected individuals. The costs of this approach have been estimated in a number of studies (Lutumba et al., 2005; Shaw and Cattand, 2001; Trowbridge et al., 2001; WHO, 1998 – which provides templates showing how the costs could be calculated for various screening strategies). The costs of screening populations range around $1–1.50 per person. The costs of treatment are usually cited as around $100 for first stage patients and $300 for second stage patients treated with melarsoprol, increasing to over $600 if eflornithine is used (WHO, 1998). More recent estimates by Robays et al. (2008) in Angola put these figures at a cost of $604 for melarsoprol and $844 for eflornithine treatment. The cost is very sensitive to assumptions made about the cost of hospitalization (Shaw and Cattand, 2001; WHO, 1998). Several attempts to calculate the cost of a patient-day in sleeping sickness treatment centers have been made and these show the costs to be highly variable – estimated at around $2 in Uganda (Odiit, 2003), while a more detailed analysis showed them to be $10 in DRC and $66 in a well-equipped treatment centre in Angola (Schmid et al., 2004). Overall, looking at the breakdown of costs for case-finding and treatment (Shaw and Cattand, 2001, WHO, 1998), it is clear that once the prevalence exceeds 1%, hospitalisation costs become the main component of overall costs. Prolonged hospitalisation also places a very heavy burden on households. The newly developed 10-day melarsoprol schedule thus represents a very welcome improvement, making it possible to substantially reduce the cost of treatment, for example from $12 to 7 per DALY averted in DRC (Schmid et al., 2004). The cost-effectiveness of case-finding and treatment approaches has been calculated in several studies (Lutumba et al., 2007b; Politi et al., 1995; Schmid et al., 2004; Shaw and Cattand, 2001; Trowbridge et al., 2001; WHO, 1998). In southern Sudan, the cost per DALY averted was estimated at $10 for periodic screening and $17 for emergency intervention after a 9-year interval (Trowbridge et al., 2001). In DRC, a similar figure of $17 per DALY averted was calculated for active case-finding and treatment using mobile teams (Lutumba et al., 2007a). A range of situations was modeled (Shaw and Cattand, 2001), showing that for most situations the cost fell below the $25 threshold of good value for money. This thus places HAT control firmly in the category of highly cost-effective interventions, comparing favorably with immunization programmes and usually cheaper, for example, than malaria control using treated bed-nets (Goodwin et al., 2000). However, the benefit of active screening for gambiense goes beyond simply finding and treating existing patients, to reducing the size of the human reservoir, thus preventing new infections. Accurately quantifying this component would require epidemiological modeling. There is ample historical evidence to show that active screening maintains a low prevalence, so that these costs per DALY averted would be even lower if the effect in lowering the incidence of the disease were taken into account. A limited number of studies have looked at the economic aspects of > vector control in HAT. Vector control is usually undertaken alongside the above measures to find and treat affected people, a summary of costs can be found in Shaw (2004). The costs of vector control to reduce the transmission of HAT have been further discussed by Gouteux and Sinda (1990), Lancien (1991) and Laveissie`re et al. (1994). Simarro et al. (1991) compared different combinations of both active screeningandvectorcontrol.Shaw(1989)found > activecasefindingtobemorecost-effectivethan vector control for the gambiense form of HAT. For rhodesiense disease, recent research (Welburn et al., 2001) has demonstrated that cattle are the main reservoir of the disease to a far greater extent than was previously thought. This has opened up the possibility of treating cattle in order to control the disease in humans and this approach is being adopted in Uganda. From the economic point of view this involves a
1439
1440
83
The Burden of Human African Trypanosomiasis
very interesting scenario. Since the trypanocides used to deal with the pathogens affecting people will also attack those which affect animals, the health of the livestock population will be improved, leading to monetary benefits which together with the costs saved by preventing the disease in people, are likely to outweigh the costs of the control strategy (Shaw, 2004). Thus a net financial benefit is realized, without even considering DALYs and through healthier livestock, people’s livelihoods as well as their own health is improved. Controlling sleeping sickness, a disease of poor and marginalized populations, thus offers a highly cost-effective opportunity for poverty alleviation.
6
What Lessons can be Drawn for other Diseases?
The global burden of this disease is not so important in absolute terms, however, as HAT is a very focal disease, it has serious consequences at the local level, disrupting entire communities. DALY ranking at a global or even national level might lead to misguided priority setting in such a case. Secondly, when evaluating the cost effectiveness of interventions in infectious diseases it is important to realize that control efforts can have an effect on future transmission and thus the benefits of preventing epidemics or controlling them at an early stage are considerable.
Summary Points By the year 2000 Human African Trypanosomiasis (HAT), a largely forgotten disease, had made a frightening comeback.
The burden of the disease in affected individuals is high. Untreated it is always fatal and
estimates of the DALYs lost per premature death range from 25 to 33 years. The morbidity component is also high. Due to the inherent difficulties of diagnosing sleeping sickness, under-reporting has always been problematic. At the start of the present epidemic, it was thought that for every reported case, ten remained undiagnosed and untreated. Recent research in endemic rhodesiense areas and in areas where active surveillance for the chronic gambiense form of the disease occurs indicates that even in active population screening campaigns 40–50% of patients are not found. In 1999 the annual global burden was estimated at two million DALYs, based on the number of reported cases. Now it looks as though, despite the falling number of reported cases, the figure is likely to range between 0.5 and 1.5 million DALYs, reflecting new research showing a higher burden per affected individual and depending on the results of evidence-based estimates of the proportion of unreported cases. HAT affects above all economically active adults, and patients require a lot of care from the other adults in the family. Seeking a diagnosis and obtaining treatment is costly and timeconsuming. Thus the burden on households and livelihoods is high – between 1.5 and 10 months’ income, even when diagnostics and HAT drugs are provided for free and patients have been successfully treated with no long term after-effects from the treatment or the disease. Costs ranging from $10 to 17 per DALY averted for case-finding and treatment of chronic sleeping sickness places the control of HAT firmly in the category of highly cost-effective
The Burden of Human African Trypanosomiasis
83
health interventions, well below the $25 threshold of “very good value for money.” This cost estimate would be even lower if the effect in reducing the human reservoir of the disease, and thus preventing future epidemics, were taken into account.
Appendix Key Facts for Human African Trypanosomiasis (HAT)
HAT is transmitted by tsetse flies. There are two forms of the disease. In East Africa HAT presents as an acute syndrome
caused by T.b. rhodesiense, and is maintained by an animal reservoir. West African HAT is caused by T.b. gambiense, has a much more protracted course than East African HAT and is transmitted in a man-fly-man cycle. The role of an animal reservoir still needs to be clarified. There are two stages of the disease, an early stage (hemolymphatic stage), with few or aspecific symptoms, and a late (meningo-encephalitic) stage. Main symptoms of HAT are neuro-psychiatric, including behavioral problems, personality disorders, sleep disturbances and coma. It almost invariably results in death, if left untreated. Treatment of patients in the meningo-encephalitic stage requires more expensive and toxic drugs. Melarsoprol, still the most commonly used drug for second stage, kills between 5 and 10% of the patients. The only alternative, eflornithine, is expensive and difficult to administer. Ninety percent of HAT cases in the world are caused by T.b. gambiense. T.b. gambiense often occurs in epidemics. Active case finding conducted by mobile teams has been the cornerstone of the control of HAT since colonial times. If treated at a late stage, sequelae, mainly persistence of personality disorders and decreased mental capacities, are very frequent and pose a huge burden on the community long after the HAT epidemics are brought under control.
References Fe`vre EM, Odiit M, Coleman PG, Woolhouse ME, Welburn SC. (2008). BMC Public Health. 8: 96. Fe`vre EM, Picozzi K, Fyfe J, Waiswa C, Odiit M, Coleman PG, Welburn SC. (2005). Lancet. 366: 745–747. Goodwin C, Coleman P, Mills A. (2000). Economic Analysis of Malaria Control in Sub-Saharan Africa. Global Forum for Health Research, Geneva, Switzerland. Gouteux JP, Bansimba P, Noireau F, Fre´zil JL (1987). Med Trop. 47: 61–63. Gouteux JP, Sinda D. (1990). Trop Med Parasitol. 41: 49–55. Lancien J. (1991). Ann Soc Belg Me´d Trop. 71 (Suppl. 1): 35–47.
Laveissie´re C, Gre´baut P, Lemasson J. (1994). Les communaute´s rurales et la lutte contre la maladie du sommeil en foreˆt de Coˆte-d’Ivoire, WHO/TRY/94.1, Gene`ve. Lutumba P, Makieya E, Shaw A, Meheus F, Boelaert M. (2007a). Emerg Infect Dis. 13: 248–254. Lutumba P, Meheus F, Robays J, Miaka C, Kande V, Buscher P, Dujardin B, Boelaert M. (2007b). Emerg Infect Dis. 13: 1484–1490. Lutumba P, Robays J, Miaka mia BC, Mesu VK, Molisho D, Declercq J, Van der Veken W, Meheus F, Jannin J, Boelaert M. (2005). Emerg Infect Dis. 11: 1382–1389.
1441
1442
83
The Burden of Human African Trypanosomiasis
Moore A, Richer M. (2001). Trop Med Int Health. 6: 342–347. Moore A, Richer M, Enrile M, Losio E, Roberts J, Levy D. (1999). Am J Trop Med Hyg. 61: 315–318. Odiit M. (2003). The Epidemiology of Trypanosoma brucei rhodesiense in Eastern Uganda. Unpublished PhD thesis, University of Edinburgh, pp. 210. Odiit M, Coleman PG, Liu WC, McDermott JJ, Fe`vre EM, Welburn SC, Woolhouse ME. (2005). Trop Med. Int Health. 10: 840–849. Odiit M, Shaw A, Welburn SC, Fe`vre EM, Coleman PG, McDermott JJ. (2004). Ann Trop Med Parasitol. 98: 339–348. Politi C, Carrin G, Evans D, Kuzoe FA, Cattand PD. (1995). Health Econ. 4: 273–287. Priotto G, Pinoges L, Fursa IB, Burke B, Nicolay N, Grillet G, Hewison C, Balasegaram M. (2008). BMJ. 336: 705–708. Robays J, Bilengue MM, Van der SP, Boelaert M. (2004). Trop Med Int Health. 9: 542–550. Robays J, Raguenaud ME, Josenando T, Boelaert M. (2008). Trop Med Int Health. Feb; 13(2): 265–271. Schmid C, Shaw A, Santercole C, Kwete J, Lutumba P, Burri C. (2004). In: Schmid C. (2004) 10-Day Melarsoprol Treatment of Trypanosoma brucei gambiense Sleeping Sickness. PhD Thesis, University of Basel. Basler Schnelldruck, Basel.
Shaw AP. (1989). Ann Soc Belg Med Trop. 69 (Suppl. 1): 237–253. Shaw AP, Cattand P. (2001). Med Trop (Mars). 61: 412–421. Shaw APM. (2004). In: Maudlin I, Holmes PH, Miles MA (eds.). The Trypanosomiases. CAB International, Wallingford, UK, pp. 369–402. Simarro PP, Sima FO, Mir M, Mateo MJ, Roche J. (1991). Bull World Health Organ. 69: 451–457. Trowbridge M, McFarland D, Richer M, Adeoye M, Moore A. (2001). Am J Trop Med Hyg. 62 (3 Suppl.): 312. Van Nieuwenhove S, Betu-Ku-Mesu VK, Diabakana PM, Declercq J, Bilenge´ CM. (2001). Trop Med Int Health. 6: 335–341. Welburn SC, Fe`vre EM, Coleman PG, Odiit M, Maudlin I. (2001). Trends Parasitol. 17: 19–24. WHO. (1998). World Health Organ Tech. Rep. Ser. 881: I-114. WHO. (2000) WHO Report on Global Surveillance of Epidemic-prone Infectious Diseases. WHO Department of Communicable Disease Surveillance and Response, WHO/CDS/CSR/ISR/2000.1. WHO. (2004) Wkly Epidemiol Rec. 79: 297–300. WHO. (2006) Wkly Epidemiol Rec. 81: 71–80.
84 The Economic Burden of Malaria in Nigeria and Willingness to Pay A. Jimoh 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1444
2 2.1 2.2 2.3 2.4 2.5
Concept of Economic Burden of a Disease and Willingness to Pay . . . . . . . . . . . . . 1445 Types of Economic Burden . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1445 Three Approaches to Quantifying the Burden . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1445 Production Function Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1446 Cost-of Illness Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1447 Willingness to Pay Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1447
3 3.1 3.2 3.3
Extent of Economic Burden of Malaria in Nigeria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1448 Limitations of Leighton and Foster Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1449 The Study by Onwujekwe and Others . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1449 WHO Funded Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1450
4 4.1 4.2 4.3
Willingness to pay for Malaria Control and Treatment Programs . . . . . . . . . . . . . . 1453 Onwujekwe’s WTP Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1453 Onwujekwe and Others’ WTP Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1453 Jimoh and Others’ WTP Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1454 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1456
#
Springer Science+Business Media LLC 2010 (USA)
1444
84
The Economic Burden of Malaria in Nigeria and Willingness to Pay
Abstract: This chapter reviews available evidence on the size of the economic burden of > malaria in Nigeria and the citizens’ > willingness to pay (WTP) for malaria control and treatment programs. It found that the direct and indirect economic burden may be in excess of 13% of the GDP. Furthermore, the review suggests that the average amount that people are willing to pay for malaria control and treatment programs are below their ruling market prices and that economic status (wealth and income) is one of the major factors determining Nigerians’ WTP for the treatment or total elimination of malaria. Consequently, it recommends the promotion of universal access for the treatment of malaria in Nigeria by abolishing or reducing payment for malaria treatment at the point of need through insurance-based financing arrangement. It observes that the National Health Insurance Scheme that is currently being put in place is a step in the right direction but that progress has been very slow. List of Abbreviations: AAAS, American Association for the Advancement Of Science; BWFU, CBN, Central Bank Of Nigeria; COI, > cost of illness; CV, > contingent valuation; DW, Durbin–Watson; FGD, focus group discussion; FMoH, Federal Ministry of Health, Nigeria; GDP, > gross domestic product; HH, household; HIV/AIDS, acquired immune deficiency syndrome; ITN, > insecticide-treated net; M, million; N, Naira, Nigeria local currency; NGOs, non-governmental organizations; PF, > production function approach; PHCs, primary health care centers; UNICEF, United Nations children’s fund; WHO, Word Health Organization; WTP, willingness to pay > binary-with-follow-up;
1
Introduction
Nigeria covers a total area of about 923,770 sq km and a land mass of about 910,770 sq km. It shares boundaries with Niger Republic and Chad Republic in the north, Benin Republic in the west and the Cameroon in its east. Its southern border is with the Atlantic Ocean (Federal Office of Statistics, FOS, 1995, p. 1). Its population was put about 140 million by 2006 Census. On this account, it is the most populous nation in Africa and the ninth most populous in the world. Nigeria has a tropical climate with three broad types of vegetation zones, namely, the forest zones in the southern coastal areas, the savannah zone in the middle areas and the grass-land zones in the northern-most areas of the country. Corresponding to these vegetation zones are its three major malaria epidemiological zones, namely, the forest, the savannah and the grass-land zones. Like most parts of Africa, malaria is a serious problem in Nigeria which is noted for its high malaria prevalence rate (AAAS, 1991; Federal Ministry of Health 1992, 2001; Gallup and Sachs, 2001; Onwujekwe et al., 2000; WHO/UNICEF, 2003; etc.). A number of studies have quantified the economic burden of malaria in Nigeria (Jimoh, 2004; Jimoh et al., 2007; Onwujekwe et al., 2000). Similarly, some studies have measured Nigerians willingness to pay for some components of malaria control programs, malaria treatments and malaria eradication programs (Onwujekwe, 2001; Onwujekwe et al., 2004; etc.) but all these are scattered in the literature. It is desirable to bring the results of these studies together to provide a bird-eye view of the current state of our knowledge of the issues involved as well as to assist policy makers in understanding these issues and how best to resolve outstanding problems. Towards these ends, what follows consists of five Sections: Concept of Economic Burden of a Disease and Willingness to Pay; Extent of Economic Burden of Malaria in Nigeria; Willingness to Pay for Malaria Control and Treatment Programmes; and Policy Implications, Summary and Conclusions.
The Economic Burden of Malaria in Nigeria and Willingness to Pay
2
84
Concept of Economic Burden of a Disease and Willingness to Pay
In general, the major effects or end results of an illness may be morbidity or mortality. In a morbid state, the victim may seek treatment and therefore incur health care > costs. Beside this, normal economic activities of the victim may cease totally or hours of work are cut down or productivity is reduced. Furthermore, victims of a disease suffer physical and psychological pains in the morbid state. Because of these unpleasant consequences, people under-take preventive expenditures and modify their habits, travel plans, and other economic and social activities to remain safe from the disease. When an illness results in death – mortality – the entire human capital investment in the victim is lost as well as the associated potential future outputs that are foregone. Furthermore, grieving relations under go mental pains and other knock-down effects, sometimes it may result in the dissolution of the household.
2.1
Types of Economic Burden
In general, the above mentioned types of burden are classified into three, namely, the direct cost, the > indirect cost and intangible cost. The direct costs include health care costs (cost of preventive actions, diagnosis, consultation, drugs, cost of transportation and other incidental expenses for the victim and the caretaker, etc.). The indirect costs are values of lost output due to morbidity (for the victim and caretaker) as well as output lost to mortality. This will include items such as the values of lost time (of victims and caretakers) while seeking care and value of reduced output in any form it may come. Though intangible burdens also have economic values because they are difficult to measure, only direct and indirect costs are usually summed up as the economic burden of the disease in question. These costs are shared among the households, the firms and public institutions (governments, NGOs, insurance institutions, etc.) depending on the health financing arrangement that is in place. But in the main, intangible costs are borne by the individual victims and their households. The nature of the malaria burden is diagrammatically represented in > Figure 84‐1 In that Figure, the second line shows the three possible states with regard to malaria that come with costs. Starting with the first box it shows that before the morbid state, some individuals may undertake malaria prevention activities and behavior which come with costs. The other two boxes show that malaria attack may result in morbidity or mortality and these have their costs. In the third line the cost of illness box collects together all the costs from the three sources and these represent the burden of the malaria. In the fourth line, the figure shows the three types of malaria economic burden (costs). In the last line of the figure, the parties that bear the costs are indicated.
2.2
Three Approaches to Quantifying the Burden
There are three major approaches to quantifying the economic burden of malaria (or any disease for that matter). These are the Production Function (PF) approach, the Cost-of-Illness (COI) approach and Willingness-to-Pay (WTP) approach (WHO/AFRO, 2001).
1445
1446
84
The Economic Burden of Malaria in Nigeria and Willingness to Pay
. Figure 84‐1 An overview of the conceptual framework for the COI approach. It is a diagrammatic representation of the nature of the malaria burden. It shows the three possible health states with regard to malaria and the corresponding three types of economic burden (costs) as well as the parties that bear the costs
2.3
Production Function Approach
The PF approach measures a disease burden by determining the effect of the disease on the output production function. It does this by establishing an empirical statistical relationship between the output level (as the dependent variable) and a set of explanatory variables (which usually include the major inputs to the production process and other exogenous shocks). The inputs to the production process will usually include capital stock, labor, among other factors of production while the exogenous shocks may be weather, policy variables, and a measure of a disease prevalence, among others. The production function so specified and estimated is then calibrated at the historical values of the disease prevalence to determine the impact of the disease on output levels. Studies that have employed the PF approach to determine the output effects of malaria include Audibert (1986), Gallup and Sachs (2001), Jimoh (2004), Wang’ombe and Mwabu (1993), etc. The major challenge for studies of this variety is getting a measure of the malaria variable that truly reflects the degree to which malaria has an impact on output (Chima et al., 2003, p. 14–15). Audibert et al. (1999) suggest that the extent to which the density of parasitemia in the blood exceeds a threshold of 500 parasites per micro liter may be a good measure of the malaria variable that is appropriate for PF study employing cross-sectional data. But even then, the coexistence of other illnesses with malaria may result in positive bias (over-statement of) in the estimated output effect of malaria in such studies. Similarly, Gallup ands Sachs (2001) suggest that an index of malaria prevalence that reflects the extent of malaria risk – based on historical data on malaria prevalence rates – may be a good
The Economic Burden of Malaria in Nigeria and Willingness to Pay
84
measure of the malaria variable for a PF study using cross-country data. This was confirmed by Jimoh (2004; 2005) which suggests that an index of malaria prevalence may be a good measure of the malaria variable in a PF study using time series data. But as with all PF studies on the output effect of malaria, this does not overcome or reduce the probability that the output-reducing effect of malaria is over-stated as there is no mechanism for netting out the effects of other diseases that may coexist with malaria.
2.4
Cost-of Illness Approach
The COI approach measures the economic burden of a disease by measuring and summing up the total direct cost of treating disease and the indirect costs of lost outputs, resulting from malaria morbidity (and mortality), value of time lost in the course of obtaining treatment as well as the cost of protecting oneself against disease. Studies that have employed the COI approach to measure the economic burden of malaria are numerous (Ettling and McFarland, 1992; Ettling and Shepard, 1991; Omwujekwe et al., 2000; Shepard et al., 1991; etc.). These have typically relied on surveys or Focus Group Discussions (FGD) to determine the average number of malaria episodes per period, the average cost of treatment per episode, average time lost to morbidity and in the process of seeking care, average transport cost and other incidental expenses – for the malaria victims and the caretakers. The major challenge to this type of studies is obtaining high-quality data. First is the problem of ensuring that the respondents do not lump together many other illnesses – especially those that have similar symptoms with malaria – along with true malaria thereby overstating the number of malaria episodes that they experienced as well as the corresponding costs of treatment. Second is the problem of the timing of the survey and its time horizon. To aid memory recall by respondents, the time horizons do not typically cover more than a maximum of 1 month to the date of interview. But the covered period may not reflect the seasonal variations in the malaria disease episodes thereby resulting in either under- or over-statement of average number of malaria episodes. The same can be true of average cost of treatment if the time covered by the survey data do not reflect seasonal variations in the distribution of episodes of more complicated and severe malaria that come with higher treatment costs. Third is the problem associated with valuing time lost by malaria victims (and their caretakers). Furthermore, there is the likelihood of under-stating the economic burden of more severe and complicated malaria as the long-term effects are usually not accounted for e.g., the long-tern effects on the intellectual development of child-victims, on the pregnant women and the products of such pregnancies, etc. (Chima et al., 2003, p. 9–11).
2.5
Willingness to Pay Approach
The WTP approach to measuring the burden of a disease is a method of contingent valuation (CV). It measures the burden of a disease by the amount people are ready to pay to avoid being a victim of the disease. Such individual valuation is expected to reflect both the direct and indirect costs of the disease to the respondents as well as a valuation of the intangible costs. Of course there are many other uses of the WTP approach especially in the valuation of
1447
1448
84
The Economic Burden of Malaria in Nigeria and Willingness to Pay
health-care programs. Studies employing the WTP approach to value health outcomes or health programs include Donaldson (1990), Johannesson et al. (1991), Morrison and Gyldmark (1992), Onwujekwe et al. (2004), etc. For a more comprehensive list and review of such studies see Klose (1999). Very recently, Jimoh et al. (2007) employed the values that respondents stated that they are willing to pay to avoid a disease to estimate the value they place on the burden of the disease. However, there are some challenges to the use of this approach in measuring the economic burden of a disease. First, because WTP involves asking individuals to state the maximum amount that they would be willing pay to prevent an undesirable health outcome (e.g., illness), it is desirable that the relevant questions be correctly framed so as to be close to the real life market pricing situations that the respondents face. In this regard, the issues that are of prime importance are the > elicitation format, scenario presentation, and the payment methods. To enhance the validity of the stated WTP values, the questioning format – whether openended, binary-with-follow-up (BWFU) or bidding game, etc. – must be appropriate for the nature of the problem being studied. For instance, for a typical African market situation where haggling is the rule, a bidding game may better resemble the pricing situation that respondents face every day (Onwujekwe, 2001, p. 147). On the contrary, close-ended binary elicitation format may be more suitable for other markets where price-tagging is the normal mode of pricing. Notwithstanding the apparent simplicity of the logic behind the choice of appropriate elicitation format, the literature on this is less than being conclusive. For instance, Brown et al. (1996), Frykblom (1997) and Loomis et al. (1997) found that actual outcomes were similar regardless of the elicitation format adopted. However, there appears to be some agreement that there could be starting-point bias or range bias with the bidding game and payment card format respectively (Klose, 1999, p. 100–103). As regard scenario specification and payment methods, the essential requirement is that the features and characteristics of the product being offered should be clearly spelt out to the respondents to enable an informed, valid and consistent response (see Klose, 1999, p. 102–105 for a more detailed discussion of this). A unique feature of WTP approach to valuing the disease burden is its ability to measure the intangible costs which none of the other two approaches can measure. It is worth noting that the COI and WTP approaches are by their very nature implemented through crosssectional (micro) studies.
3
Extent of Economic Burden of Malaria in Nigeria
In Nigeria malaria imposes great burden. It is a leading cause of mortality and morbidity especially among children and pregnant women (Ejezie et al., 1991; FMoH, 2001). Leighton and Foster (1993) was probably one of the earliest efforts to measure the economic burden of malaria in Nigeria using a combination of FGD and secondary data. Two possible scenarios – low and high prevalence regimes – were investigated. They found that the malaria episodes per adult per year ranges between 1 and 4 and the number of work days missed per adult episode ranges between 1 and 3. Furthermore, they reported that the number of adult work days missed per child episode of malaria for caretaking ranges between 1 and 3. Relying on official statistics on wages by sectors, Leighton and Foster (1993) estimated average daily wages to be N30.8 per day in the agriculture sector, N23.1 per day in the service
The Economic Burden of Malaria in Nigeria and Willingness to Pay
84
sector and N53.1 per day in the industrial sector. These wage rates were then used to value the lost work days of malaria victims and their caretakers which was estimated to range between 1 and 5% of the national output (GDP). The value of reduced output (in days of work but with lower productivity) was similarly estimated to range between 0.1 and 0.4% (Leighton and Foster, 1993, p. 58–61). This implies that the indirect cost of malaria (i.e., the value of lost output) in Nigeria in 1991 ranged between 1.1 and 5.4% of the GDP. The study also found that treatment costs for malaria consumed between 4 and 13% of the households annual incomes in 1991 (Leighton and Foster, 1993, p. 76).
3.1
Limitations of Leighton and Foster Study
Results from Leighton and Foster (1993) are at best regarded as preliminary. First, because the study was based on FGD, its estimates of key parameters need to be validated with surveybased data before its outcome can be accepted as reliable. Second, it is well known that given the pervasiveness of malaria episodes in Nigeria people undertake various protective actions which come at some costs. Therefore, the failure of the study to cover protection expenditures means that the malaria burden was under-stated to that extent. Furthermore, issues could be raised on the method adopted for the valuation of lost work time. It is most likely that the official statistics on average would not reflect the marginal wages at the periods of malaria episodes yet those are the appropriate wage rates to use (Sauerborn et al., 1991). Also, the study did not cover the mortality costs of malaria. However, this may not be a major draw back because it is generally accepted that while malaria has some costs associated with mortality in areas where it is endemic, these are not profound because adults seldom die of malaria i.e., the indirect costs associated with malaria mortality are likely to be minimal – unlike a disease like HIV/AIDS (McGregor, 1988). But given that it is known that mortality rate is however high among Nigerian children below the age of 5 years (Ejezie et al., 1991; FMoH, 2001), the seemingly low cost of mortality may become material if the output lost during grieving and bereavement period of victims’ relations and well wishers are accounted for – as relations and well wishers would normally not engage in economic activities in such periods, resulting in lost outputs. Despite these limitations the study filled in some very important gaps in our knowledge as at 1993 as it invited further studies.
3.2
The Study by Onwujekwe and Others
Another study that examines the malaria burden in Nigeria is Onwujekwe et al. (2000). The study conducted a survey of households in five rural communities in an eastern state of Nigeria (Enugu State) in 1998. It shows that malaria treatment cost per household per month was N174.35 ($1.84) and that the value of man-days lost was about N121.23 ($1.28) per household per month giving an average total cost of N295.58 ($3.11) per household per month (Onwujekwe et al., 2000, p. 152–154). With an average household size of 7.07 persons reported in the study (p.154), these translated to a per capita treatment cost of N24.66 ($0.26), indirect cost of N17.15 ($0.18) and a total cost per capita of N41.81 ($0.44) per month. This amounts to a per capita total cost of N501.72 ($5.28) per annum. If this data were to be representative of the entire country, with a total estimated population of 108 million in 1998,
1449
1450
84
The Economic Burden of Malaria in Nigeria and Willingness to Pay
this translates to a total cost of N54,186 million for Nigeria which constitutes about 2.0% of the country’s GDP of N2,708.4 billion in 1998 current prices (see Central Bank of Nigeria (CBN), 1999 for GDP statistics). However, since only rural communities were sampled, these results cannot be representative of the country because the average treatment costs are much higher in the urban areas. Similarly, the value of man-hour in the urban areas is likely to be higher. Like Leighton and Foster (1993) before it, Onwujekwe et al. (2000) did not investigate protection and mortality costs with the attendant under-statement of the economic burden. Similarly, the value of lost time was determined using an average wage rate implied by the then government minimum wage of N3,000. Therefore, most of the issues that could be raised with Leighton and Foster (1993) apply to that study.
3.3
WHO Funded Study
In 2003, a nationally representative survey was funded by the World Health Organisation (WHO) with a view to determining the economic burden of malaria in Nigeria. The findings of the study were reported in Jimoh (2004). The study used a combination of household survey, health facility survey and official statistics (secondary data) to arrive at its results. The study employed the three alternative approaches to measuring a disease burden – PF, COI and WTP approaches. The household survey was conducted in August 2003 covering 1,582 households spread over all the three major malaria epidemiological zones of Nigeria. The health facility survey was conducted in September 2003 and it covered five primary Health Centres (PHCs) in Kwara State in the Central Area of Nigeria. The major findings when the COI approach was applied are that the average number of malaria episodes is 1.08 per household per month and this translates to average of 13 episodes per household per year. At an average household size of about 4.4, this translates to about three malaria episodes per capita per annum. The study suggests that average cost of treating a malaria episode by self-medication is N45.86 ($1.15), N300.60 ($2.37) by herbalist/spiritualist and N1551.26 ($12.21) by clinic/hospital. With estimated population of 120 million for Nigeria in 2003, it was estimated that treatment cost (including cost of transport) was about N284,992 million ($2,244 million) which constituted about 3.9% of the GDP in 2003 (see CBN (2003) for 2003 GDP statistics). The same study found that on the average, 2.64 days are lost by victims per an adult episode of malaria with a corresponding loss of 2.08 days by caretakers; sick students are reported to miss on the average 1.67 school days per episode. Projecting this for the entire country gives a total of N303,910 million ($2,343 million) constituting about 4.1% o the country’s GDP. Furthermore, the value of protection expenditure was estimated at N3,208 ($25.26) per capita per annum. This translated to about N384,965 million ($3,031 million) per annum for the entire country representing about 5.2% of the GDP. Finally, institutional (or public) cost was estimated at about 0.27% of the GDP. > Table 84‐1 presents the main results of Jimoh (2004) along side with the results from two other similar studies for ease of comparison. The Table shows that Leighton and Foster’s (1993) estimate of indirect costs of malaria in Nigeria (which is put at between 1.1 and 5.4 per of the GDP) is similar to that of Jimoh (2004) which is about 4% of the GDP. Similarly, the estimates of the two studies in respect of the average number of malaria episodes per capita per annum are similar. However, the table shows that Onwujekwe et al. (2000) estimates – when
The Economic Burden of Malaria in Nigeria and Willingness to Pay
84
. Table 84-1 Estimated indicators of malaria burden by different studies Description
Leighton and Foster (1993)
Average number of adult malaria episodes
1–4
Average missed work days per episode
1–3
Average treatment cost per household (N) Average treatment cost per episode (N)
Onwujekwe et al. (2000)
Jimoh (2004) 3 4.72
2,092.20
8,220.94
295.93
634.33
Average protection cost per household (N)
–
14,243.52
Average protection cost per capita (N)
–
3,208.00
Average lost output per household (N)
1,454.76
11,228
205.77
2,529
Average lost output per capita (N) Institutional cost per episode (N)
–
Total treatment cost for the country (N’m)
31,960
Total indirect for the country (N’m)
22,223
44.40 284,993 303,458
Total Prevention Cost for the Country (N’m)
384,960
Total institutional cost for the country (N’m)
19,943
Total cost for the country (N’m) Total treatment cost as % of HH income
54,183 4–13
Total treatment cost as % of GDP Total indirect as % of GDP
1.1–5.4
Total prevention cost as % of GDP
–
Total institutional cost as % of GDP
–
Total cost as % of GDP
993,354
1.18
3.89
0.82
4.14 5.25 0.27
2.00
13.54
Computed from Leighton and Foster (1993), Onwujekwe et al. (2000) and Jimoh (2004) It presents major indicators of malaria burden in Nigeria as estimated or derived from three existing studies on that subject. The first column presents estimates by Leighton and Foster (1993) while the second column presents estimates derived form Onwujekwe et al. (2000). The third column presents estimates by Jimoh (2004). An increase in any of these variables implies an increase in the burden of malaria. HH household; GDP gross domestic product; N Naira; m Million
they are comparable – are generally lower than those of Jimoh (2004) which suggests that the burden of malaria per capita may be significantly higher in the urban areas than in the rural areas – from where Onwujekwe (2004) drew its sample. It seems therefore that Jimoh (2004) provides more comprehensive and representative estimates of the economic burden of the malaria illness in Nigeria than any of the existing studies. However, like all other studies on this subject, it does not investigate mortality costs.
1451
1452
84
The Economic Burden of Malaria in Nigeria and Willingness to Pay
> Table
84‐2 presents a summary of such estimates based on the COI approach and this shows that the economic burden of malaria may be consuming up to 13.5% of Nigeria’s GDP. But we must caution that this does not translate to 13.5% in lost output because expenditures on health care (direct treatment cost of a Disease) are expected to stimulate economic activities and outputs in the healthcare sector and therefore expand the GDP. Similarly, protection (as well as institutional) expenditures will stimulate economic activities in the sectors producing the relevant goods and services. Therefore, whether or not malaria has an overall negative effect on output will depend on the relative strength of its output-creating effect (which is yet to be estimated by any study that has come to our knowledge) and its output-reducing effect. One of the ways to determining the net output effect of malaria is to apply the PF approach. In terms of the distribution of the economic burden of malaria by types, > Chart 84‐1 shows that private > protection costs take the lion share at about 38.7% of the total costs which is followed by private indirect costs that account for about 30.6%. Private treatment costs account for about 28.8% of the burden while > institutional costs about 2%.
. Table 84-2 Estimated economic burden of malaria in Nigeria by types Description
N(million)
$(million)
% of GDP
Direct treatment cost
284,992
2,244
3.9
Protection cost
384,965
3,031
5.2
Indirect cost (lost output)
303,910
2,393
4.1
Institutional (public lost)
19,943
167
0.3
993,810
7,825
13.5
Total
Source: Jimoh (2004) > Table 84‐2 presents the value of the economic burden of malaria by four broad types as estimated by Jimoh (2004). The first column presents these estimates in millions of local currency (Naira) while the second column presents the same estimates in millions of US dollars. The third column expresses the estimated economic burden as a percentage of Nigerian GDP. GDP gross domestic product; N Naira; m million
. Chart 84‐1 Nigeria Malaria Burden by Types (Percentages). Chart 1 uses a pie diagram to show the distribution of the economic burden of malaria by types, namely, private protection costs, private indirect costs, private treatment costs and institutional costs. The relative sizes of the components of the chart reflect their relative shares of the burden
The Economic Burden of Malaria in Nigeria and Willingness to Pay
84
The results from the application of the PF approach by Jimoh (2004) to evaluate the economic burden of malaria in Nigeria puts the negative (net) output effect (indirect cost) of malaria at about 4% of the GDP and Jimoh (2005, p. 41) suggests that the bulk (75%) of the output loss is borne by the Nigerian agricultural sector which is the largest employer of labor and in which peasant farmers are in the majority. Consequently, the economic burden of the malaria disease seems to fall disproportionately more on the poor in Nigeria. But it is important to note that even with the application of the PF approach, the chances are there that the negative output effects that are often attributed to malaria alone may the combined effects of some other important diseases who impacts are not specifically provided in the estimated production function. Yet, except the impacts of these other diseases are truly randomly distributed over time or they are not significant even when put together, the resulting estimates may overstate the output effect of malaria. Though Gallup and Sachs (2001) expressed some concern about the large size of the estimated negative output effect of malaria and concluded that the associated mechanism by which it has such a large effect remains unclear, it might well be that such effect include the effects of other diseases.
4
Willingness to pay for Malaria Control and Treatment Programs
There have been a number of studies that have investigated the willingness of Nigerians to pay for malaria treatment and control programs or some components of such programs (Jimoh et al., 2007; Onwujekwe, 2001; Onwujekwe et al., 2004, Onwujekwe and Uzochkwu, 2004).
4.1
Onwujekwe’s WTP Study
Onwujekwe (2001) in an attempt to compare the theoretical and predictive validity of two alternative elicitation formats – BWFU and bidding game – found, among others, that over 83% of respondents who were selected from two rural communities in Enugu State were willing to pay for the large nets for their own use and that the mean slated WTP value was N242.47 ($2.55) which is below its market of N300 ($3.16) at the period of study. The implication of this is that ITNs would have to be significantly subsidized for it to be used by the majority. A similar pattern was observed in respect of respondents’ willingness to pay for the small-sized ITNs. Though the study attempted to determine the major determinants of respondents’ WTP for ITNs, the poor statistical properties of the estimated regression model (adjusted R2 of less than 0.2 and the statistical insignificance of most of the estimated coefficients) suggest that little valid inferences could be made from the regression results. Similarly, because the sample were drawn from two rural communities, the mean stated WTP for ITNs that were reported in the study are at best indicative estimates of the WTP for ITNS in the entire country.
4.2
Onwujekwe and Others’ WTP Study
Onwujekwe et al. (2004) reported that less than 5% of respondents surveyed were willing to buy ITNs at the price of N450 ($4.09). It also found that economic status of the respondents
1453
1454
84
The Economic Burden of Malaria in Nigeria and Willingness to Pay
was a significant explanation of whether they own ITNs or not with only 14.9% of the poor respondents having ITNs in contrast to 21.1% of the rich who had ITNS. The mean stated WTP for ITNs by the poor was found to be about N105 ($0.95) while the corresponding figure for the rich is N230 ($2.09). It also found that the demand for ITNs among the poor was more price sensitive than those of rich. The study concluded that the poorest socio-economic groups were less likely to purchase ITNs and consequently recommends that a universal or targeted ITN subsidy program (for the poor and biologically vulnerable groups) would increase the demand for ITNs in Nigeria. Onwujekwe and Uzochukwu (2004) found that the mean WTP for ITN for own use ranged between N179.8 ($1.63) (obtained from open-ended elicitation format) to N188.8 ($1.72) (that was obtained from the BWFU elicitation format); the market price then was N450 ($4.09). This again suggests that subsidy would be required to increase ITN coverage in Nigeria. Furthermore, it found that some well-to-do households were prepared to pay some amounts to enable poor households purchase their own ITNs. Mean stated WTP for such altruistic contributions ranges between N38.8 ($0.35) and N103.7 ($0.94). On the basis of these findings, they concluded that the potential for well-to-do people to contribute for the poor to own ITNS actually exists and that malaria control program should explore altruistic contribution as a mechanism to increase ITN coverage. The conclusion that can be drawn from these studies is that an ITN subsidy program is inevitable if ITN coverage is to increase significantly in Nigeria and targeted subsidy programs may be the way to start. However, all the above mentioned studies have been carried out in Enugu State and all have virtually been in rural communities which may limit their values for generalization.
4.3
Jimoh and Others’ WTP Study
Jimoh et al. (2007) was directly concerned with obtaining the WTP of Nigerians for malaria treatment and control. It uses the data obtained from Jimoh (2004) which is nationally representative. The study investigated four major questions. These are: What households were willing to pay to be covered if there is a program of medical care which offers free malaria treatments for members of households that subscribe to it. What they were willing to pay to be supplied with the number of ITNs that would be sufficient for the needs of all members of the households. What they are willing to pay to be protected from malaria via area spraying that would be undertaken twice a year. What they were willing to pay per household per annum for the total eradication of malaria. The study found that households were willing to pay an average of N1,112 per month (USD 9.3) for the treatment of an adult and N1,132 (USD 9.4) for a child. Similarly, households were willing to pay an average N1,325 (USD 11) per household to be supplied with ITNs that would be sufficient to meet the needs of all members of the households and would be willing to pay about N1,068 (USD 8.9) per annum for area spraying. Also, households were willing to pay N7,324 (USD 61) per household per annum for total eradication of malaria. Comparing the stated WTP values with relevant households’ actual expenditures (see > Table 84‐3), the study concluded that the difference between the actual cost of protection, treatment and indirect cost to the households and what the households are willing to pay for the eradication of malaria which was about N2,715 (USD 22.6) per month per household represents the household valuation of the intangible costs. This implies that intangible costs may be about 12% of the Nigerian GDP. Similar differences in respect of malaria treatment
The Economic Burden of Malaria in Nigeria and Willingness to Pay
84
. Table 84-3 Estimates of what households are willing to pay and their corresponding actual expenditures Cost items (N)
Actual expenditures (N)
Treatment – adulta,b
685.19
0
1 131.70
–
10 325.00
324.10
Area spraying (2/year) Room spraying
643.86
–
–
10 068.00
806.89 c
426.91
–
Window/door nets
Eradication of malaria
10 112.10
Excess over actual (N)
10 000.90
Treatment – child Bed nets
Amount they are willing to pay (N)
0
4 609.19
– 0
7 324.20
20 715.01
Source: Jimoh et al. (2007) > Table 84‐3 presents estimates of what Nigerians are willing to pay for various types of malaria interventions and services as estimated by Jimoh et al. (2007). The first column presents the levels of actual expenditure while the second column presents stated WTP values. For the first row – treatment cost for adult – the reported actual cost of treatment is a weighted average of the costs of treatment by self-medication, by herbalist/spiritualist and by clinic/ hospital. The average cost of treatment in clinic/hospital is about N1,340 ($1.17). N Naira a Actual is the weighted average cost of obtaining treatment and cure from the three major health-care providers multiplied by average malaria cases per household (1.08). It is calculated that 0.53, 0.07 and 0.40 of patients are treated and cured by self-medication, Herbalist/Spiritualist and Clinic/Hospital respectively b Treatment cost in Clinic/Hospital not involving admission is used c Actual is the sum of protection expenditures, weighted treatment costs and indirect costs – all per household and per month
and ITNs were interpreted as price for elimination of payment risk (insurance policy) and payment for the inconvenience involved in self procurement. At an average household size of 4.4 reported in the study, the reported stated WTP for ITNs implies a per capita WTP for ITNs of N301.14 ($2.37) which is below its market price of N450 ($3.54) at the close of 2003. Thus, the study provides further evidence that an ITN subsidy program would be required to increase ITN coverage in Nigeria. Similarly, the stated WTP for malaria treatment per capita per month is less than the average cost of treating an episode of malaria in a clinic/hospital that was reported by Jimoh (2004) to be about N1,340 ($11.17) for cases that do not involve hospitalization and N1,551 ($12.93) for cases that involve hospitalization. This suggest that subsidy would be required if the utilization rate of modern healthcare facilities for the treatment of malaria is to be increased significantly. Another interesting outcome of the study is the identification of the major determinants of the stated WTP for malaria eradication. The major findings are presented in > Table 84‐4. These results indicate that the major determinants of households willingness to pay for malaria eradication and control are their: level of education, income, cost of protection, self assessment of their status relatively to others in the society and total cost of obtaining treatment in the hospital with married households ready to pay more than the singles. Similarly, those currently attending public medical facilities are willing to pay less for malaria control as well as strangers (i.e., those who have not stayed in the community more than 1 year). One important implication of these estimates is that, other things held constant,
1455
1456
84
The Economic Burden of Malaria in Nigeria and Willingness to Pay
. Table 84-4 Estimates of the log-linear regression model Coefficientsa, b Unstandardized coefficients
Standardized coefficients
B
Std. error
0.125
0.069
0.010
Log of HH highest level of education
0.408
0.066
0.124
6.203 0.000
Log of HH income
0.408
0.022
0.604
18.466 0.000
Log of cost of spraying
0.224
0.028
0.167
7.877 0.000
0.578
0.077
0.077
7.460 0.000
0.210
0.032
0.167
6.652 0.000
0.300
0.185
0.005
1.624 0.104
0.171
0.067
0.019
2.552 0.011
Model Public medical facilities
Log self-assessment rating Log of total cost of hospital treatmentc Stranger Married household
Beta
t
Sig.
1.822 0.069
Source: Jimoh et al. (2007) > Table 84‐4 presents the results of a regression analysis undertaken by Jimoh et al. (2007). The estimated unstandardized coefficients (B) are the effects of a one percentage change in the corresponding explanatory variable on the dependent variable (WTP). The last column of the table reports the level at which the explanatory variable is statistically significant with a value above 0.05 suggesting that variable is not statistically significant at the conventional 5% level of significance. Dependent Variable: Log of WTP to eradicate malaria; Other Statistics: R2 = 0.987; DW = 1.522; F = 15341.9 a Linear regression through the origin b Treatment cost in clinic/Hospital not involving admission is used
as the economic status of respondents improves (as indicated by income and wealth levels), their WTP for malaria treatment and control increases. Consequently, the poor have significantly lower WTP than the rich.
Summary Points Evidence available from existing studies suggests that the burden of malaria illness in Nigeria may be in excess of 25% of the GDP.
Direct and indirect costs (excluding the costs attributable to mortality) constitute about 13% of the GDP while the value placed on intangible costs comes to about 12% of the GDP.
Of the direct and indirect costs about 98% is borne by the private citizens and the lion share of this is on protection expenses.
Also, available evidence suggests that that average WTP for malaria treatments is lower than the average cost of treatment in clinic/hospital and that economic status (wealth and income) is one of the major factors determining the amount Nigerians are willing to pay for the treatment or total elimination of malaria. In a society where most (over 65% of) healthcare expenses are financed through household out-of-pocket expenditures (Soyibo et al., 2005), majority of the poor who have low WTP may indeed be unable to pay for the cost of malaria treatment at ruling market rates in modern clinics/hospitals when the need arises.
The Economic Burden of Malaria in Nigeria and Willingness to Pay
84
Such inability to pay probably explains why self-medication appears to be a popular first
line of action for the treatment of malaria by the poor as well as explaining why herbal and spiritual care providers still attract some patronage (Jimoh, 2004, p. 30–32). There is therefore a need to promote universal access for the treatment of malaria in Nigeria (and may be for healthcare in general) by abolishing or reducing payment for malaria treatment (or healthcare) at the point of need. By abolishing or reducing payment at the point of need, not only are the chances of catastrophic health expenditures among the poor reduced but by promoting economic growth via improvement in the health of the poor, it may also be a poverty-alleviation strategy. However, this narrows the viable financing options to two, namely, tax-financed free-atthe-point-of use treatment for malaria or insurance-based financing arrangement for the same purpose. The latter appears more attractive because it may be more fiscally sustainable than the former. Nigeria is currently putting in place a national health insurance scheme (NHIS) starting with the formal sector employees but progress has been very slow. With the informal sector still virtually uncovered by NHIS and the requirement that some co-payment be made at the point of treatment (NHIS, 2002), it is yet to be seen how the needs of the very poor will be catered for.
References American Association for the Advancement of Science (AAAS). (1991). Malaria and Development in Africa: A Cross-Sectoral Approach. AAAS, Washington, DC. Audibert M, Mathonnat J, Nzeyimana I, Henry M. (1999). Revue D’Economie du Development. 4: 121–148. Audibert M. (1986). J Dev Econ. 24: 275–291. Audibert M, Mathonnat J, Nzeyimana I, Henry M. (1999). Revue D’Economie du Development. 4: 121–148. Brown T, Champ P, Bishop R, McCollum D. (1996). Land Econ. 72: 152–166. Central Bank of Nigeria (CBN). (1999). Statistical Bulletin, CBN. Vol. 10, No. 1. Central Bank of Nigeria (CBN). (2003). Annual Reports and Statement of Accounts, CBN. Chima RI, Goodman CA, Mills A. (2003). Health Policy. 63: 17–36. Donaldson C. (1990). J Health Econ. 9: 103–118. Ejezie GC, Ezednachi EN, Usanga EA, Gemade EI, Ikpatt NW, Alaribe AA, et al. (1991). Acta Tropica. 48: 17–24. Ettling M, McFarland DA. (1992). Economic Impact of Malaria in Malawi. Vector Biology Control Project, Virginia. Ettling MB, Shepard DS. (1991). Trop Med Parasitol. 42: 214–218.
Federal Ministry of Health. (1992). The National Health Policy of Nigeria. Federal Ministry of Health, Lagos. Federal Ministry of Health. (2001). Strategic Plan for Rolling Back Malaria in Nigeria 2001–2005. FMOH, Abuja. Federal Office of Statistics (FOS). (1995). Annual Abstract of Statistics. FOS, Lagos. National Health Insurance Scheme. (2002). Operational Guidelines on NHIS. Abuja. Frykblom P. (1997). J Environ Econ Manage. 34: 275–287. Gallup JL, Sachs JD. (2001). Am J Trop Med Hyg. 64: 85–96. Klose T. (1999). Health Policy. 47: 97–123. Jimoh A. (2004). The Economic Burden of Malaria in Nigeria. WHO Final Technical Report. Jimoh A. (2005). Agrosearch. 7: 37–46. Jimoh A, Sofola O, Petu A, Okorosobo T. (2007). Cost Eff Resour Alloc. 5: 1–17. Johannesson M, Jonsson B, Borgquist L. (1991). J Health Econ. 10: 461–473. Leighton C, Foster R. (1993). Economic Impacts of Malaria in Kenya and Nigeria. Health Financing and Sustainability Project 6, Bethesda, Abt Associates, Maryland. Loomis J, Brown T, Lucero B. (1997). Environ Resour Econ. 10: 109–123.
1457
1458
84
The Economic Burden of Malaria in Nigeria and Willingness to Pay
McGregor IA. (1988). In: Wernsdorfer WH, McGregor IA (eds.) Malaria: Principles and Practice of Malariology. Churchill Livingstone, Edinburgh, pp. 753–767. Morrison GC, Gyldmark M. (1992). Health Econ. 1: 233–243. Onwujekwe O. (2001). Health Econ. 10: 147–158. Onwujekwe O, Chima R, Okonkwo P. (2000). Health Policy. 54: 143–159. Onwujekwe O, Hanson K, Fox-Rushby J. (2004). Malaria J. 3: 6. Onwujekwe O, Uzochukwu B. (2004). Health Econ. 13: 477–492. Sauerborn R, Shepard D, Ettling M, Brinkmann U, Nougtera A, Diebfeld H. (1991). Trop Med Parasitol. 42: 219–223.
Shepard DS, Brinkmann U, Ettling M, Sauerborn R. (1991). Trop Med Parasitol. 42: 199–203. Soyibo A, Odious O, Ladejobi F, Lawanson AO, Oladejo B, Alayande S. (2005). National Health Accounts of Nigeria, 1998–2002. Final Report of World Health Organization (WHO) Sponsored Study, Geneva. Wang’ombe JK, Mwabu GM. (1993). Soc Sci Med. 37: 275–291. WHO/AFRO. (2001). A Framework for Estimating Economic Burden of Malaria in the African Region. WHO, Harare, Zimbabwe. WHO/UNICEF. (2003). Africa Malaria Report. WHO/ CDS/MAL/2003.1093.
85 Measurement of Adverse Health Burden Related to Sexual Behavior S. H. Ebrahim . M. McKenna 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1460 2 History of Analysis of Sexual Behavior Related Health Burden . . . . . . . . . . . . . . . . . . . 1461 3 Challenges to Assessment: Conceptual . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1463 4 Challenges to Assessment: Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1464 5 Estimation of Attributable Fractions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1465 6 DALY-Specific Computations and Adjustments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1465 7 Comment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1466 8 Way Forward . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1467 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1468
Disclaimer: The findings and conclusions in this report are those of the authors, and do not necessarily represent the official position of the U.S. Department of Health and Human Services. #
Springer Science+Business Media LLC 2010 (USA)
1460
85
Measurement of Adverse Health Burden Related to Sexual Behavior
Abstract: As part of an analysis of the burden of disease and injury in the United States, we established a method to identify and quantify the incidence of adverse health events, deaths, and disability adjusted life years (DALY) attributed to unprotected sexual behavior (USB). Estimates for the incidence, morbidity and mortality from sexually transmitted diseases (STD) were derived from vital statistics, notifiable disease surveillance system, national surveys, outpatient and hospital discharge summaries, prevalence data from individuals attending health care services, the published literature, and expert opinion. Because many conditions for which USB is implicated as a cause have multifactorial causes, a major challenge is developing estimates for the attributable fraction of disease that is exclusively related to USB. For conditions with risk factors other than sexual behavior (e.g., HIV), or etiology (e.g., liver disease), sexual behavior-attributable fraction was estimated based on the literature or through expert consultations. Maternal and infant conditions related to unintended pregnancy and all voluntary abortions were considered as adverse outcomes of sexual behavior. A comprehensive estimation including morbidity, mortality, and summary measures such as DALYs can identify a larger burden related to sexual behavior and more profound disparities by gender than that can be captured by traditional measurement of mortality or morbidity alone. However, because epidemiology of health events related to sexual behavior varies by countries or regions, statistics such as attributable fractions should be developed based on these geographic units. Further, summary measures involve the synthesis of parameters considered objective and science-driven with culture-specific and value oriented inputs. Many countries have limited information regarding either of these topics and must develop estimates in this context. This chapter provides only a roadmap for such estimation. List of Abbreviations: DALY, disability adjusted life years; HIV, human immunodeficiency virus;; ICD, international classification of diseases; STD, sexually transmitted diseases; USB, unprotected sexual behavior
1
Introduction
Health systems and health policies are increasingly being challenged by increasing demand for health care, rapid epidemiologic transition, exploding costs associated with the delivery of medical services and greater accountability for effective use of resources. The evidence base for health policy decisions can only as sophisticated as the comprehensiveness and relevance of the information provided to those individuals responsible for these decisions. The past four decades experienced significant changes in how we measure health and prioritize health challenges. Improvements in epidemiology and biomedical sciences have heralded conceptual changes in our understanding of the depth and scope of health service needs, prevention opportunities and research. Many factors the interconnectedness among diseases in terms of causation, the potential opportunities of a bundled approach to their combined prevention, and a life-span approach to evaluating the impact of public health measures have helped move health initiatives from a disease and end stage oriented approach to more of an ‘‘upstream,’’ health determinants approach. Such an approach is epidemiologically sound, rational, and potentially cost-effective. Health determinants that have generally received the most attention include sexual behavior, smoking, alcohol consumption, caloric over consumption inadequate fruit and vegetable intake, the built environment, and poverty. The influence of sexual behavior on reproductive health and sexually transmitted diseases (STDs) is obviously one of the strongest
Measurement of Adverse Health Burden Related to Sexual Behavior
85
relationships. For the purposes of this review USB is defined as sexual activities that result in outcomes considered undesirable from a societal perspective regardless of the intent of the participants at the time of the sexual encounter. This includes encounters where contraceptives and barrier methods were used incorrectly, or ineffectively – as well as sexual encounters resulting in unintended pregnancies. An important clarification about the DALYs attributed to unintended pregnancies is that only the negative health outcomes associated with these pregnancies (e.g. infection, hemorrhage, infant mortality) were denominated. There are no valuations based on the negative, or positive, emotional relationships, or productivity outcomes that develop as a result of the pregnancies being unintended. Similarly, the DALYs attributed to elective abortions only include the direct, negative health outcomes (e.g. infection, hemorrhage, post-surgical complications) which may result from such procedures. The epidemiologic focus of this approach calls for programmatic interventions aimed at promoting healthy sexual behavior. Because the adverse outcomes of sexual behavior usually affect younger individuals than many other major diseases such as cancer or heart disease, relative assessments of the impact of these various conditions that only count deaths or incident cases of disease alone are inadequate for a proper understanding of the dimensions of the issue. Some adverse effects of prevalent disease (for example Chlamydia trachomatis infection in women who may remain asymptomatic and untreated) affect individuals prior to and even during some of their most potentially productive years. In addition, some of these conditions will have long term effects that diminish the quality of life of the affected for many years (for example infertility related to sexually transmitted infections). This chapter builds upon past efforts to quantify the burden attributable to sexual behavior and presents a comprehensive approach to assess the public health burden related to USB. This paper presents a roadmap for the estimation of summary statistics on adverse health burden from sexual behavior including the incidence of adverse health events (morbidity and their sequelae), mortality; and disability-adjusted life- years. DALY is a time-based, composite indicator of the burden of disease that adds loss of life-years due to premature death and loss of healthy life due to morbidity and associated disability.
2
History of Analysis of Sexual Behavior Related Health Burden
Composite examination of the burden from sexual behavior was not considered until the1980s during which the emergence of HIV and related expansion in research lead to broadening of the spectrum of disorders associated with USB. Until then, diseases related to sexual behavior were subsumed under the umbrella of reproductive health. To our knowledge the first attempt to focus attention on comprehensive burden assessment was an analysis by Grimes of the U.S. Centers for Disease Control in 1986, who studied STD-related mortality in the United States for the years 1955, 1965, and 1975, expanding the concept of reproductive mortality to include deaths related to sexually transmitted diseases. Under this new classification which counted both infectious and other conditions related to reproductive health within the framework of USB, STDs caused 32% of reproductive mortality in 1965. As the debate on the causation of cervical cancer was not conclusive at that time, Grimes excluded cervical cancer from this analysis. However, Grimes underscored that if cervical cancer was included, then cervical cancer deaths alone would alone outnumber all other reproductive causes combined. Cervical cancer deaths have declined substantially in the U.S. and other developed
1461
1462
85
Measurement of Adverse Health Burden Related to Sexual Behavior
countries and its contribution to the burden of USB will continue to decline in developed countries. However, the burden form cervical cancer remains significant in developing countries. Globally, the 1994 International Conference on Population and Development expanded the reproductive health concept to include ‘‘sexual and reproductive health.’’ The recognition of cervical cancer as being related to USB, the emergence of sexually transmitted cases of HIV, recognition of and improvements in laboratory diagnosis of Chlamydia trachomatis, better understanding regarding the transmission dynamics of hepatitis viruses in the 1980s prompted the reexamination of the concept of USB as a major root cause of ill health. In 1993, McGinnes and Foege developed a crude estimate of deaths by causative factors and identified USB as a rapidly increasing factor. This estimate was not based on International Classification of Diseases (ICD) codes of individual case records; rather they attributed deaths from specific causes based on the role of of sexual behavior on the total number of deaths. In 1997, Ebrahim and colleagues at the U.S. Centers of Disease Control refined the earlier methodologies and published a revised analysis of deaths attributable to USB in the U.S. for the 20 year period from 1973 through 1992. They used ICD-coded actual data reported on death certificates of individuals. Due to various challenges in attribution of a cause-disease link and limitation of data, they categorized the quality of their findings as ‘‘hard’’ data indicating verifiable statistics, and ‘‘soft’’ data that indicates best estimates based on various assumptions. The analysis categorized diseases related to USB into three categories; those conditions that are always sexually transmitted (e.g., gonorrhea), conditions that are not always transmitted sexually (e.g., hepatitis viruses) for which attribution was based on epidemiological estimates, and HIV infections acquired sexually based on patient information provided from case reports on HIV infected individuals. Although Chlamydia was recognized as STD pathogen by this time, it was not included in the ICD eighth revision. Therefore, it was not possible to identify persons who had the diagnosis of Chlamydia in that analysis. Given the burden of Chlamydia infections in women and its relevance to infertility, the inability to include an estimate Chlamydia burden in that analysis was a major limitation. By the late 1990s a new concept for the measurement of ill health was being articulated by the Global Burden of Disease (GBD) study. A metric called the disability-adjusted life year (DALY) was developed which attempted to capture both fatal and non-fatal health outcomes, separately and in combination. The GBD provided comprehensive data on sexual and reproductive health related DALYs worldwide. Using this methodology, Ebrahim et al. (2005), published a detailed account of all measurable health effects attributable to USB in the U.S. Their findings indicated that DALYs were able to elucidate previously unrecognized burdens from the non-fatal losses in health associated with the adverse effects of USB. Overall, in the United States, in 1998, mortality estimates attributed 1.3% of U.S. deaths to USB, whereas DALY estimates put the proportion of the health burden associated with USB at 6.2% of the total. Some disparity was seen by gender between mortality and DALYs; females suffered the majority of incident USB-related adverse health events (62%) and DALYs (57%), whereas males suffered the majority of deaths (66%). If HIV-related mortality were excluded, more than 80% of USB-related mortality would be those among women. Among females, more than half of the incident events and DALYs were contributed by curable infections and their sequale. However, deaths from curable STDs were rare; the majority of USB deaths are from viral infections and their sequelae (men 99.5%; women 96.8%). Cervical cancer and HIV were the leading causes of such mortality among females, whereas HIV was by far the single leading cause among males. Also, viral diseases caused the majority of health burden attributed to USB among males. This chapter summarizes that methodology.
Measurement of Adverse Health Burden Related to Sexual Behavior
3
85
Challenges to Assessment: Conceptual
A major conceptual challenge in defining USB related health burden is the overlap between reproductive health and USB-related health. Current definition of USB-related health burden includes all conditions or aspect of health condition for which USB is epidemiologically significant. Therefore, the list of conditions includes sexually transmitted diseases, medical complications of pregnancy and child birth, and some reproductive outcomes (> Table 85-1). A comprehensive approach to calculate burden from root causes such as USB is challenging from a measurement perspective. The health metrics that are in use today by most health systems are not comprehensive enough to capture the full dimensions of adverse effects of root causes of diseases. This is because of both the challenges in quantifying certain behaviors and because the adverse effects of behaviors or root causes cuts across the lines along which diseases have traditionally been measured. Public health and medical practice is dominated by vertical disease-focused programs that yield separate and poorly coordinated data collection efforts. Measurement of burden of chronic diseases that is attributable to USB is much . Table 85-1 Attributable fractions used for estimation of sexual behavior-related adverse health burden in the United States Attributable to sex (%) Female Curable diseases and their sequelae Chlamydia
100
Gonorrhea
100
Trichomoniasis
100
Syphilis
100
Other curable STDs
100
PID
65
Infertility
10
Viral diseases and their sequelae Genital herpes
100
HPV
100
-Cervical cancer
100
Hepatitis B Virus
33b,58c
Hepatitis C Virus
9b,20c
HIV
57d
Pregnancy-related Pregnancy outcome’ Elective abortions’
e
94f
Male Curable diseases and their sequelae Chlamydia
100
1463
1464
85
Measurement of Adverse Health Burden Related to Sexual Behavior
. Table 85-1 (continued) Attributable to sex (%) Gonorrhea
100
Trichomoniasis
100
Syphilis
100
Other curable STDs
100
Urethral stricture
–h
Epididymitis
–h
Infertility
–h
Viral diseases and their sequelae Genital herpes HPV
100 100
Hepatitis B Virus
33b,58c
Hepatitis C Virus
9b,20c 72d
HIV
These conditions include events attributed to adverse effects of unprotected sex and not necessarily attributed to sexually transmitted infections or their sequelae. All other conditions are attributed to infections and their sequelae a Not included in the number of women with HPV b Chronic liver disease and hepatocellular carcinoma c Acute conditions d McGinnis and Foege (1993) e Based on age-specific rates of unintendeness of pregnancy (Henshaw 1998) and includes effects on both mother and child f Abortions performed for medically not indicated reasons are attributed as adverse effects of sexual behavior g Allocated to pregnancy outcome h Based on natural history of gonorrhea and chlamydia (Murray 1998) i Parameters are inadequate to compute DALYs
more challenging than infectious diseases that have a clear causal pathogen, more defined incubation period and pathological process. Also, mixing burden from chronic disease causes based on events that occurred several decades ago with current statistics from acute conditions presents a challenge to policy assessment when considering the time frame for events. For example, USB that results in papilloma virus infections and subsequent cervical cancers occurs one or two decades ago, prior to the diagnosis of the neoplasm, whereas pregnancy-related health events result within less than one year of the occurrence of the USB. Therefore, overall DALY measurements of current morbidity and mortality associated with USB may not capture all of the relevant changes in such behaviors in recent years.
4
Challenges to Assessment: Data
Statistics needed for estimation of sexual behavior-related burden may come from (1) reportable infections (such as syphilis, gonorrhea, Chlamydia) which may vary by countries (2) diagnosis made during outpatient or inpatient visits to medical facilities
Measurement of Adverse Health Burden Related to Sexual Behavior
85
(3) national surveys of representative population samples, and (4) data from mathematical models of STD epidemiology based on knowledge about the natural history of these diseases. A simple comprehensive approach for attributing adverse health outcomes to specific root causes such as USB is not available. USB plays a role in many adverse health outcomes that have a multifactorial etiology. USB leads to a variety of harmful consequences, ranging from unintended pregnancy, social stigma, infections and chronic psychological or pathological sequelae including cancers. The consequences are considered not because all adverse effects are classical diseases or pathological process that can be easily defined but because they affect the quality of life of the affected individual. The current scheme for the classification of diseases ICD is based on diseases and this is inadequate to assess disease burden from a root cause. For example, only a proportion of hepatitis B virus is transmitted sexually, and only a proportion of liver cancer is caused by hepatitis B virus. Sorting out the relative contribution of sexual behavior for some conditions and quantification of such conditions require disparate approaches and varying data sources. Similarly, calculation of DALY’s for a given condition require information on the natural history of that condition including incidence and duration of illness, the extent of disability, and the number of deaths, some of which are not readily available.
5
Estimation of Attributable Fractions
The analytic methodology presented here adapted and updated the methodology used in the global burden of disease study—initiated by the World Bank and the World Health Organization, and all other attempts to assess various components of USB- related health burden. Data were reviewed from published estimates of incidence of STD and reproductive morbidity and mortality, national vital statistics, notifiable disease surveillance system, national surveys, outpatient and hospital discharge summaries, prevalence data from individuals attending health care services, and expert opinion. We included 100% of all major sexually transmitted infections. For conditions with risk factors other than USB (e.g., HIV), or etiology (e.g., liver disease), we identified the sexual behavior-attributable fraction from the literature (> Table 85-1). Consistent with assumptions in the report of the World Health Organization, maternal and infant conditions related to unintended pregnancy and all voluntary abortions were considered as adverse outcomes of sexual behavior. Sexual behavior-related conditions with insufficient data on attributable fractions (e.g., bacterial vaginosis, rape-associated violence and post-traumatic stress disorders) were generally excluded. For conditions that are not always attributable to sexual transmission, data shown in > Table 85-1 are prepared for the United States based on various epidemiological parameters. These numbers need to be reassessed for other countries based on regional or local epidemiologic patterns of respective conditions. When adequate epidemiologic data is not available, expert opinion may be used to come up with most appropriate estimates.
6
DALY-Specific Computations and Adjustments
Comprehensive measures that attempt to combine the impact of non-fatal as well as fatal health conditions are known as summary health measures. The DALY is one example of such a summary measure. The methodologic and ethical complexities as well as long-standing
1465
1466
85
Measurement of Adverse Health Burden Related to Sexual Behavior
controversies associated with the computation of these measures, and especially the DALY, have been reviewed in greater detail elsewhere and are beyond the scope of this chapter (Field and Gold, 1998; Murray et al., 2002, 2006). Briefly, the DALY is a time-based measure that attempts to enumerate the years lost to premature death (years of life lost-YLLs), and the years of healthy life ‘‘lost’’ to living in a state of less-than-perfect health (yeast lost to disabilityYLDs). In order to calculate cause-specific YLLs the number of years each person was expected to live at each age of death was summed across a common set of conditions listed as the underlying cause of death on the decedent’s death certificate. The life-expectancy is calculated separately for men and women and is based on the longest observed life-expectancy in the world (i.e., Japanese women). The YLDs were estimated by multiplying incident, non-fatal health conditions by the estimated, average duration of disability associated with each type of condition. The impact of the non-fatal conditions on lost health was obtained by multiplying these durations by disability weights that range between 0 (perfect health) and 1 (severity equivalent to death). These valuations were developed for the Global Burden of Disease study and were validated for the United States as part of the overall U.S. Burden of Disease and Injury project. Various adjustments are needed when developing epidemiologic estimates for DALY-based epidemiologic studies to assure that plausible figures result. For example, because cardiovascular diseases are such a major source of disease burden in these analyses, and there is tremendous variation between, and within, countries in the attribution of deaths to these conditions, various statistical techniques are required to standardize the crude numbers for ischemic heart disease mortality. In addition to modifications of the epidemiologic measurements, because DALYs are time-based, they are usually adjusted for time preference, or discounting in accordance with widely accepted standards. Finally, many Burden of Disease analyses, also incorporates weights that provide differential value for years of life at specific ages according to the following equation: "
Cxe-Bx
where C and B are constants equal to 0.1658 and 0.04, respectively, and x is the value for a particular year of age. This calculation is not the same as assuming a death at a younger age imposes a greater burden because of the longer expectation of life associated with youth. The DALYage-weights adjust for the fact that young and middle-age adults are generally the caretakers of children and the elderly within societies. Therefore, the loss of those years of life represents a greater threat to the stability and productivity of the society than years lived in the more dependent stages of life. Given the controversy surrounding this component of the DALY measure, many investigators do not execute this adjustment. In the original Global Burden of Disease Study, age-weighted and non-age-weighted DALYs were provided.
7
Comment
Assessment of USB as a composite measure (DALY) helps magnify its burden on the health care system and the population more than that achieved by mortality measurements or morbidity measurements. This analysis framework is the most comprehensive account of USB burden to date and included more USB‐related conditions than were considered in previous analyses for the United States (such as hepatitis, infertility, reproductive sequelae
Measurement of Adverse Health Burden Related to Sexual Behavior
85
in men, maternal conditions and abortions). Even so, these estimates likely reflect the lower bound of the USB-related public health burden and call for reassessing the attributable burden for each geographical location and year of calculation. For instance, for HIV infection, we included only the percentage attributable to sexual transmission alone, and not in combination with other risk factors such as intravenous drug use. These factors change over time. Therefore, the attributable proportions listed in this paper should be revised to match the epidemiology of sexual transmission of HIV in target populations. Furthermore, USB attributed DALYs, particularly those attributed to HIV/AIDS, may vary slightly for more recent years due to the declines in HIV/AIDS deaths, increases in the number of people living with HIV/AIDS, and the resurgence of bacterial STDs and unsafe sex among some population groups. DALYs may not capture the full social and economic dimensions of reproductive morbidity and those arising from some prevalent viral infections such as genital herpes human papilloma virus infections. Of note, attributable fractions provided in this chapter used epidemiological data relevant to the U.S. and are not universal. Therefore, other geographical regions or countries interested in developing such estimates need to adjust the data according to the regional epidemiology.
8
Way Forward
Crude analyses of the data from the current public health information system cannot provide a comprehensive picture of the disease burden from underlying causes of health outcomes in order to effectively guide health planning. Health policies will be more effective if they are guided by reliable information on levels, patterns, and causes of ill health and suffering and how they change over time. Estimation of health burden with respect to a underlying causes (such as USB) is limited by the conceptual difficulties in defining the link between exposure and health event, deficiencies in the primary databases, and to the variable approaches used in various studies to derive attributable fractions. It is generally acknowledged that the field of sexual and reproductive health should not only address the key adverse outcomes but pay attention to positive, life enhancing aspects such as ‘‘safe and satisfying sex life’’ and ‘‘the enhancements of life and personal relations.’’ Methods to estimate such benefits and, more so, how to translate such findings to action for programs remain unclear. In the meantime, the quantification attempts, such as this one which focused on the adverse outcomes, compel us to further refine the methodology. The findings offer a mandate that help us focus public health policy on fundamental factors that affect a wide variety of diseases. Despite the many criticisms of the DALY as a measurement unit, it represents a major conceptual advance to assess public health burden providing a common metric for fatal and no-fatal outcomes. Both the assessment of burden and assessment of sexual-behavior attributable component of a given condition are evolving disciplines and hence subject to change based on methodologic enhancements and greater availability of data. Therefore, attempts to use the methodology at the national or public health level should consider the following recommendations.
The need for full delineation of the natural history of USB related conditions and associated death, disability and mortality.
1467
1468
85
Measurement of Adverse Health Burden Related to Sexual Behavior
Country level case studies are needed to identify, evaluate and utilize more existing and prospective data sources. For the estimation of attributable fractions such data should be subject to secondary analysis and expert panel discussions. Long term research strategies may be needed to determine the incidence and sequeale of some conditions with long incubation periods and complex, extended natural histories.
Summary Points Unprotected sexual behavior (USB) is defined as sexual activities that result in outcomes considered undesirable from a health or societal perspective.
The list of USB-related conditions expanded in recent decades to include emerging or
newly recognized infections, unintended pregnancy, the acute or chronic complications or sequelae arising from such outcomes including cancers, and efforts to minimize some outcomes such as medical termination of pregnancy. Continues assessment and monitoring of the health burden from USB is critical to evidence-based health policy because USB is among the top 10 causes of disease burden globally including in many developing countries and the U.S. Measurement of USB is challenging due to the conceptual difficulties in delineating the link between exposure and a health event, the multifactorial etiology of some outcomes that necessitate computation of geographically appropriate attributable fractions, and deficiencies in primary databases. Use of summary measures (DALY) that capture both fatal and non-fatal outcomes can demonstrate the full extent of the USB-related burden and profound disparities by gender than that measured by mortality and morbidity statistics. Summary measures of USB helps to capture the epidemiologic transition in USB; declines in burden from bacterial causes and increases in burden attributable to virally mediated conditions. All USB-related health events are preventable through sexual behavioral change.
Further Reading AbouZahr C, Vaughan JP. (2000). Bull World Health Org. 78: 655–665. Barendregt JJ, Bonneux L, Maas PJ Van der. (1996). Bull World Health Organ. 74(4): 439–443. Cates W Jr. (1999). Sex Transm Dis. 26(4): S2–S7. Centers for Disease Control and Prevention. (2001). Pregnancy-related deaths among Hispanic, Asian/ Pacific Islander, and American Indian/Alaska Native women-United States, 1991–1997. MMWR. 50: 361–364.
Centers for Disease Control and Prevention. (1998). Recommendations for prevention and control of hepatitis C virus (HCV) infection and HCV-related chronic disease. MMWR. 47: 1–39. Ebrahim SH, Peterman TA, Zaidi AA, Kamb ML. (1997). Am J Public Health. 87: 938–944. Ebrahim SH, MCkenna MT, Marks JS. (2005). Sex Transm Infect. 81: 38–40. Field M, Gold MR. (1998). Summarizing Population Health: Directions for the Development and
Measurement of Adverse Health Burden Related to Sexual Behavior Application of Population Metrics. National Academy, Washington DC. Grimes DA. (1986). JAMA. 255: 1727–1729. Gold MR, Siegel J, Russell L, Weinstein MC (ed.). (1996). Cost-effectiveness in Health and Medicine. Oxford University Press, New York. Henshaw SK. (1998). Fam Plann Perspect. 30: 24–29. Koran JM, Fleming PL, Steketee RW, De Cock KM. (2001). Am J Public Health. 91: 1060–1068. McGinnis J, Foege WH. (1993). JAMA 270: 2207–2217. Michaud CM, Murray CJL, Bloom BR. (2001). JAMA. 285: 535–539. Murray CJ, Kulkarni SC, Ezzati M. (2006). Circulation. 113(17): 2071–2081. Murry CJL, Lopez AD (ed.). (1998). Health dimensions of sex and reproduction: the global burden of
85
sexually transmitted diseases, HIV, maternal conditions, perinatal disorders, and congenital anomalies. World Health Organization, Geneva. Murray CJL, Salomon JA, Mathers CD, Lopez AD (ed.). (2002). Summary Measures of Population Health: Concepts, Ethics, Measurement and Applications. World Health Organization, Geneva. Rein DB, Kassler WJ, Irwin KL, Rabiee L. (2000). Obstet Gynecol. 95: 397–402. Torres A, Forrest JD. (1988). Fam Plann Perspect. 20: 169–176. Williams A. (1999). Health Economics. 8: 1–8. Wilcox LS, Mosher WD. (1993). Obstet Gynecol. 82: 122–127.
1469
2
Disease Burdens and Economic Impacts 2.6 Psychosocial, Social, Behavioural, Psychiatric, Neurological and Addictions
86 Global Burden of Mental Health M. Kastrup 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1474
2
Epidemiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1475
3 3.1 3.2 3.3 3.4 3.5
Main Diagnostic Groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1475 Schizophrenia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1475 Affective Disorders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1477 Nervous Disorders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1478 Abuse . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1479 Suicide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1479
4 4.1 4.2 4.3 4.4
Factors Related to Mental Disorders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1480 Migration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1480 Poverty . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1481 Comorbidity with Other Health Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1481 Gender Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1481
5
Prevention . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1482
6 Consequences of Globalization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1482 6.1 Cultural Competence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1483 7
Health Economics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1483
8
Therapeutic Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1485
9
Primary Health Care . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1485
10
Treatment Gap . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1485
11
Ethical Concerns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1487
12
Perspectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1488
13
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1489 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1489
#
Springer Science+Business Media LLC 2010 (USA)
1474
86
Global Burden of Mental Health
Abstract: Mental disorders contribute significantly to the Global Burden of Disease, as four out of the ten diseases with the highest burden are psychiatric. About 25% of all develop one or more psychiatric and behavioral disorders during their lifetime. Unipolar depression ranges as the leading mental disorder with respect to disability adjusted life years. The major psychiatric disorders like schizophrenia and depression are found in all cultures and result in significant disability. The cost of mental disorders worldwide needs receiving increasing recognition. Mental disorders are highly associated with other health problems and increase the risk for both communicable and non-communicable diseases as well as for both intentional and unintentional injuries. Vice versa, many physical disorders increase the risk for the development of mental disorders. A large proportion of mental health problems go unnoticed and the recognition hereof is close related to social and cultural factors. Similarly, the treatment gap is closely associated with economic, social and cultural background, and access to adequate care in many settings a rarity. Awareness of mental health issues should be an integrated and significant part of any health and social policy both at international as well as national level. The WHO Health for all targets focusing on disparities in health and strategies to overcome them is still pertinent. List of Abbreviations: DALY, disability adjusted life years; DOSMED, determinants of outcome of severe mental disorders; GBD, global burden of disease; IPSS, international pilot study of schizophrenia; OCD, obsessional compulsive disorder; PTSD, post traumatic stress disorder; US, United States; WHO, World Health Association; YLD, years of life lived with disability
1
Introduction
Until the middle of the twentieth century, it was a commonly held belief that mental disorders were less frequent in developing countries compared to industrialized countries. But since then it has been documented that mental disorders are frequently found worldwide and that stress factors precipitating mental distress may occur irrespective of developmental level (Sartorius, 2000a). The global burden of mental illness has been further documented in e.g., the three international reports: World Mental Health (World Bank, 1993), Institution of Medicine (Murray and Lopez, 1996) and World Health Report (WHO, 2001). These comprehensive reports have been accompanied with regional and national reports where the report from the European Ministerial Conference in 2005 deserves particular mentioning as it was signed by all Ministers of Health in the WHO European Region (WHO, 2005a). But mental health problems reach further than the mere mental health, they have an impact on the physical health, and the general mortality of those afflicted. Further, the existence of mental health problems are linked with poverty, stigmatization, marginalization in most areas of the world (Lancet Global, 2007). And the impact is not only affecting the mentally ill, but also their families and surroundings. The large epidemiological studies initiated by WHO on schizophrenia and depression (Jablensky et al., 1992; Sartorius et al., 1983) indicate the similarities in frequency across culture. Socio-cultural factors on the other hand have a significant impact on the incidence of minor mental disorders and culture specific disorders.
Global Burden of Mental Health
86
Despite geographical differences in the occurrence of mental disorders, these disorders are characterized by generally having an early onset compared to many other severe long-lasting diseases, thereby having a great impact on the life of those affected (WHO, 2001).
2
Epidemiology
Mental disorders contribute significantly to the global burden of disease, and currently it is estimated that about 450 million people worldwide suffer at a certain point of a neurological, psychiatric or behavior related disease (WHO, 2001). The lifetime risk for the worlds population of developing at a certain moment of their life a psychiatric or behavioral disorder is estimated to be around 25% (WHO, 2001). Lifetime prevalence rates for any kind of psychological disorder are higher than previously thought, the disorders are increasing in recent cohorts and affect nearly half the population. And yet, the true burden of disease is likely to be far greater, as there is a large proportion of mental problems that go unnoticed, and there is a widespread lack of recognition of the link between mental disorders and other health conditions. Despite being common, mental illness is under diagnosed by doctors, and less than half of those who meet diagnostic criteria for psychological disorders are identified by doctors (WHO, 2008a). Patients, too, appear reluctant to seek professional help. Only 2 in every 5 people experiencing a mood, anxiety or substance use disorder, seek assistance in the year of the onset of the disorder. It is noteworthy that in low and middle income countries mental problems are frequently given low priority and communicable diseases, maternity and child health are among conditions that are ranked higher (Prince et al., 2007). And yet 85% of the worlds population lies in countries that according to the World bank criteria are classified as low-income or middleincome countries (Jacob et al., 2007). Mental disorders contribute to severe morbidity, to long-term disability, and to mortality. According to WHO (2005a) neuropsychiatric disorders contribute to 31.7% of all years lived with disability (DALY), with unipolar depression ranking the highest (11.8%), followed by alcohol related conditions (3.3%), and schizophrenia with (2.8%). Among the ten most important diseases measured by YLD (Years of Life lived with Disability) psychiatric conditions comprise four, namely uni-polar depression, alcohol abuse, schizophrenia and bipolar disorder (WHO, 2001). The burden of neuropsychiatric disease is greater in upper middle-income countries than lower middle-income countries that for their part is higher than in low-income countries (Jacob et al., 2007), and in the European Union, psychiatric conditions constitute 25% of the total burden of disease (Andlin-Sobocki et al., 2005) (> Tables 86-1 and 86-2).
3
Main Diagnostic Groups
3.1
Schizophrenia
Schizophrenia is the most disabling and serious psychiatric disorder and is found in all cultures. Today, it is estimated that schizophrenia comprises 1.1% of the global burden of disease and 2.8% of YLD.
1475
1476
86
Global Burden of Mental Health
. Table 86-1 Leading causes of disability-adjusted life years (DALYs) 15–44 years old, estimates for 2000 Males 15–44 years
% total
Females 15–44 years
% total
3. Unipolar depressive disorders
6.7
2. Unipolar depressive disorders
10.6
4. Alcohol use disorders
5.1
5. Schizophrenia
2.8
7. Self-inflicted injury
3.0
7. Bipolar affective disorders
2.5
8. Schizophrenia
2.5
9. Bipolar affective disorders
2.4
16. Drug use disorders
9. Self-inflicted injury 14. Panic disorder
2.4 1.6
1.6
The table shows how the mental disorders are to be found among the leading causes of life years adjusted for disability in both sexes with unipolar depression being the top cause in both sexes. Source: WHO (2001)
. Table 86-2 Leading causes of years lived with disability (YLD) in males and females, estimates for year 2000 Males all ages
% total
Females all ages
% total
1. Unipolar depressive disorders
9.7
1. Unipolar depressive disorders
14.0
2. Alcohol abuse disorders
5.5
6. Schizophrenia
2.7
7. Schizophrenia
3.0
7. Bipolar affective disorders
2.4
9. Bipolar affective disorders
2.6
16. Dementias & Alzheimer
1.8
20. Drug abuse disorders
1.6
9. Dementias & Alzheimer 20. Panic disorder
2.3 1.6
The table show how the mental disorders contribute as leading causes of years lived with disability in both sexes Again unipolar depression stands at the top. Source: WHO (2001)
One of the largest epidemiological studies of mental health disorders is the World Health Organization investigation carried out in the 1960s in which the incidence and course of schizophrenia were compared in ten different countries (the IPSS project and DOSMED project) (Jablensky et al., 1992). This investigation demonstrated similarities in the presentation form and incidences of schizophrenia across cultures using a narrow delineation of schizophrenia, but the differences of these variables increased significantly when using a wider delineation of schizophrenia from 4.2 per 100,000 in Chandigarh, India to 1.6 per 100,000 in A˚rhus (Jablensky et al., 1992). In the European region it is estimated that 3.7% million persons suffer from schizophrenia (Olesen et al., 2006). In the WHO Outcome study, the outcome of schizophrenia was found more favorable in developing countries compared to highly industrialized countries (Jablensky et al., 1992), and the differences in outcome were interpreted as due to genetic factors as well as specific environmental conditions. This more favorable prognosis in developing countries has recently been challenged. Later it has been reported that also countries as Japan and Singapore report a more favorable outcome of schizophrenia which could be attributed to their ability to adhere to a traditional culture despite high technological development (Jilek, 2001).
Global Burden of Mental Health
86
Considerable interest has also been paid to the high preponderance of schizophrenia among Afro-caribian immigrants to UK (Bhugra, 2004). No simple explanation has been found for this phenomenon, but issues like misdiagnosis, migratory stress and increased tendency of substance abuse have been mentioned. The preponderance is however likely to be linked to an interaction of biological and sociocultural factors as well as increased stress. The preponderance of schizophrenia in migrants have also been reported in a Danish register study in particular among young male descendants of migrants (Helweg-Larsen et al., 2007). Brief psychotic episodes are reported worldwide, but with a tendency to increase in frequency, the less technologically developed the society, and the psychotic episodes are frequently found to be related to sociocultural factors.
3.2
Affective Disorders
Affective disorders comprise a group of disorders characterized by significant mood disturbances. The burden of depression depends upon region having a relatively smaller burden in poorer regions. The prevalence of two of the major disorders of the group: major depression and bipolar disorders amounts to 21 million persons in the European region (Olesen et al., 2006). For example depression amounts to 1.2% of the total burden in Africa to 8.9% in high-income countries, but depression in the developing countries is predicted to become a leading cause of disease burden in these regions as well (Murray and Lopez, 1997). Uni-polar depression has a point prevalence amounting to 1.9% for males and 3.2% for females; and 5.8% of males and 9.5% of females are likely to develop a depressive episode within a 12-month period. Every year 5–8% of the adult population gets a depression (AndlinSobocki et al., 2005). It is estimated that the lifetime risk for a severe depression amounts to 12–16%. The frequency of depression may be measured in the community using well-defined diagnostic categories and reliable operational criteria. WHO carried out a study on mental illness in general health care comprising 14 countries around the world. A main finding was that the prevalence of current depression ranged from 2.6% in Nagasaki, 4.0% in Shanghai to 16.9% in Manchester and 29.5% in Santiago (Goldberg and Lecrubier, 1995). Four possible explanations were given as contributing to this finding: namely that there indeed are true differences in depression; that the concept of disease may differ according to culture; that there are differences in help-seeking behavior; and finally that the demographic characteristics of the studied populations differ (Goldberg and Lecrubier, 1995). Symptoms like lowered mood, lack of joy, anxiety, lack of energy and interest in surroundings are reported in most cultures, and depressive disorders occur at different levels of frequency in all known cultures (Jablensky et al., 1992). WHO reports of quite similar symptomatology in different cultural contexts, but considerable variation in prevalence, and the question is whether these differences are due to differences in social or cultural conditions, and the cultural context may have a pathoplastic effect on the manifestations of the depression (Tseng, 2001). One should keep in mind that in many cultures, depressive traits may be interpreted as a natural reaction following losses or other severe life events. Also severe pathological grief reactions
1477
1478
86
Global Burden of Mental Health
can be seen as a sign of genuine grief more than as a pathological phenomenon, which may lead to the situation that persons suffering from major depressive episodes may not receive adequate treatment. Also depressive traits in old age may be interpreted as a condition not requiring any therapeutic intervention. Thus, it is important to recognize the interaction between culture and depression whenever interviewing a patient suspect of a depressive episode. The measures of disability may differ but persons with depressive disorders demonstrate impaired physical health, impaired level of functioning, increased work impairment and high utilization of health services (Lecrubier, 2001). Race or ethnicity do not significantly influence the rate of occurrence of depression. Socioeconomic or educational differences may contribute to some differences observed between ethnic groups, but on statistical correction of these factors, there is no variation in risk by ethnic groups (WHO, 2008b). WHOs investigation of depression in five countries showed that signs of tristesse, lack of joy anxiety and tension were central in all settings. In Switzerland guilt feelings were common in contrast to Iran whereas somatic complaints were more common in Iran compared to Canada. It appears that guilt feelings are more common among persons of Jewish-Christian faith (Sartorius et al., 1983). The variation in prevalence was not easily explained, but women had generally a higher prevalence Lifetime prevalence varied between 1.5% in Taiwan to 19% in Beirut (Wittchen et al., 1999).
3.3
Nervous Disorders
Anxiety disorders are a major group of the nervous disorders. Anxiety problems comprise panic disorder, generalized anxiety, phobic disorders, OCD and PTSD. Anxiety disorders comprise the largest group of persons with mental distress worldwide. About one third of absenteeism due to sickness is due to anxiety disorders. In the large US Epidemiological Catchment Area study (Tseng, 2001) general anxiety was reported in 2.7% of Afro-Americans and in 1.6% of Anglosaxon-American (1 year prevalence). In the European region the anxiety disorders including OCD comprise approximately 41 million persons and the 1-year prevalence amounts to 12% of the population (Olesen et al., 2006). In China it is reported that 4 out of 1,000 men and 39 out of 1,000 women suffer from anxiety (Tseng, 2001). Anxiety problems are frequently found in general practice – ranging from 22% in Rio de Janeiro to 1% in Ankara. Particular interest has been paid to the frequency of Post traumatic stress disorders (PTSD) in recognition of that an increasing proportion of war related deaths and war-casualties are civilians, and women are in particular risk. It is estimated that 80% of the 50 million people affected by violent conflicts, civil wars, disasters, and displacement are women and children. Lifetime prevalence rate of violence against women ranges from 16 to 50% (WHO, 2008a). The symptomatology of PTSD shows clear communalities across culture, but over the years the delineation and overall existence of the diagnosis have been the subject of repeated discussion. The need for treatment is also closely linked to the cultural background, coping strategies, political/religious conviction and existence of social network. The prevalence of PTSD following traumatic events vary considerably but is generally reported between 9 and 37% (Ekblad and Jaranson, 2004). In the general population PTSD
Global Burden of Mental Health
86
point prevalence amounts to 0.4% and the life-time prevalence to 10.4% in women and 5% in men (Ekblad and Jaranson, 2004).
3.4
Abuse
Abuse is a pattern of symptoms linked to the use of the substance with impaired capacity to control the abuse, a tolerance of withdrawal from preoccupation with substance and continuing use of the harmful substances. Alcohol abuse is estimated on a worldwide basis to comprise about 76 million people and the point prevalence is deemed to be 1.7% globally, 2.8% in men and 0.5% in women. The lifetime prevalence shows similar gender ratio with reported prevalences as 31 and 7% respectively in males and females in Canada and 43% compared to 3% in South Korea. The preponderance of males is reported in general and across cultures, both with respect of consumption as well as physical and social consequences (WHO, 2001). Gender differences however seem to diminish in some regions with womens’ increasing educational background. Regional differences in the accepted behavior and attitudes to the abuse of ethanol are significant which has a heavy impact on help-seeking behavior and the socio-cultural context plays a great role in the overall consequences of the abuse (Tseng, 2001). The prevalence in Europe amounts to 9 million people abusing alcohol and illicit substances (Olesen et al., 2006). Increase in alcohol consumption is often seen following societal disintegration and destabilization as seen in the former Soviet Union. But alcohol abuse is growing rapidly in low-income and middle-income countries – in particular among males – and in these regions many may avoid to seek help either because of shame or because of ignorance about treatment facilities (Patel et al., 2007). Thus, we see large ranges in the frequency of alcohol abuse revealed in consultations to general practitioners, in a study of general practice we saw approximately 10% of patients from large cities like Paris, Seattle or Berlin compared to 2% in Ankara (Kastrup and Baez, 2007). Alcohol abuse is in the global burden of disease project estimated as the cause of about 1.5% of all deaths. Alcohol ranks high in the burden of disease with 3.5% of all DALY’s (WHO, 2001).
3.5
Suicide
About 877,000 people die by suicide every year, and suicide is the 13th leading cause of death globally (WHO 2005), the median suicide rate being 6.55 per 100,000 population (Jacob et al., 2007). WHO (2001) have reported age-standardized rates of suicide in 24.0 males per 100,000 and 6.8 females per 100,000. Suicide is globally on the increase and we know that suicide is linked to psychosocial stressors, family structure and abuse. Certain groups run a particular risk due to e.g., migration, stay in prison or due to a feeling of lack of coherence with the sociocultural group. In fact, sociocultural factors as well as biological and psychological ones all contribute to suicidal behavior. It is also well known that there is a close link between suicide and the existence of mental disorders and about 15% of those suffering from depression is estimated to commit suicide at a certain point (WHO, 2008b). This has also been demonstrated among others in studies using case registers (Helweg-Larsen, 2006). Suicide trends has varied considerably over time. Thus
1479
1480
86
Global Burden of Mental Health
whereas we see a steady decline in recent decades in many Western countries, clear increases are reported in other parts of the world, e.g. India, Mexico and the former Soviet Union (Helweg-Larsen, 2006; WHO, 2001). In many low-income and middle-income countries suicide is now one of the leading causes of death among the young, and in the European region among those aged 15–44 years, suicide is the second most frequent cause of death. Presently suicide rates for males are above 35 per 100,000 in Belarus, Kazakhstan, Lithuania, Russia and Ukraine (Jacob et al., 2007). Regional variation is also reported when it comes to suicide attempts that are frequently seen in younger women and in persons marked by lack of stability or poverty. The experience of alienation faced with a new culture or torn between the demands of two cultures may explain the tendency to suicide attempts documented in certain ethnic groups and also among young internationally adopted (Helweg-Larsen et al., 2007).
4
Factors Related to Mental Disorders
A number of factors are associated with the presence, the kind and the course of mental and behavioral disorders and include e.g., migration, poverty, gender, conflicts and catastrophs, concomitant severe physical disorders, urbanization and family history (> Table 86-3).
. Table 86-3 Presence of mental health programs for special populations in the world Mental health programs for special populations Minority groups
Countries % 16.5
Refugees
26.2
Disaster affected populations
37.7
Indigenous people
14.8
Elderly persons
50.5
Children
62.4
This table shows that only a small proportion of countries worldwide have developed mental health programs focusing on special populations that frequently are more vulnerable, such as refugees. Source: WHO (2005b)
4.1
Migration
Among these factors the role of migration on mental health deserves particular attention. Migration is influenced by an interaction of several different interactive factors (Bhugra and Becker, 2005). On a pre-migratory level, factors such as psychological, social and biological vulnerability, political conditions in home country, education etc all are contributory. The time after the migration may often be marked by several losses (e.g., identity, family, language, values, social network), insecurity, social, work or economic problems, language isolation or adjustment to the new life circumstances. Migration implies a feeling of alienation and despair and it may imply an increased risk to develop depressive disorders.
Global Burden of Mental Health
86
It has been brought forward that those who migrate are the particularly gifted, robust individuals, who are rich in initiatives. This could explain why the morbidity is frequently low particularly immediately following the migration. Migration research that has been broad in its approach has hitherto primarily related to the emergence of psychotic disorders in particular schizophrenia (e.g., Cantor-Graa et al., 2003) The importance of the interaction between migration for the development of a depression is not evident (Bhugra, 2003) and investigations and findings that relate to other mental disorders are far less clear, and they often relate to restricted populations. It has been demonstrated that limited social contact is a main contributory factor for the increased frequency of depression among migrants compared to native Norwegians and investigations have reported a relation between lack of a sense of coherence and the depression in young refugees (Roth and Ekblad, 2006). A factor, that may be contributing is that persons who migrate from collectivistic, sociocentriske societies to egocentric societies may experience alienation and worries, that makes adjustment to new societies difficult (Bhugra and Becker, 2005). There is increasing evidence that not only migrants are exposed to stress also their descendants run an increased risk of mental disorders (Helweg-Larsen et al., 2007). Similarly international adoptees have been found to have an increased prevalence of mental disorders (e.g., Helweg-Larsen et al., 2007).
4.2
Poverty
It has been demonstrated persons that with low education and low income had a particular vulnerability to common mental disorders, no matter what society we deal with. Results from low-and middle-income countries in Latin America had shown a relationship between high rate of mental illness and poor education (Saxena et al., 2007).
4.3
Comorbidity with Other Health Conditions
Approximately 15% of patients seen in primary care exhibit medically unexplained somatic conditions linked with psychological distress (Prince et al., 2007), and somatization is seen to increase significantly health care costs. Furthermore, adding to the burden and disability is the fact that persons with mental disorders have a high degree of co-morbidity with both physical and mental disorders.
4.4
Gender Issues
It has again and again been documented that gender is a major determinant of mental health and mental illness, and that depression, anxiety, psychological distress, violence occurs more frequently in women across countries. Gender has decisive impact on social position, status, opportunities, and in some areas female gender per se prevents access e.g., adequate mental heath care. With respect to depression, women have on a global level a 1.5–2 times higher risk of getting a depression compared to men. In all countries of the Cross National Collaborative Group women had a higher prevalence ranging from a ratio female: male of 3.1 in West Germany to 1.6 in Beirut and Lebanon (Weissman et al., 1996).Women are under pressure due to their many roles and obligations, their risk of violence and sexual abuse and the
1481
1482
86
Global Burden of Mental Health
discrimination encountered in many countries (WHO, 2008a). The increased risk among women to develop PTSD is repeatedly reported and yet findings of gender differences in the prevalence of PTSD among non-Western adults exposed to political violence are inconsistent. There is little gender difference in the prevalence of schizophrenia, but it manifests earlier in males. It is also documented that males have an about two times higher life-time prevalence rates of alcohol dependence, and men are more likely to be diagnosed with antisocial personality disorder. Gender also has a critical input on compliance to treatment, social adjustment and longterm outcome. In case of war, disaster or exile, women fulfill the role of nurturers and providers of emotional support, and they have usually the main responsibility for care giving in the family. Women’s capacity to cope may be overloaded, as preoccupation with the needs of the family may lead to that they are not able to consider their own needs. On the other hand the care giving role may have a protective function as women contrary to their partners have a natural role and identity in the new environment.
5
Prevention
Despite the fact that severe mental disorder take an increasing toll also in low-income and middle-income countries, preventive trials focusing on how to prevent the psychotic disorders have not been carried out in these regions (Patel et al., 2007). In the preventive research emphasis should be placed on interventions that focus on the need of the low-income and middle-income countries. Combining medication, psycho-education and support to the family the relapse rate of psychotic manifestations can be reduced from about 50% to less than 10% after a year which could be seen as a very cost-effective intervention (WHO, 2001).
6
Consequences of Globalization
We live in a world that undergoes drastic changes, and it is essential to avoid that walls are built between modern societies and traditional cultures, whereby certain cultural groups are excluded from participating in the societal development (Brundtland, 1999). As Brundtland points out globalization is a challenge that some may take up more than others. Those that come from less developed regions may face a future that is so different from the world in which they grew up, that most of the skills they have taken over from previous generations are of no use to cope with the tasks of a globalized world. In the age of globalization, one definition of culture deserves attention, namely that it is an ever-changing construct that emerges from interactions between individuals, communities and institutional practices (Kirmayer, 2001). In that context culture plays an important role on the symptoms and manifestation of mental problems, the explanation models used, the coping mechanisms and help seeking behavior, and the social response to mental distress and disability (Kirmayer, 2001). The question is frequently asked whether we are ready for this global culture. For certain groups globalization will have as a consequence that they cannot cope with the changes
Global Burden of Mental Health
86
and thereby their mental health will be influenced. The demands for change and adjusting to the consequences of globalization are particularly pronounced among the populations of the developing world. On a global level, such groups run an increased risk of developing stress related mental problems, both in their home country and in relation to migration. This is an issue that in the light of the global burden deserves further attention.
6.1
Cultural Competence
As a consequence of globalization mental health professionals are increasingly confronted with multicultural patient populations. A certain proportion of them may have had contact with alternative approaches of treatment or indigenous healers partly because this approach to treatment may reflect their belief systems and explanatory models better, and partly because they may have experienced limited access to mental health services. Today, the majority of mental health professionals working in urban settings will encounter such patients in their daily clinical life (Kastrup, 2008). It is important that mental health care staff possesses competencies to deal with this diversity of patients. The question remains whether the training about appreciation of cultural aspects will enter medical curricula and whether those who provide the training will consider this sufficiently medically relevant (Fox, 2005). Few have outlined the aspects of cultural competence better than Tseng (2003) who has presented some of the needed ingredients including the need to focus on cultural sensitivity and acknowledge the existence of different lifestyles and attitudes, and accept them without prejudice; the acquisition of cultural knowledge; the need for cultural empathy and understanding the patients own perspective; the recognition of culturally relevant relations and interactions with consideration of cultural background; and cultural impact on help-seeking, Mental health professionals in Western societies face a doctor-patient relationship that has undergone a considerable transition. Previously, the patient accepted the authority of the doctor, and an asymmetrical position characterized the relationship. Today, the relationship between therapist and patient is strongly focused on informed consent with the aim to promote patient autonomy and permit the individual patient to make rational decisions. The responsibility for the treatment is in some way shared, and it may be described as a kind of contract where the two agents negotiate conditions and come to an agreement (Kastrup, 2008). The outcome of disorder is mainly described as self-determined and the patient is encouraged to take an active part. In contrast to this we see the traditional approach that is often acknowledging dependence on God (or other agents outside the individual) regarding health and illness and stress the spiritual/divine role for the outcome (Okasha, 2000). Communication with female patients may in some traditional settings be very authoritarian, thereby making it difficult for women to tell about their emotional distress. Distress may be interpreted as a critique of her husband or male members of the family which may result in that these women are either under- over over-treated.
7
Health Economics
Since the World Bank and WHO in the 1990s carried out their analyses of the Global Burden of Disease (GBD) the attention to the health economic consequences of mental disorders
1483
1484
86
Global Burden of Mental Health
has increased. Low-income countries spend significantly less of their Gross Domestic Product on health than middle-income countries and they in turn spend significantly less than high-income countries. The proportion of the health budget allocated to mental health varies in the WHO regions from 0.76% in the African, over 2% in the Western Pacific and the Eastern Mediterranean region, to 6.3% in the European region, or when distributed according to income level from 1% in low-income over 2.15% in lower-middle income, to 3% in uppermiddle income, and 6.8% in high-income countries (Jakob et al., 2007). Availability of mental health beds differ similarly with 0.24 per 10,000 in the low-income countries over 1.59 in lower-middle income to 7.7 in upper-middle and 7.5 per 10,000 in highincome countries (Jakob et al., 2007) (> Table 86-4).
. Table 86-4 Income group and psychiatric beds, proportion in mental hospitals and number of psychiatrists Income group of countries
Median no psychiatric beds per 10,000 pop
Proportion psychiatric beds in mental hospitals
Median no psychiatrists per 100,000 pop
Low
0.24
74.4
0.05
Lower middle
1.59
82.7
1.05
Higher middle
7.70
78.8
2.70
High
7.50
55.0
10.50
This table shows how the number of psychiatric beds, and how many of them are placed in mental hospitals differ in the different income group of countries. Similarly with the number of psychiatrists in the different income group of countries. Source: WHO (2005b)
The European Brain Council too has recently published an overview of the impact of mental disorders on the European health economy. Projections towards 2020 show that the neuropsychiatric disorders will increase in proportion from the present 10.5% of the total burden of disease in 1990 to 15% in 2020. As a consequence, these disorders have achieved a high ranking on the international health political and economic agenda. Most middle-income and low-income countries devote less than 1% of their health expenditure to mental health, and many also lack mental health policy and legislation (Jakob et al., 2007). Till now the vast majority of research on effectiveness of mental health services are carried out in high-income countries and it is not necessarily so that results may be extrapolated to low-income countries (Saxena et al., 2007). But even within the same income level, components of mental health care differ greatly (Jakob et al., 2007). But the increased focus has not been guaranteeing that persons with mental disorders in the developing world will have a more fortunate destiny, and there is still a significant distance between the size and amount of the illnesses and the possibility for adequate treatment hereof, not the least in the non-industrialized countries where access of psychiatrically trained staff is very limited.
Global Burden of Mental Health
8
86
Therapeutic Issues
According to the survey of Patel et al. (2007) who identified 11,501 trials globally focusing on the assessment of therapeutic or preventive interventions for mental health problems, the majority of the evidence stems from high-income countries. Less than 1% were carried out in low-income countries. Breaking the studies according to diagnostic groups it is characteristic that about 3 out of 4 trials in low-income or middle-income countries were dealing with the treatment of schizophrenia and about 1 out if 4 with depression, whereas other disorders were hardly focused upon (Patel et al., 2007). And yet examples do exist of effective community-based interventions for psychotic and depressive disorders, programs that are feasible and affordable in such settings. Patel el al. (2007) in their survey emphasize as examples hereof primary care depression program in Chile and community-based program for rehabilitation of chronic schizophrenics in rural India. But we also have limited experience with the integration of treatment programs developed in high-income countries, in low-income or middle-income countries with different organization and availability of services. As regards treatment of depression, group psychological interventions were reported to be efficacious in some trials in low-income and middle-income countries, and this finding was related to the fact that this kind of intervention may resemble traditional social mechanisms such as support through collective action (Patel et al., 2007).
9
Primary Health Care
About 20% of all patients who consult the primary health sector have a psychiatric disorder, but a considerable part of them are never diagnosed or treated properly (WHO, 2001). To overcome the treatment gap one strategy is to train, assist and supervise primary care physicians to identify and treat mental illness (Saxena et al., 2007). This training need to be followed by regular supervision in order to provide mental care in low-income countries.
10
Treatment Gap
Since the publication of the three reports on global mental health (World Bank, 1993, Murray and Lopez 1996, WHO, 2001) the economic and social consequences of mental illnesses has been increasingly recognized. Worldwide there is a vast treatment gap between those needing care and those receiving adequate care for their mental disorders. This gap differs according to region and whereas it is estimated to 35–50% in developed countries, it is 76–85% in low-and middle-income countries (WHO, 2004). Even in the US it is estimated that 2/3 of those with mental disorders are not treated (Kessler et al., 2005). In Europe up to 74% of those suffering from mental disorder are not treated (Wittchen and Jacobi, 2005). According to the European Ministerial Report (WHO, 2005a) the treatment gap in Western Europe is estimated to 17.8% for schizophrenia, 45.4% for severe depression, and 62.3% for generalized anxiety (> Table 86-5). In low-income and middle-income countries we may see that even among those treated, only a small proportion are adequately treated (Lancet Global, 2007). Thus, the increased
1485
1486
86
Global Burden of Mental Health
. Table 86-5 Median percentage treatment gap for selected psychiatric disorders based on community based surveys Psychiatric disorder
Median rate untreated
Schizophrenia and non-affective psychotic disorders
32.2
Major depression
56.3
Bipolar disorders
50.2
General anxiety disorders
57.5
Alcohol abuse
78.1
This table shows how large a proportion of persons with a particular disorder are left untreated. Source: Kohn et al. (2004)
focus on the burden of mental health has not ensured better conditions for the mentally ill in the low-income and lower-middle income countries (Kastrup and Baez, 2007). Community based services vary considerably from region to region. Thus provision of community based care ranges from 97% in high-income countries to 52% in low-income countries (Saxena et al., 2007). Saxena et al. (2007) further point out that middle income countries that have invested in the creation of large mental hospitals may have little incentive in establishing community health care (> Table 86-6). . Table 86-6 Presence of community mental care in the WHO regions, income group and the world WHO regions
Countries %
Low-income countries
51.7
Lower middle-income countries
51.9
Higher middle-income countries
90.9
High income countries
97.4
This table shows to what extent community care is available in the different income group of countries. Source: WHO (2005b)
Many countries do not have a mental health budget, and in African and South Asian countries less than 1% of the health budget is allocated to mental health (World Mental Health Atlas, 2005b). But also in the European region we see a relatively small proportion of the health budget allocated to mental health, thus among 24 European countries 5.8% of their Gross National Product is allocated to mental health, even though the disorders comprise 20% of the burden of disease (WHO, 2005a). Many low-income countries prioritize inadequately the development of mental health policies despite the fact that the consequences of mental disorders show an increasing proportion of the burden of disease. The human resources allocated to mental health are unevenly distributed at a global level. According to the WHO Mental Health Atlas (2005b) the amount of psychiatrists in low-income countries are 0.05 per 100,000 population to 1.05 in lower-middle to 10.5 in
Global Burden of Mental Health
86
high-income countries. Within a given country, the tendency is that psychiatrists and other mental health professionals live in larger urban areas which result in low access to care in the rural population. In settings with inavailability of mental health professionals one should bear in mind that specialist mental health staff in such settings preferably could concentrate on supervising less specialized staff instead of acting as clinicians, and that non-professionals need to be part of the mental health programming of the communities (Saraceno et al., 2007). Non-formal community interventions are needed and community members should be mobilized if we are to overcome the treatment gap in low-income countries. Furthermore, it needs mentioning that the shortage of trained mental health professionals also prevent many countries to introduce psychosocial interventions, psychotherapy, etc. Low-income countries are faced with another obstacle as the cost of the necessary drugs are relatively more expensive in these countries. According to Saxena et al. (2007) 1 year supply of one of the least expensive antidepressants only costs twice as much in high-income compared to low-income countries despite a 12.5 time difference in the Gross National Product. Prescribing antipsychotic medication for the treatment of schizophrenia is a key component of treatment of the disorder, but in low-income and middle-income countries, the treatment gap for schizophrenia is large and resources for medication low (Patel et al., 2007). Consequently in such countries it may be difficult to provide sufficient and adequate medication, and the long-term outcome of chronic psychoses may be less favorable.
11
Ethical Concerns
As a consequence of the increased globalization the debate is ongoing whether there should be a universal set of ethical guidelines or whether ethical guidelines rather should reflect a given cultural context and vary accordingly (Sartorius, 2000b). Should e.g., the guiding principles of the Declaration of Madrid be interpreted according to the setting in which they are applied? Or according to the cultural background of the mentally ill patient in front of us? The danger of having a culture relativistic approach in psychiatry is that it implies that all cultures and values could be seen as equally good, and that there is no such thing as absolute global moral values. Global mental health is to a large extent a question of political and economic welfare. But to what degree is mental health a fundamental human right? Access to psychiatric help is a close function of economic differences inter as well as intra nationally, and access to adequate psychiatric service is still inaccessible in larger parts of the world. And we should not forget that most psychiatrists are treating patients who belong to a different social and educational background than that of their psychiatrists. The economic disparities both between and within nations are reflected in the availability of adequate mental health care but the moral need for an equitable allocation as stressed in the Declaration of Madrid moves further away if indefinite progress in certain regions is given priority over the implementation of limited achievable goals globally. If we want to see to that the intentions of the Declaration of Madrid in their entirety are fulfilled on a worldwide basis, we should formulate mental health policy goals that to a greater extent pay attention to the problems of the poor, and use goals that are expressed in terms of enhancement of equity than what is seen today.
1487
1488
86
Global Burden of Mental Health
We are confronted with a difficult balance how to adhere to the universal standards and at the same time not disregard local values of our target population. And, on the other hand, that ensuring respect for the local culture should not become a pretext for bypassing or ignoring ethical guidelines. We have to weigh the different values and consider the diverse views in the modern multicultural world (Okasha, 2002). Judgments on values of the past cannot stand alone, and we take the challenge facing us and reconstruct the world and create perspectives we can live with (Okasha, 2002).
12
Perspectives
Changes in demographic distribution have consequences for the future projection of the burden of mental disorders. Thus the ageing of certain Western populations like the European one implies that it is likely that we shall see an increasing proportion of mental disorders. In Europe neuropsychiatric diseases are estimated in 2020 to increase to 15% of the total burden and amount to 40% of all chronic diseases (WHO, 2005a).The Global Burden of Disease study estimates that major depression would rank second among the leading ten causes of DALY in 2020 (Murray and Lopez, 1997). Most mental health programs have primarily targeted local populations and may succeed locally but till now we only have few that have undergone systematic assessments of their effectiveness in large scale populations (Patel et al., 2007). Most of the evidence for interventions unfortunately has little relevance for settings in low-income and middle-income countries. Here it is needed to focus on interventions that are relying on low-technology and mental health workers who are not specialists as such interventions may be as efficient in such contexts (McKenzie et al., 2004). In low-and middle-income countries we however still see that approximately 80% of mental beds are found in large mental hospitals (Jacob et al., 2007) and that despite recommendations for deinstitutionalized care there are large obstacles preventing this development. Some of them are found among mental health staff themselves, and there is in many settings still a culture of paternalism and a lack of belief in the ability of the mentally ill to make proper decisions (Saraceno et al., 2007). Saraceno and co-workers (2007) outlined several obstacles to the development of services. The public health agenda need to prioritize mental health, and the move away from centralized mental health hospitals towards community-based services. Further, primary care needs to integrate mental health care and the lack of human resources, clinically as well as organizationally, is pertinent. In a survey on mental health priorities on a global level (Prince et al., 2007) respondents identified a number of reasons why mental health is low on the public agenda. These included that advocating on mental health may not easily be understood by non-mental health practitioners, but also that persons suffering from mental health problems and their families are often hidden and not outspoken. Further, that the indicators for mental health may be weak and that the general public’s interest in the life and well-being of mentally disordered limited (Saraceno et al., 2007). Surveys of recent research on mental health consistently document that almost all research in this field takes place in high-income countries (Patel and Sumathipala, 2007).
Global Burden of Mental Health
13
86
Conclusion
A dialogue with all stakeholders involved in global mental health is needed in recognition of that worldwide health systems are confronted with great tasks if they are to provide adequate mental health care and ensure the protection of the human rights of the mentally ill (Lancet global, 2007). In the three high-profiled international reports on the burden of mental health (Murray and Lopez, 1996; WHO, 2001; World Bank, 1993) it is recommended that availability of mental care be increased. Furthermore, that provision of better and increased mental health be provided through the development of human resources and mental health policies. But despite such reports, the Lancet series on global mental health have pointed out that scaling-up of initiatives to improve mental health services have been slow (Saraceno et al., 2007).Unfortunately, mental health is almost absent from the international agenda for public health and mental health lobbyists may either work against one another as they focus on different mental health problems each needing different solutions or they may offer different views on the type of intervention needed (Saraceno et al., 2007). If we want that persons living in low-income and middle-income countries should get access and receive help for their mental health problems, a help that is sufficient, affordable, accessible and acceptable heavy tasks lie ahead and the health resources have to be prioritized differently. As pointed out by the Lancet Global Mental Health Group, change in public health only comes about if three core elements are present: a knowledge base, strategies to implement what we know, and the political will to act.
Summary Points Mental health problems are linked with poverty, stigmatization, with impact on physical health and general mortality of those afflicted and their families.
Mental disorders have early onset compared to many other long-lasting diseases and
contribute significantly to global burden of disease. Approximately 450 millions have at a certain point a neurological, or psychiatric disease. Neuropsychiatric disorders contribute to 31.7% of years lived with disability. Unipolar depression, alcohol abuse, schizophrenia and bipolar disorder belong to the top ten diseases measured by years lived with disability. The burden is greater in upper middle-income countries and lowest in low-income countries. Schizophrenia comprises 1.1% of the global burden of disease. Depression amounts to 1.2% of the total burden in Africa, 8.9% in high-income countries. Suicide is the 13th leading cause of death globally with 877,000 annually. Migrants and their descendants run an increased risk of mental disorders. Depression, and anxiety are more common in women, alcohol dependence, antisocial personality disorder more common in men. Culture plays an important role on manifestation of mental problems; explanatory models; coping mechanisms and help seeking behavior.
1489
1490
86
Global Burden of Mental Health
The proportion of the health budget allocated to mental health varies from 1% in lowincome to 6.8% in high-income countries.
Mental health beds ranges from 0.24 per 10,000 in the low-income, 1.59 in lower-middle income, 7.7 in upper-middle and 7.5 in high-income countries.
The treatment gap differs from 35–50% in developed, to 76–85% in low-and middleincome countries.
References Andlin-Sobocki P, Jo¨nsson B, Wittchen H-U, et al. (2005). Eur J Neurol. 12(suppl 1). 1–27. Bhugra D. (2003). Acta Psychiatr Scand 108: 67–72. Bhugra D. (2004). Acta Psychiatr Scand. 109: 243–258. Bhugra D, Becker M. (2005). World Psychiatr. 4: 18–24. Brundtland GH. (1999). International consultation on the health of indigenous peoples. Speech at Sweet Briar College. Cantor-Graae E, Pedersen CB, McNeil TF, Mortensen PB. (2003). Br J Psychiatr. 182: 117–122. Ekblad S, Jaranson J. (2004). In: Wilson J, Drozdek B (eds.) Broken Spirits. Brunner & Routledge, New York. Fox RC. (2005). New Engl J Med. 353: 1316–1319. ˆ stuu¨ TB, Goldberg D, Lecrubier Y. (1995). In: U Sartorius N (eds.) Mental Illness in General Health Care. Wiley, Chichester. Helweg-Larsen K. (2006). In: Danish (ed.) Suicide in Denmark. Significant Decrease in Suicide, But Increasing Number of Suicide Attempts, Why? The National Institute of Public Health, Copenhagen. Helweg-Larsen K, Kastrup M, Baez A, Flachs E. (2007). Etniske forskelle i kontaktmønstret til psykiatrisk behandling. Et registerbaseret studie (Ethnic differences in contact patterns A register based study). Videnscenter for Transkulturel Psykiatri, Copenhagen. Jablensky A, Sartorius N, Ehrenberg G, Anker M, Corten A, Cooper JE, Day R, Bertelsen A. (1992). Psychol Med Monogr Suppl. 20: 1–97. Jacob KS, Sharan P, Mirzo I, Garrido-Cumbrega M, Seedat S, Mari JJ, Screenivas V, Saxena S. (2007). Lancet. 370: 1061–1077. Jilek W. (2001). Cultural Factors in Psychiatric Disorders. Paper 26th Congress of World Federation of Mental Health. Available at: http://www.mentalhealth.com/ mag1/wolfgang.html/april 2002. Kastrup M, Baez A. (2007). Dan Med Bull. 54: 42–43. Kastrup M. (2008). Eur Psychiatr 23(Suppl 1): 59–68. Kessler R, Demler O, Frank RG, et al. (2005). New Engl J Med. 353: 2515–2523.
Kirmaye LJ. (2001). J Clin Psychiatr. 62(suppl 13): 22–28. Kohn R, Saxena S, Levav I, Saraceno B. (2004). Bull WHO. 82: 858–866. Lancet Global mental Health Group. (2007). Lancet. 370: 2241–2252. Lecrubier Y. (2001). J Clin Psychiatr. 62(Suppl 8): 4–9. McKenzie K, Patel V, Araya R. (2004). BMJ. 329: 1138–1140. Murray CJ, Lopez AD. (1996). The Global Burden of Disease: A Comprehensive Assessment of Mortality and Disability from Diseases, Injuries, and Risk Factors in 1990 and Projected to 2020. Harvard School of Public Health, Cambridge. Murray CJ, Lopez AD. (1997). Lancet. 349: 1498–1504. Okasha A. (2000). In: Okasha, et al. (eds.) Ethics Culture and Psychiatry. APA Press, Washington. Okasha A. (2002) In: Sartorius N, et al. (eds.) Psychiatry in Society. Wiley, Chichester. Olesen J, Baker M, Freund T, di Luca M, Mendlewicz I, Ragan Westphal M. (2006). J Neurol Neurosurg Psychiatr. 77(Suppl 1): 1–49. Patel V, Sumathipala A. (2007). Br J Psychiatr. 190: 77–78. Patel V, Araya R, Chatterjee S, Chrisholm D, Cohen A, De Silva M, Hosman C, McGuire H, Rojas G, van Ommeren M. (2007). Lancet. 370: 991–1005. Prince M, Patel V, Saxena S, Maj M, Maselko J, Phillips M, Rahman A. (2007). Lancet. 370: 859–877. Roth G, Ekblad S. (2006). J Nerv Ment Dis. 194: 378–381. Saraceno B, van Ommeren M, Batniji R, Cohen A, Gureje O, Mahoney J, Sridhar D, Underhill C. (2007). Lancet. 370: 1164–1174. Sartorius N. (2000a). In: Helmchen H, et al. (eds.) Psychiatrie der Gegenwart. Band 3: Psychiatrie spezieller Lebenssituationen. 4. Auflage. Springer, Berlin/Heidelberg/New York, pp. 425–446. Sartorius N. (2000b). In: Okasha A, et al. (eds.) Culture, Ethics and Psychiatry. APA Press, Washington. Sartorius N, Davidian H, Ehrenberg G, Fenton FR, Jujii I, Gastpar M, Guibinat W, Jablensky A, Kielholz P,
Global Burden of Mental Health Lehmann HE, Naraghi M, Shimizu M, Shinfuku N, Takahashi R. (1983). Depressive Disorders in Different Cultures: Report on the WHO-CollaborativeStudy on Standardized Assessment of Depressive Disorders. World Health Organization, Geneva. Saxena S, Thornicroft G, Knapp M, Whiteford H. (2007). Lancet. 370: 878–889. Tseng W-S. (2001). Handbook of Cultural Psychiatry. Academic Press, San Diego. Tseng WS. (2003). Clinician’s guide to cultural psychiatry. Academic Press, San Diego. Weismann MM, Bland RC, Canino GJ, et al. (1996). JAMA. 276: 293–299. WHO. (2001). World Health Report. WHO, Geneva. WHO. (2004). JAMA. 291: 2581–2590. WHO. (2005a). Mental Health: Facing the Challenges, Building Solutions. Report from WHO European
86
Ministerial Conference. WHO Regional Office, Copenhagen. WHO. (2005b). World Mental Health Atlas. WHO, Geneva. WHO. (2008a). http://www.who.int/mental_health/prevention/genderwomen/en/ 2.2.2008. WHO. (2008b). http://www.searo.who.int/en/Section1174/ Section1199/Section1567/Section1826_8101.htm) 8.1. 2008. Wittchen HU, Lieb R, Wunderlich U, Schuster P. (1999). Clin Psychiatr. 60(Suppl 7): 29–36. Wittchen HU, Jacobi F. (2005). Eur Neur psych pharmacol. 15: 357–376. World Bank. (1993). World Development Report 1993: Investing in Health. Oxford University Press, New York.
1491
87 Disease Burden and Disability-Adjusted Life Years Due to Schizophrenia and Psychotic Disorders A. Theodoridou . W. Ro¨ssler 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1494 1.1 Schizophrenia, Schizophrenia Spectrum Disorders and Other Psychotic Disorders with a More Benign, Short-Lived Course . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1495 2
Epidemiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1499
3
The Link Between Physical and Mental Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1499
4
Treatment Possibilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1501
5
Stigma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1501
6
Burden and the Role of Unmet Need . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1501
7
Family Burden . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1502
8
Quality of Life . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1503
9
The Economic Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1505 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1506
#
Springer Science+Business Media LLC 2010 (USA)
1494
87
Disease Burden and DALY Due to Schizophrenia and Psychotic Disorders
Abstract: > Schizophrenia usually starts in young adulthood. The cumulative lifetime risk for men and women is similar, although it is higher for men in the age group younger than 40 years. A common categorization of severe mental illness includes the schizophrenic and delusional disorders listed under e.g. categories F20–29 of the WHO International Classification of Diseases (ICD). Characteristic for the schizophrenic disorder is a distortion of thinking and perception and affects that are inappropriate or blunted. > Hallucinations, delusional perceptions and > delusion of controls are psychopathological phenomena that appear in schizophrenia. Diagnosis and treatment of schizophrenia are often delayed. Schizophrenia is one of the most burdensome and costly illnesses worldwide, because of onset, course and rate of disabilities. Family relationships suffer, if the burden of care is shifted to families. According to the Global Burden of Disease Study, schizophrenia causes a high degree of disability, which accounts for 1.1% of the total > DALYs (disability-adjusted life years) and 2.8% of YLDs (years lived with disability). Schizophrenia is listed as the eighth leading cause of DALYs worldwide in the age group 15–44 years, according to the WHO World Health Report: new understanding, new hope, 2001, Geneva. The risk of suicide is very high, 10–13% of people with schizophrenia commit suicide. Comorbid somatic conditions in people with schizophrenia can also lead to premature death. Optimizing the general health needs attention. List of Abbreviations: BPRS, brief psychiatric rating scale; DALY, disability-adjusted life year; DSM, diagnostic and statistical manual of mental disorders; GAF, global assessment of functioning; GBD, global burden of disease; ICD, international classification of diseases; MSQOL, > modular system for Quality of Life; > PANSS, positive and negative syndrome scale; QOL, quality of life; QUALY, quality-adjusted life year; > SF-36, Short-Form 36; WHO, world health organization; YLD, years lived with disability; YLL, years of lost life
1
Introduction
‘‘Schizophrenia and psychotic disorders’’ are a heterogeneous category. Emil Kraepelin (1896) first described an entity in using the term ‘‘dementia praecox,’’ which was later called schizophrenic. With this expression he described clinical symptoms like hallucinations and delusions (Jaspers, 1913/1963) within a long-term course. A few years later Eugen Bleuler (1911) implemented the term schizophrenia. Symptoms can be e.g. disturbances in thinking, perception and emotions. As recommended by Kane it is possible to split schizophrenic symptoms into two categories, i.e. positive and > negative symptoms. > Positive symptoms comprise hallucinations (e.g. internal voices discussing or commenting), delusions (e.g. of being controlled or persecuted) and thought disorders like incoherence. Negative symptoms reflecting a diminution or loss of normal function include poverty of speech, inappropriate or flattening affect, apathy and anhedonia. These symptoms can appear together or alternate. Crow proposed a new typology for schizophrenia in 1980, dividing the disorder into type I (among others characterized by predominantly positive symptoms, good response to treatment) and type II (characterized mainly by negative symptoms, insidious onset and tendency to drug resistance). Negative symptoms are supposed to be more stable than positive symptoms. However, the measurement of negative symptoms can be confounded by various factors like neuroleptic side-effects, depression or environmental understimulation. The term > psychosis was introduced by Ernst Feuchtersleben and has been in use since 1845, initially to differentiate from neuroses. Commonly, psychosis means to be impaired in reality testing with hallucinations or delusions occurring. In broader definitions, psychosis can also include
Disease Burden and DALY Due to Schizophrenia and Psychotic Disorders
87
disorganized speech or other positive symptoms of schizophrenia. The illness not only affects the patient, but also his or her family and the society. Due to stigma and other factors schizophrenia has been a misunderstood condition to patients, their families and to many in the medical field.
1.1
Schizophrenia, Schizophrenia Spectrum Disorders and Other Psychotic Disorders with a More Benign, Short-Lived Course
Schizophrenia and Psychotic Disorders are classified in the chapter F2 (see > Table 87-1) of the International Classification of Diseases (WHO) and analogous in the Diagnostic and Statistical Manual of Mental Disorders (> DSM IV-TR) (APA). Nevertheless these two classification systems differ regarding definition, duration and subtypes of schizophrenia. > ICD-10 e.g. avoids criteria based on social and occupational dysfunction because of the difficulties to equate these criteria between different cultures. Regarding the longitudinal course and the subtypes both classification systems are broadly similar. Schizophrenia functions as the main item in this group, the remaining categories are related to schizophrenia. Characteristic for the schizophrenic disorder is a distortion of thinking and perception and affects that are inappropriate or blunted. The core psychopathological phenomena of Schizophrenia includes thought echo, thought insertion or withdrawal, thought broadcasting, delusional perception and delusions of control (for further details see Kaplan & Sadock’s Comprehensive Textbook of Psychiatry, 8th edition). Depressive symptoms are often among the earliest signs of schizophrenia onset (Yung and McGorry, 1996). Impairments in attention and information processing, which can often occur in schizophrenia, are associated with social and vocational impairments. In the further course depressive symptoms can appear as a sign of an imminent relapse (Gaebel et al., 2000). They can also appear in a psychotic episode or following a psychotic episode. The course of schizophrenic disorders can be either continuous or episodic with progressive or stable deficits. One or more episodes can occur with complete or incomplete remission. A better course of the illness has been observed in women, in patients who had an acute, stressrelated onset of their illness, low levels of negative symptoms, higher social class, better premorbid social development and no cannabis use (Kelly et al., 2001). Onset usually occurs in late adolescence or early adulthood. Often a prodromal period appears several years before the diagnosis can be made. The initial phase of schizophrenia, with the prodromal stage and the psychotic prephase, varies greatly in type and length (see > Figure 87-1). In the ‘‘ABC Study,’’ a retrospective epidemiological study called the Mannheim Age – Beginning – Course Study, it was shown that the impairments often already occur in the early preclinical course of the disease and the duration of untreated psychosis (DUP) had a mean of 1.1 years in this study (Ha¨fner et al., 1995). The DUP is seen as a predictor of an unfavorable illness course (Ha¨fner et al., 2002). The longer the DUP, the more of persisting symptoms and the longer the inpatient treatment. Although this association is not seen consistently in all groups, early detection and early intervention could be helpful in reducing the pronounced social consequences of the disorder. With regard to the mental health of young people we have to realize that mental disorders in young people have a substantial effect on economic and social outcome, since most young people take first steps regarding their career, job, friendships and relationships during this period of their life. The functional impairment, exposure to stigma and discrimination can persist into adulthood and have an impact on public health (Patel et al., 2007).
1495
1496
87
Disease Burden and DALY Due to Schizophrenia and Psychotic Disorders
. Table 87-1 Block of ICD-10 categories relating to schizophrenia, schizotypal and delusional disorders Schizophrenia, schizotypal and delusional disorders F 20 Schizophrenia F 20.0 Paranoid schizophrenia F 20.1 Hebephrenic schizophenia F 20.2 Catatonic schizophrenia F 20.3 Undifferentiated schizophenia F 20.4 Post-schizophrenic depression F 20.5 Residual schizophrenia F 20.6 Simple schizophrenia F 20.8 Other schizophrenia F 20.9 Unspecified schizophrenia F 21 Schizotypal disorder F 22 Persistent delusional disorders F 22.0 Delusional disorder F 22.8 Other persistent delusional disorders F 22.9 Persistent delusional disorder, unspeciefied F 23 Acute and transient psychotic disorders F 23.0 Acute polymorphic psychotic disorder without symptoms of schizophrenia F 23.1 Acute polymorphic psychotic disorder with symptoms of schizophrenia F 23.2 Acute schizophrenia-like psychotic disorder F 23.3 Other acute predominantly delusional psychotic disorders F 23.8 Other acute transient psychotic disorders F 23.9 Acute and transient psychotic disorders unspecified F 24 Induced delusional disorder F 25 Schizoaffective disorders F 25.0 Schizoaffective disorder, manic type F 25.1 Schizoaffective disorder, depressive type F 25.2 Schizoaffective disorder, mixed type F 25.8 Other schizoaffective disorders F 25.9 Schizoaffective disorder, unspecified F 28 Other nonorganic psychotic disorders F 29 Unspecified nonorganic psychosis Course: continuous/episodic(with progressive deficit, stable deficit, remittent)/incomplete remission/complete remission/other This is an excerpt from ICD-10 Classification of Mental and Behavioral Disorders (1992) (Data source: ICD-10; Clinical Descriptions and Diagnostic Guidelines, pp. 84–85, WHO Geneva. Reprinted with permission from World Health Organization)
. Figure 87-1 The prephases of schizophrenia
Disease Burden and DALY Due to Schizophrenia and Psychotic Disorders
87
1497
1498
87
Disease Burden and DALY Due to Schizophrenia and Psychotic Disorders
The prephases of schizophrenia from the first sign of mental disorder, the first positive symptom to first admission. On average 5 years before the first positive symptoms occur patients reported unspecific symptoms (N = 232) (Data source: Search for the Causes of Schizophrenia, vol III (1995), pp. 43–56, (ed: Ha¨fner H, Gattaz H) Onset and early course of schizophrenia. Ha¨fner H, Maurer K, Lo¨ffler W, Bustamante S, an der Heiden W, RiecherRo¨ssler A, Nowotny B. (p. 53, > Figure 87-5) With kind permission of Springer Science and Business Media.) Diagnosis and treatment of schizophrenia are often delayed. Patients can show symptoms like delusions or hallucinations for years before the disorder is diagnosed and treatment is initiated. This duration of untreated psychosis (DUP, see > Figure 87-2) seemed to be associated with impaired quality of life and poor outcome (Browne et al., 2000) and higher risk of depression and suicide (Addington et al., 1998). Treatment during the prodromal period is hoped to prevent psychosis or at least reduce its severity. In the past years a lot of efforts have been made to understand symptoms of the schizophrenia prodrome. McGorry and his colleagues developed a set of criteria for identifying a prodrome. Other groups in Europe and USA developed different assessments, e.g. the Instrument for the Retrospective Assessment of the Onset of Schizophrenia (IRAOS) (Ha¨fner et al., 1992) or the Structured Interview for Prodromal Symptoms (SIPS) and Scale of Prodromal Symptoms (SOPS) (Miller et al., 1999). Multiple interactions between genes and environment are involved in the etiology of schizophrenia. Triggering factors and environmental influences, e.g. like birth complications, . Figure 87-2 Early course of schizophrenia – phases and definitions. DUP duration of untreated psychosis; DUI duration of untreated illness. Phases and their definitions pointing out the different stages of schizophrenia (Data source: modified from McGlashan and Johannessen (1996), with permission from Oxford University Press.)
Disease Burden and DALY Due to Schizophrenia and Psychotic Disorders
87
drug abuse or urban background have been identified. Further knowledge about possible genotype-environment interactions is required in order to develop and improve strategies for the prevention and early treatment of schizophrenia (van Os and Marcelis, 1998). Beside the role genetic predisposition plays in the etiology of schizophrenia, strong evidence exists for a number of co-factors that influence manifestation and course of the illness (> Figure 87-3) (Ehrenreich et al., 2002).
2
Epidemiology
Schizophrenia has been observed worldwide in all societies. In most psychiatric textbooks the incidence and lifetime prevalence rates are defined as equal worldwide. Lifetime prevalence of schizophrenia can be estimated at 1%. McGrath and colleagues point out in their systematic review from 2004, that the median incidence rate was 15.2 per 100,000, with many studies reporting rates in the upper range (10–90% quantiles = 7.7–43.0). They describe higher incidence rates in men compared to women, higher rates in urban sites compared to mixed urban/rural sites and higher rates in migrants compared to native-born individuals (McGrath et al., 2004). Due to the sometimes chronic course the patients can be ill over a long period of time, although the life expectancy for a person with schizophrenia is 20% shorter than for the general population. The incidence or prevalence of schizoaffective disorder, schizophreniform disorder and brief psychotic disorder is not yet clear today because of the change in diagnosis over time. Due to differing observation times, methodology and true differences, the prevalence rates show more than a 50-fold variation (Wittchen and Jacobi, 2005). In a WHO study the determinants of outcome of severe mental disorders (DOS) were obtained with a cohort size of 1,379 incidence, based on service contacts and clinical information on the diagnosis (Jablensky et al., 1992). The incidence ranged between 0.016 and 0.042% per year with a tendency toward earlier onset in males for a broadly defined diagnosis of schizophrenia. Wittchen et al. identified 27 eligible studies with quite variable designs and methods including over 150,000 subjects from 16 European countries. Only 26% of all cases had had any consultation with professional health care services, a finding suggesting a considerable degree of unmet need. Severity, disability and comorbidity are necessary information for determining the degree of met and unmet needs. These needs are most pronounced for the new EU member states as well as more generally for adolescent and older populations (Wittchen and Jacobi, 2005). Nearly 30% of patients diagnosed with schizophrenia had attempted suicide at least once during their lifetime (Radomsky et al., 1999) and about 10% of the patients commit suicide (Caldwell and Gottesmann, 1990).
3
The Link Between Physical and Mental Problems
There are great variations in life expectancy worldwide. Compared to the general population life expectancy of people with severe mental illness is reduced. In a systematic review of mortality in schizophrenia Saha et al. conclude that people with schizophrenia have a twofold to threefold increased mortality risk. Both sexes are equally affected (Saha et al., 2007). People with severe mental illness suffer an increased risk of having e.g. diabetes and hypertension. They have a ten times a greater risk of cardiovascular disease (STAKES, 2003). According to published data the greatest cause of death among persons with schizophrenia is cardiovascular disease followed by suicides (Brown, 1997). Caldwell and Gottesmann
1499
. Figure 87-3 Etiology and pathogenesis of schizophrenic psychosis. Co-factors that influence manifestation and course of schizophrenia (Data source, modified from: Risk and Protective Factors in Schizophrenia (2002, p. 259), Neuroprotection in Schizophrenia – What does it mean? – What means do we have? Ehrenreich H, Siren A-L, > Figure 87-1. With kind permission of Springer Science and Business Media.)
1500
87 Disease Burden and DALY Due to Schizophrenia and Psychotic Disorders
Disease Burden and DALY Due to Schizophrenia and Psychotic Disorders
87
showed in their review that 10–13% of people with schizophrenia committed suicides. In comparison to the specific-cause standardized mortality ratios, suicide was estimated 12 times greater than expected in the general population (Saha et al., 2007). The risk of suicide increased with e.g., depressive symptoms, alcohol and drug dependency and socioeconomic difficulties.
4
Treatment Possibilities
Today psychosocial interventions and medication are the most common treatment approaches. They include, tailored to each individual, medication, psychotherapy, family counseling and community-based service. Although a number of promising interventions seem to be efficient for different domains, the fragmented care can often represent a barrier to access care. An integrated approach to treatment may be helpful in trying to optimize recovery and outcome for individuals. Another area of intense interest is the attempt to identify premorbid symptoms early and intervene before the illness reaches full manifestation. Evaluation of the use of low-dosage antipsychotic treatment in high-risk individuals is under way.
5
Stigma
Patients, their families and the society can be affected through negative attitudes and stereotypes. They can also internalize false beliefs like being defective and undeserving. It is wellknown that negative attitudes towards the mentally ill still exist, especially towards persons with schizophrenia (Stuart and Arboleda-Florez, 2001). This creates a vicious cycle of discrimination leading to social isolation, unemployment, drug abuse, long-lasting institutionalization or even homelessness, which further decreases the chances of recovery and reintegration into normal life. Several strategies were developed to fight the stigma and discrimination because of schizophrenia, e.g. the global anti-stigma program from the World Psychiatric Association.
6
Burden and the Role of Unmet Need
With the Global Burden of Disease Study (GBD) a new metric called disability-adjusted life year (DALY) was introduced to quantify the burden of diseases. In a standardized approach to epidemiological assessment DALY should aid comparisons. DALY combines information on the impact of mortality (years of life lost because of premature death = YLL) and disability (years lived with disability = YLD). One DALY can be thought as one lost year of ‘‘healthy’’ life. According to the Global Burden of Disease Study (Murray and Lopez, 1996, 1997), schizophrenia causes a high degree of disability that accounts for 1.1% of the total DALYs and 2.8% of YLDs. In the World Health Report (2001) schizophrenia is listed as the eighth leading cause of disability-adjusted life years worldwide in the age group 15–44 (Ro¨ssler et al., 2005). ¨ stu¨n et al., 1999) active psychosis ranked higher In further studies (Harwood et al., 2004; U
1501
1502
87
Disease Burden and DALY Due to Schizophrenia and Psychotic Disorders
than paraplegia and blindness in the general population and ranked in the highest disability class, requiring daily care. Maybe Policy makers and public health experts are interested in tools for estimating the burden of disease in the context of evaluation and priority setting for health disorders. Though DALYs and QALYs (quality-adjusted life years, the complement to DALY, a measure for the impact of a disease in terms of duration and quality of life) are common measures, we have to take the values (not only monetary) into account which play a critical role in decision making (Frohberg and Kane, 1989). Another limitation is the fact that basic parameters for mental health epidemiology like incidence, duration and change in parameters over time and treatment effectiveness depend on the based classification system and can change, and we also have to take the fact into account that up to now these do not exist in many developing countries. It is possible to distinguish between the total and the ‘‘avoidable’’ burden. ‘‘Avoidable’’ in the sense of Hollinghurst et al. (2000) means the element of the total burden that ought to be amenable to prevention and treatment. The authors distinguish between incidence-based DALYs and prevalence-based DALYs. The first are appropriate where prevention should be able to reduce the burden of disease. The prevalence-based DALYs are appropriate when a disease cannot be prevented but effective treatment is available. Several assumptions have to be made, while keeping in mind that there are still considerable gaps in knowledge about the natural history of schizophrenia. Hollinghurst et al. consider that ‘‘avoidable’’ DALYs are relevant in deciding which health care to pay for, and total DALYs offer useful information to indicate how research funds ought to be spent. Both kinds of DALYs can be useful in providing information about monitoring the health of populations.
7
Family Burden
The burden of care has often shifted to families. Family relationships suffer. They sustain a substantial amount of the burden due to the schizophrenic illness. Examples of this kind of burden are emotional reactions to the illness, the stress of coping with disturbed behavior, the disruption of household routine, the stigma they too are confronted with, the restriction of social activities and economic difficulties. The most important predictor of burden on relatives according to Lauber et al. (2003) was the distress and changes in the relationship between the caregiver and the affected individual that occur during acute illness. Recent research has studied burden in relatives of persons with schizophrenia (Jungbauer et al., 2004; Schulze et al., 2005; Thornicroft et al., 2004). In a qualitative study Jungbauer et al. investigate the effect of care giving in spouses of people with schizophrenia, they discuss partnership difficulties and resources and compare burden between spouses and parents. Thornicroft et al. investigated the personal impact of schizophrenia in a multi-centre study. The authors compared caregiver burden across five European countries. They found caregiver burden in schizophrenia to be almost identical across five countries. Scores were higher when patients lived with their family and in regions with fewer psychiatric beds. For the practical and psychological burden experienced by relatives of patients with schizophrenia see > Table 87-2 (Magliano et al., 2002). In another study with a representative sample the authors compared burden and social networks in families of patients with schizophrenia or a longterm physical disease, and found higher subjective burden as well as dramatically lower social support and help in emergencies concerning the patient with schizophrenia (Magliano et al., 2005).
87
Disease Burden and DALY Due to Schizophrenia and Psychotic Disorders
. Table 87-2 Practical and psychological burden experienced by relatives of patients with schizophrenia (N = 709)
Variables
Alwaysoften/yes (%)
Sometimes Never/ (%) no(%)
Practical problems I have had to wake up during the night
22
34
44
I have had to neglect my hobbies and things I like doing in my free time
45
28
27
I have had difficulty in going on Sunday outings
38
29
33
I found it difficult to have fiends at home
31
24
45
I found it difficult to meet my friends and people I like to spend my leisure time
33
29
38
I found it difficult to carry out my usual work or household activities or I had to stay at home from work or school
22
33
45
I had to neglect other family members
21
30
49
I had difficulty in going on holiday
47
21
32
I felt that I would not be able to bear this situation much longer
38
37
25
I cried or felt depressed
45
38
17
I worry for the future of other family members
40
33
27
When I went to a public place with my ill relative, I felt that everyone was watching us
21
31
48
I feel guilty because I believe that I or my spouse may have passed on the illness to our relative
9
18
73
I think that if our relative didn’t have this problem, everything would be all right in our family
62
23
15
When I think of how our ill relative was beforehand and how he/ she is now, I feel disappointed
83
14
3
Psychological problems
Practical and psychological problems experienced by relatives of patients with schizophrenia (Data source: Magliano et al. (2002), with permission from Blackwell Publishing, Oxford)
8
Quality of Life
Quality of life (QOL) not only depends on the disorder but also on the individual, economic and sociocultural conditions that the patients live in. It is a complex multidimensional construct that consists of various domains including physical health, psychological health and general health perceptions of the individual, social functioning and role functioning (Wilson and Cleary, 1995). Subjective QOL can be measured reliably and reflect clinically relevant changes responsively by the MSQOL and the SF-36, although discriminant validity with regard to depression and current mood is questionable (Pukrop et al., 2003).
1503
1504
87
Disease Burden and DALY Due to Schizophrenia and Psychotic Disorders
Individuals with first episode spectrum disorders (schizophrenia, schizoaffective and schizophreniform disorders show a negative correlation between the QOL and negative symptomatology on the PANSS (Ho et al., 1998; Sim et al., 2004). In a study published 1999, Ro¨ssler et al. assessed the Quality of life (QOL) of 96 schizophrenic patients in different treatment settings (community mental health care vs. long-term hospital care) with the > Munich Quality of Life Dimensions List (MLDL, Heinisch et al., 1991) and searched for the factors which influenced their QOL. The patients in community care reported a better QOL than those in long-term hospital care. They showed a greater satisfaction with life. The dimensions with which the patients were most satisfied and more dissatisfied were not significantly different for the two groups. The most dissatisfying domains were marriage/partnership, sex life, financial situation and job situation, and the patients were most satisfied with their medical treatment and leisure time (see > Figure 87-4). Regression analysis showed that, when other factors influencing QOL (like age, gender, marital status, level of education, professional training, current paid employment, duration of illness, number of previous inpatient stays, first contact with psychiatric services when under 25 years of age, total number of supporting individuals) were included, the place of treatment was no longer significant, but rather the social support, the severity of the illness, educational level and certain illness concepts. Salokangas et al. (2006) interviewed a national sample of 2,221 persons with schizophrenia in Finland 3 years after discharge. Subjective life satisfaction was measured. They conclude that being female and having a good psychosocial functioning, confidants, good physical health,
. Figure 87-4 Quality of life in community-care patients (dark line) vs. long-term hospital-care patients. *Significant differences in QOL domains ( 50: least satisfied, 50: most satisfied) (Data source: modified from Ro¨ssler et al. (1999), with permission from Blackwell Publishing, Oxford)
Disease Burden and DALY Due to Schizophrenia and Psychotic Disorders
87
. Figure 87-5 Service cost comparison for schizophrenic patients in five European countries. Average direct cost of mental health care per annum and patient (in English pounds, for the year 1998). Costs were adjusted for center, gender, marital status, ethnicity, language, employment, age, education, GAF and BPRS (Data source: The EPSILON Study, Knapp et al. (2002), with permission from Blackwell Publishing, Oxford)
and living arrangements in the community that offer support are important factors of life satisfaction (> Figure 87-5).
9
The Economic Perspective
Schizophrenia is found to be a worldwide public health problem that causes enormous economic costs. It begins early in life and can cause long-lasting impairments. Direct and indirect costs due to schizophrenia have been examined in different studies. It is estimated that indirect costs (expressed in financial terms) are five times higher than the direct costs of treatment and care. Schizophrenia has financial consequences for the patients, their relatives and the national economy (Ro¨ssler et al., 1998). In Germany in the year 2002 the health expenditure for schizophrenia patients came to an estimated 1.3% (Statistisches Bundesamt, 2004). In the EPSILON Study the authors compared patterns and costs of schizophrenia care in five European countries and found differences in the annual cost of care per patient between different countries (see > Figure 87-5) (Knapp et al., 2002). With regard to the relatively high prevalence and the fact, that schizophrenia often leads to mental and social disability, the illness is one of the most burdensome and costly illnesses worldwide (Ro¨ssler et al., 2005). In a cost-effectiveness study of current and optimal treatments for mental disorders and
1505
1506
87
Disease Burden and DALY Due to Schizophrenia and Psychotic Disorders
the proportion of burden avertable by each, Andrews et al. compared data (cost, burden averted, efficiency of current and optimal treatment) for affective disorders, anxiety disorders, alcohol use disorders and schizophrenia. They conclude that the efficiency of treatment varied more than tenfold across disorders. Treatment of schizophrenia was ten times the average per case in this investigation. Optimal treatment of everyone with a mental disorder would still leave 60% of the burden unaverted (Andrews et al., 2004). In another modeling study of the same group, they conclude that optimal, evidence-based care for people with schizophrenia would not cost more but would increase the health gain by two-thirds (Andrews et al., 2003). Schizophrenia has a high burden. Knowledge of optimal intervention today allows an improvement, but more work has to be done to better understand the mechanisms and to provide prevention strategies.
Summary Points Schizophrenia usually starts in young adulthood. The initial phase shows variations in type and length with an estimated lifetime prevalence of 1%.
Diagnosis and treatment of schizophrenia are often delayed with a long duration of untreated psychosis (DUP).
The etiology of schizophrenia seems to be multifactorial based on interactions between genes and environment.
Schizophrenia is listed as the eighth leading cause of disability-adjusted life years (DALYs). Values in decision making and changes in the based classification system have to be taken into account as limitations regarding the effort to optimize the estimation of the burden of disease. Burden of care often falls to families. Indirect costs (expressed in financial terms) are five times higher than the direct costs of treatment and care in schizophrenia.
References Addington D, Addington J, Patten S. (1998). Br J Psychiatry 172/Suppl. 33: 90–92. American Psychiatric Association. (2000). Diagnostic and Statistical Manual of Mental Disorders, Fourth Ed., Text Revision. American Psychiatric Association, Washington, DC. Andrews G, Sanderson K, Corry J, Issakidis C, Lapsley H. (2003). Br J Psychiatry. 183: 427–435. Andrews G, Issakidis C, Sanderson K, Corry J, Lapsley H. (2004). Br J Psychiatry. 184: 526–533. Bleuler E. (1911). Dementia praecox oder Gruppe der Schizophrenien. Deutike, Leipzig-Wien. Brown S. (1997). Br J Psychiatry. 171: 502–508. Browne S, Clarke M, Gervin M, Waddington JL, Larkin C, O’Callaghan E. (2000). Br J Psychiatry. 176: 173–176. Caldwell CB, Gottesman II. (1990). Schizophr Bull. 16 (4): 571–589.
Ehrenreich H, Sire´n AL. (2002). In: Ha¨fner H (ed.) Risk and Protective Factors in Schizophrenia. Steinkopff Verlag Darmstadt, pp. 257–262. Frohberg DG, Kane RL. (1989). J Clin Epidemiol. 42(4): 345–354. Gaebel W, Ja¨nner M, Frommann N, Pietzcker A, Ko¨pke W, Linden M, Mu¨ller P, Mu¨ller-Spahn F. (2000). Compr Psychiatry. 41: 76–85. Harwood RH, Sayer AA, Hirschfeld M. (2004). Bull World Health Organ. 82(4): 251–258. Ha¨fner H, Maurer K, Lo¨ffler W, an der Heiden W, Ko¨nnecke R, Hambrecht M. (2002). In: Ha¨fner H (ed.) Risk and Protective Factors in Schizophrenia. Steinkopff Verlag Darmstadt, pp. 207–228. Ha¨fner H, Nowotny B, Lo¨ffler W, Heiden an der W, Maurer K. (1995). Eur Arch Psychiatry Clin Neurosci. 246(1): 17–28.
Disease Burden and DALY Due to Schizophrenia and Psychotic Disorders Ha¨fner H, Riecher-Ro¨ssler A, Hambrecht M, Maurer K, Meissner S, Schmidtke A, Faktenheuer B, Lo¨ffler W, an der Heiden W. (1992). Schizophr Res. 6: 209–223. Heinisch M, Ludwig M, Bullinger M. (1991). Bullinger M, Ludwig M, von Steinbu¨chel N (ed.) Lebensqualita¨t bei Kardiovaskula¨ren Erkrankungen. Grundlagen, Messverfahren und Ergebnisse. Go¨ttingen, Hogrefe, pp. 73–90. Ho BC, Nopoulos P, Flaum M, Arndt S, Andreasen NC. (1998). Am J Psychiatry. 155: 1196–1201. Hollinghurst S, Bevan G, Bowie C. (2000). Health Care Manag Sci. 3: 9–21. Jablensky A, Sartorius N, Ernberg G, Anker M, Korten A, Cooper JE, Day R, Bertelsen A. (1992). Schizophrenia: Manifestations, Incidence and Course in Different Cultures. Cambridge University Press. Jaspers K. (1913/1963). General Psychopathology. Manchester University Press (orig.: Allgemeine Psychopathologie, Springer Verlag, Berlin). Jungbauer J, Wittmund B, Dietrich S, Angermeyer MC. (2004). Schizophr Bull. 30(3): 665–675. Kaplan & Sadock’s Comprehensive Textbook of Psychiatry, 8th ed., Lippincott Williams & Wilkins, pp. 1096–1232. Knapp M, Chisholm D, Leese M, Amaddeo F, Tansella M, Schene A, Thornicroft G, Vasquez-Barquero JL, Knudson HC, Becker T. (2002). Acta Psychiatr Scand. 105(1): 42–54. Kelly J, Murray RM, van Os J. (2001). The outcome of psychotic illness. In: Lieberman J, Murray RM (ed.). The Comprehensive Care of Schizophrenia. Dunitz, London, pp. 1531–1539. Kraepelin E. (1896). Psichiatrie, 5th ed. Barth, Leipzig. Lauber C, Eichenberger A, Luginbuhl P, Keller C, Ro¨ssler W. (2003). Eur Psychiatry. 18(6): 285–289. McGlashan TH, Johannessen JO. (1996). Schizophr Bull. 22(2): 201–222. McGrath J, Sukanta S, Welham J, El Saadi O, MacCauley C, Chant D. (2004). BMC Med. 2: 13. Magliano L, Fiorillo A, De Rosa C, Malangone C, Maj M. (2005). Soc Sci Med. 61(2): 313–322. Magliano L, Marasco C, Fiorillo A, Malangone C, Guameri M, Maj M. (2002). Acta Psychiatr Scand. 106: 291–298. Miller TJ, McGlashan TH, Woods SW, Stein K, Driesen N, Corcoran CM, Hoffman R, Davidson L. (1999). Psychiatr Q. 70: 273–287. Murray CJL, Lopez AD. (1996). The Global Burden of Disease: A Comprehensive Assessment of
87
Mortality and Disability from Diseases, Injuries and Risk Factors in 1990 and Projected to 2020. Global Burden of Disease and Injury Series, Vol. 1. Harvard School of Public Mental Health, Cambridge, MA. Murray CJL, Lopez AD. (1997). Lancet 349(9064): 1498–1504. Patel V, Flisher AJ, Hetrick S, McGorry PM. (2007). Lancet. 369: 1302–1313. Pukrop R, Schlaak V, Mo¨ller-Leimku¨hler AM, Albus M, Czernik A, Klosterko¨tter J, Mo¨ller HJ. (2003). Psychiatry Res. 119: 63–79. Radomsky ED, Haas GL, Mann JJ, Sweeny JA. (1999). Am J Psychiatry. 156(10): 1590–1595. Ro¨ssler W, Salize HJ, Cucchiaro G, Reinhard I, Kernig C. (1999). Acta Psychiatr Scand. 100: 142–148. Ro¨ssler W, Salize HJ, Knapp M. (1998). Fortschr Neurol Psychiattr. 66(11): 496–504. Ro¨ssler W, Salize HJ, van Os J, Riecher-Ro¨ssler A. (2005). Eur Neuropsychopharmacol. 15: 399–409. Saha S, Chant D, McGrath J. (2007). Arch Gen Psychitry. 64(10): 1123–1131. Salokangas RK, Honkonen T, Stengard E, Koivisto AM. (2006). Psychiatr Serv. 57(3): 373–381. Schulze B, Ro¨ssler W. (2005). Curr Opin Psychiatry. 18: 684–691. Sim K, Mahendran R, Siris SG, Heckers S, Chong SA. (2004). Psychiatry Res. 129: 141–147. Statistisches Bundesamt. (2004). Krankheitskosten 2002. Wiesbaden. Stuart H, Arboleda-Florez J. (2001). Can J Psychiatry. 46(3): 245–252. The ICD-10 Classification of Mental and Behavioural Disorders. (1992). World Health Organization, Geneva. The WHO World Health Report: New Understanding, New Hope. Geneva, 2001. Thornicroft G, Tansella M, Becker T, Knapp M, Leese M, Schene A. (2004). Schizophr Res. 69(2–3): 125–132. ¨ stu¨n TB, Rehm J, Chatterji S, Saxena S, Trotter R, U Room R, Bickenbach J. (1999). Lancet. 354(9173): 111–115. Van Os J, Marcelis M. (1998). Schizophr Res. 32: 127–135. Wilson IB, Cleary PD. (1995). JAMA. 273: 59–65. Wittchen HU, Jacobi F. (2005). Eur Neuropsychopharmacol. 15: 357–376. Yung AR, McGorry PD. (1996). Schizophr Bull. 22: 353–370.
1507
88 The Societal Costs of Anxiety and Mood Disorders: An Epidemiological Perspective R. C. Kessler . P. S. Wang . H.-U. Wittchen 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1510 2 Methods of Assessing Anxiety and Mood Disorders in Epidemiological Surveys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1511 3 Lifetime and 12-Month Prevalence of Anxiety and Mood Disorders . . . . . . . . . . . . . . 1512 4 Age-of-Onset Distributions of Anxiety and Mood Disorders . . . . . . . . . . . . . . . . . . . . . . . 1515 5 The Course of Anxiety and Mood Disorders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1518 6 Comorbidity Among Anxiety and Mood Disorders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1519 7 The Adverse Effects of Anxiety and Mood Disorders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1520 8 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1522 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1522
#
Springer Science+Business Media LLC 2010 (USA)
1510
88
The Societal Costs of Anxiety and Mood Disorders: An Epidemiological Perspective
Abstract: This chapter reviews the epidemiological literature on the prevalence and adverse societal consequences of anxiety and mood disorders. A special emphasis is put on the recently collected data from the completed World Health Organization (WHO) World ¨ stu¨n, 2008). The first section of the chapter Mental Health (WMH) Surveys (Kessler and U discusses the methods used in community epidemiological surveys to estimate the prevalence and correlates of mental disorders. The next section reviews the recent worldwide epidemiological literature on the general population prevalence of anxiety and mood disorders. This is followed by reviews of recent epidemiological evidence on age-of-onset distributions, illness course, and comorbidity among the anxiety and mood disorders. The last section reviews the recent epidemiological literature on the adverse effects of anxiety and mood disorders. The evidence presented in these sections documents clearly that anxiety and mood disorders are commonly occurring, that many of them start at an early age and have a chronic-recurrent course, and that they have a number of adverse effects that make them among the most costly of all health problems from a societal perspective. List of Abbreviations: AOO, Age-Of-Onset; BPD, Bipolar Disorder; CDS, The National Institute of Mental Health Collaborative Depression Study; CIDI, The Composite International Diagnostic Interview; DIS, The Diagnostic Interview Schedule; DSM, The American Psychiatric Association Diagnostic and Statistical Manual of Mental Disorders; GAD, Generalized Anxiety Disorder; HARP, The Harvard/Brown Anxiety Disorders Research Program; ICD, The World Health Organization International Classification of Diseases; ICPE, The World Health Organization International Consortium of Psychiatric Epidemiology; IQR, Inter-Quartile Range; MDD, Major Depressive Disorder; NCS-R, The National Comorbidity Survey Replication; NIMH, The National Institute of Mental Health; OCD, Obsessive-Compulsive Disorder; PTSD, Post-Traumatic Stress Disorder; SAD, Separation Anxiety Disorder; WHO, World Health Organization; WMH Surveys, The World Mental Health Surveys
1
Introduction
This chapter reviews the epidemiological literature on the societal costs of anxiety and mood disorders. Interest in the costs of illness – not only direct treatment costs, but the human costs as well – has increased dramatically over the past decade as part of the larger movement to rationalize the allocation of treatment resources and maximize benefit in relation to cost. Much of the current interest in anxiety and mood disorders among health policy makers is based on the fact that these disorders have consistently been found in cost-of-illness studies to be among the most costly health problems in the population (e.g., Ormel et al., 2008). A number of factors account for these results and have important implications for the design of treatment programs for anxiety and mood disorders: that anxiety and mood disorders are commonly occurring, often begin at an early age, often are quite persistent throughout the life course, and often have substantial adverse effects on functioning. This chapter reviews the epidemiological evidence regarding these points, with a special emphasis on data from the recently completed World Health Organization (WHO) World Mental Health (WMH) ¨ stu¨n, 2008). Surveys (Kessler and U
The Societal Costs of Anxiety and Mood Disorders: An Epidemiological Perspective
2
88
Methods of Assessing Anxiety and Mood Disorders in Epidemiological Surveys
Information about the epidemiology of anxiety and mood disorders has proliferated over the past two decades. The reason for this can be traced to modifications in the criteria for diagnoses of mental disorders in the DSM system beginning with DSM-III that made it much easier than previously to operationalize diagnostic criteria. Fully structured research diagnostic interviews appropriate for use by trained lay interviewers were subsequently developed for this purpose. The first of these interviews was the Diagnostic Interview Schedule (DIS; Robins et al., 1981), an instrument developed for use in a large community epidemiological survey in the US (Robins and Regier, 1991) and subsequently used in a number of similar surveys in other parts of the world (e.g., Horwath and Weissman, 2000). The WHO subsequently developed the Composite International Diagnostic Interview (CIDI; Robins et al., 1988), which was based on the DIS, in order to have an instrument that could be used to generate diagnoses according to the definitions of both the DSM and ICD systems and that could be used reliably in many different cultures throughout the world (Wittchen, 1994). As general population surveys were carried out in a number of countries with the first version of CIDI, WHO developed a cross-national research consortium to carry out systematic comparisons of CIDI survey results (WHO International Consortium in Psychiatric Epidemiology, 2000). Results based on these comparisons led to the expansion and refinement of the CIDI and to a new generation of cross-national CIDI surveys in the WHO World Mental Health (WMH) Survey Initiative. The latter is an initiative aimed at carrying out and analyzing the results of parallel CIDI surveys in countries throughout the world. Twenty-eight countries have completed WMH surveys as of the time this chapter is being written and close to 200 countryspecific reports have been published from these surveys (www.hcp.med.harvard.edu/wmh). Although only a small number of cross-national comparative WMH reports have been published so far (e.g., Nock et al., 2008; Ormel et al., 2008; Wang et al., 2007), the first volume in a new series of WMH books was recently completed that provides very useful comparative ¨ stu¨n, 2008). We draw heavily on data on disorder prevalence and treatment (Kessler and U these data in this chapter. As the CIDI has become so predominant in psychiatric epidemiological surveys, a few words need to be said about the extent to which diagnoses based on the CIDI are consistent with diagnoses based on independent clinician-administered research diagnostic interviews. Clinical reappraisal studies of the original version of CIDI were quite mixed in this regard, some showing concordance of anxiety and mood disorder diagnoses with clinical diagnoses to be low and others moderate to good. Concordance has been considerably better for more recent versions of CIDI in clinical reappraisal studies carried out in Western countries (Haro et al., 2006). Both individual-level diagnostic concordance and consistency of CIDI anxiety disorder prevalence estimates with prevalence estimates based on clinical interviews have been good in these studies. Much less is known about the clinical relevance of diagnoses based on fully-structured diagnostic interviews in developing countries. Prevalence estimates in epidemiological surveys in some developing countries seem implausibly low, raising concerns either that research diagnostic interviews are not valid in countries where there is no tradition of public opinion research or that the Western diagnostic constructs embedded in existing interviews have low
1511
1512
88
The Societal Costs of Anxiety and Mood Disorders: An Epidemiological Perspective
relevance to some developing countries. Methodological studies are underway to investigate these possibilities. Caution is needed in interpreting results of epidemiological studies carried out in developing countries until methodological investigations have reached a conclusion.
3
Lifetime and 12-Month Prevalence of Anxiety and Mood Disorders
With these cautions as a backdrop, we consider the prevalence estimates of anxiety and mood disorders reported in published community epidemiological surveys. Several recent literature reviews have presented detailed summary tables of prevalence estimates for individual anxiety and mood disorders across a number of epidemiological surveys (Somers et al., 2006; Waraich et al., 2004; Wittchen and Jacobi, 2005). A number of patterns are consistent in these reviews. One is that anxiety disorders have consistently been found to be the most prevalent class of mental disorders in the general population. The estimated lifetime prevalence of any anxiety disorder averages approximately 16% and that of 12-month prevalence approximately 11% in community surveys. The estimated lifetime prevalence of any mood disorder, in comparison, averages approximately 12% and that of 12-month prevalence approximately 6%. There is wide variation around these averages, though, with prevalence estimates generally higher in Western developed countries than in developing countries. This can be seen clearly by examining the range of lifetime and 12-month prevalence estimates of anxiety disorders (> Table 88-1) and mood disorders (> Table 88-2) in the WMH surveys. As shown there, the median lifetime prevalence estimate is somewhat higher for anxiety disorders than in the larger literature – 14.3%, with an inter-quartile range (QR; 25th–75th percentiles) of 9.9–16.7%. The Median WMH lifetime prevalence estimate for any mood disorder, in comparison, is somewhat lower than the average in the larger literature – 10.6% with an IQR of 7.6–17.9%. Twelve-month prevalence estimates (with IQR in parentheses) in the WMH surveys average 8.3% (6.5–12.1) for any anxiety disorder and 5.1% (3.4–6.8) for any mood disorder. Focusing on individual disorders, specific phobia is generally found to be the most prevalent anxiety or mood disorder in community epidemiological surveys, with lifetime prevalence estimates usually in the 6–12% range and 12-month prevalence estimates in the 4–8% range (Silverman and Moreno, 2005). Major depressive disorder (MDD) is generally found to be the next most prevalent lifetime anxiety or mood disorder, with lifetime prevalence estimates usually in the 4–10% range and 12-month prevalence estimates in the 3–6% range (Judd and Akiskal, 2000). Social phobia is generally found to be the next most prevalent anxiety or mood disorder, with prevalence estimates sometimes approaching those of MDD (Furmark, 2002). At the other extreme, Bipolar I disorder is usually found to be the least prevalent anxiety or mood disorder, with lifetime prevalence estimates averaging approximately 1% and 12-month prevalence averaging 0.6% (Benazzi, 2007). Obsessive compulsive disorder (OCD) is usually found to be the least common anxiety disorder, with lifetime prevalence typically less than 2% and 12-month prevalence of approximately 1% (Fontenelle et al., 2006). Prevalence estimates for the other anxiety and mood disorders are higher than these, but lower than those for specific phobia, MDD, and social phobia. The WMH estimates are generally quite consistent with these more general estimates (Kessler et al., 2008). Controversy exists regarding the appropriate diagnostic thresholds for some anxiety and mood disorders, such as post-traumatic stress disorder (PTSD; Mylle and Maes, 2004) and generalized anxiety disorder (GAD; Ruscio et al., 2007). In both these cases, good evidence
The Societal Costs of Anxiety and Mood Disorders: An Epidemiological Perspective
88
. Table 88-1 Lifetime and 12-month prevalence estimates of DSM-IV anxiety disorders in the WMH surveysa Lifetime %
12-month (SE)
%
(SE)
I. WHO Region: Pan American Health Organization (PAHO) Colombia
25.3
(1.4)
14.4
(1.0)
Mexico
14.3
(0.9)
8.4
(0.6)
US
31.0
(1.0)
19.0
(0.7)
II. WHO Region: African Regional Office (AFRO) Nigeria South Africab,c
6.5
(0.9)
4.2
(0.5)
15.8
(0.8)
8.2
(0.6)
12.2
(1.2)
III. WHO Region: Eastern Mediterranean Regional Office (EMRO) Lebanon
16.7
(1.6)
IV. WHO Region: European Regional Office (EURO) Belgium
13.1
(1.9)
8.4
(1.4)
France
22.3
(1.4)
13.7
(1.1)
Germany
14.6
(1.5)
8.3
(1.1)
Israeld,b,c
5.2
(0.3)
3.6
(0.3)
Italy
11.0
(0.9)
6.5
(0.6)
Netherlands
15.9
(1.1)
8.9
(1.0)
9.9
(1.1)
6.6
(0.9)
10.9
(0.8)
6.8
(0.7)
Spain Ukraineb,c
V. WHO Region: Western Pacific Regional Office (WPRO) PRCe
4.8
(0.7)
3.0
(0.5)
Japanc
6.9
(0.6)
4.2
(0.6)
24.6
(0.7)
15.0
(0.5)
New Zealand
c
a
Anxiety disorders include Agoraphobia without a history of panic disorder, adult and childhood separation anxiety disorder, generalized anxiety disorder, panic disorder, post-traumatic stress disorder, social phobia, and specific phobia. Organic exclusions were made as specified in the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition b Specific phobia was not assessed c Adult and childhood separation anxiety disorder not assessed d Social phobia was not assessed e People’s Republic of China
exists from epidemiological surveys that one or more particular diagnostic criteria define a much more restrictive set of cases than the other criteria, calling into question the wisdom of including the restrictive criteria. In the case of GAD, for example, the original diagnostic criteria in DSM-III required a minimum duration of 1 month that was changed in the DSMIII-R and DSM-IV to 6 months in an effort to reduce the high comorbidity found in clinical samples (but not, as it was subsequently discovered, in community samples) between GAD and MDD. The ICD-10 criteria for practice split this difference by requiring a minimum
1513
1514
88
The Societal Costs of Anxiety and Mood Disorders: An Epidemiological Perspective
. Table 88-2 Lifetime and 12-month prevalence estimates of DSM-IV mood disorders in the WMH surveysa Lifetime %
12-month (SE)
%
(SE)
I. WHO Region: Pan American Health Organization (PAHO) Colombia Mexico US
14.6
(0.7)
7.0
(0.5)
9.2
(0.5)
4.7
(0.3)
21.4
(0.6)
9.7
(0.4)
II. WHO Region: African Regional Office (AFRO) Nigeria
3.3
(0.3)
1.1
(0.2)
South Africab
9.8
(0.7)
4.9
(0.4)
(0.9)
6.8
(0.7)
III. WHO Region: Eastern Mediterranean Regional Office (EMRO) Lebanon
12.6
IV. WHO Region: European Regional Office (EURO) Belgiumb,c
14.1
(1.0)
5.4
(0.5)
Franceb,c
21.0
(1.1)
6.5
(0.6)
Germany
b,c
Israel Italyb,c b,c
9.9
(0.6)
3.3
(0.3)
10.7
(0.5)
6.4
(0.4)
9.9
(0.5)
3.4
(0.3)
17.9
(1.0)
5.1
(0.5)
Spainb,c
10.6
(0.5)
4.4
(0.3)
Ukraineb
15.8
(0.8)
9.0
(0.6)
Netherlands
V. WHO Region: Western Pacific Regional Office (WPRO) PRCd
3.6
(0.4)
1.9
(0.3)
Japan
7.6
(0.5)
2.5
(0.4)
20.4
(0.5)
8.0
(0.4)
New Zealand a
Mood disorders include bipolar disorders, dysthymic disorder, and major depressive disorder. Organic exclusions were made as specified in the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition b Bipolar disorders were not assessed c Dysthymia was not included d People’s Republic of China
duration of ‘‘several’’ months. Considerable epidemiological research has now shown, though, that the original 1-month requirement makes more sense from an epidemiological perspective than the more strict duration requirements in that the predictors of lifetime GAD (e.g., childhood adversity, family history of anxiety disorder, primary comorbid disorders) and the clinical correlates of GAD (e.g., age-of-onset, course, role impairment, suicidality) are all very similar for cases with 1-month durations versus 6-month durations. Many of the people with episodes of GAD that do not last as long as 6 months report having many episodes in their lifetimes, suggesting that this is a chronic-recurrent episodic disorder characterized by extreme stress reactivity that triggers episodes of excessive worry that remit (in less than 6 months) and then recur repeatedly over time. The 6-month duration requirements results in these people failing to be defined as clinically meaningful cases. If these diagnostic
The Societal Costs of Anxiety and Mood Disorders: An Epidemiological Perspective
88
controversies were resolved in the direction of broadening the criteria for PTSD and GAD, prevalence estimates would increase substantially: as much as 50% in the case of PTSD and as much as 150% in the case of GAD. A related issue is that considerable evidence exists for the existence of clinically significant sub-threshold manifestations of many anxiety and mood disorders that are much more prevalent than the disorders themselves. Unlike the situations with PTSD and GAD noted in the previous paragraph, the more general issue is that many anxiety and mood disorders appear to be extremes on underlying dimensions rather than categorical manifestations that are qualitatively distinct from these distributions (Brown and Barlow, 2005). For example, even though OCD is almost always estimated to be fairly rare in general population surveys, sub-threshold manifestations of OCD, some of them appearing to be clinically significant, are fairly common (Matsunaga and Seedat, 2007). The same is true for bipolar spectrum disorder, where even though the lifetime prevalence of BP-I is estimated to be only about 0.8–1.5%, the combined prevalence of BP-I, BP-II, and clinically significant sub-threshold BPD is likely in the range 4–6% (Skeppar and Adolfsson, 2006). However, as community epidemiological surveys have for the most part not explored these sub-threshold manifestations systematically, we do not currently have good estimates of the proportion of the population that would meet criteria for one or more anxiety and mood spectrum disorders.
4
Age-of-Onset Distributions of Anxiety and Mood Disorders
While the results reviewed in the last section document that anxiety and mood disorders are highly prevalent, it is also important to examine age-of-onset (AOO) distributions (the distribution of when the disorder first occurred across people with a lifetime history of a disorder) for three reasons. The first reason is that commonly occurring lifetime disorders might have much less effect on the overall lives of the people who experience them if they only occur late in life. All else equal, earlier-onset disorders are more burdensome. Second, and related, AOO information allows us to distinguish between lifetime prevalence (the proportion of the population who had a disorder at some time in their life up to their age at interview) and projected lifetime risk (the estimated proportion of the population who will have the disorder by the end of their life). The estimates reported above were for lifetime prevalence rather than for lifetime risk. Lifetime risk cannot be estimated directly from community surveys because respondents differ in age and, therefore, number of years at risk. Projections of estimated future risk can be made from AOO distributions, though, using standard statistical methods. Third, an understanding of AOO is important for targeting research on prevention of mental disorders, early intervention with prodromal or incipient mental disorders, and primary prevention of secondary disorders. In the absence of AOO information, we would have no way to know the appropriate age range to target these interventions. Although AOO is routinely assessed retrospectively in community surveys, only a few reports have been published over the years that describe AOO distributions based on these data. These studies are reviewed elsewhere (Kessler et al., 2007). Recently, though, comprehensive AOO data were published from the WMH surveys (Kessler et al., 2008). These data are remarkably consistent across countries as well as consistent with the AOO data reported in previous studies in showing that some anxiety disorders, most notably the phobias and separation anxiety disorder (SAD), have very early AOO distributions. In the WMH data, these disorders had a median AOO in the range 7–14 and an IQR between 4 and 20 years.
1515
1516
88
The Societal Costs of Anxiety and Mood Disorders: An Epidemiological Perspective
Inspection of the country-specific AOO curves (> Figure 88-1) shows striking consistency. No significant associations were observed between the position of the curve in the swarm and either prevalence of the disorders in the country or the county’s level of economic development. The other common anxiety disorders (panic disorder, generalized anxiety disorder, and post-traumatic stress disorder), in comparison, have considerably later AOO distributions (> Figure 88-2). In addition, the cross-national variation in both median AOO in the WMH surveys (ages 25–53) and in IQR (15–75 years) is considerably wider than for the phobias or SAD. As with the earlier-onset disorders, no significant associations were observed between the position of the curve in the swarm and either prevalence of the disorders in the country or the country’s level of economic development. The mood disorder AOO distributions in the WMH surveys are quite similar to those for the later-onset anxiety disorders (> Figure 88-3). Mood disorder AOO curves show consistently low prevalence until the early teens followed by a roughly linear increase through late middle age and a declining increase thereafter. The median AOO of mood disorders has a very wide range across countries (ages 25–45) and an even wider IQR (17–65 years) but again without any consistent association between the shape of the curve and either prevalence of the disorders in the country or the county’s level of economic development. A note of caution is needed in interpreting these results, as they are based on retrospective lifetime recall and thus are subject to bias. Indeed, somewhat earlier AOO estimates are generally found in prospective-longitudinal studies than in the analysis of retrospective AOO reports (Wittchen et al., 1999). Nonetheless, these prospective data are generally quite consistent with the AOO distributions seen in the retrospective WMH data.
. Figure 88-1 Standardized age-of-onset distributions of DSM-IV phobias and separation anxiety disorder in the WMH surveys (no significant association was found between the position of the curve and either the prevalence of the disorders in the country or the level of economic development of the country. See > Tables 88-1–> 88-3 for a list of the countries included in the figure)
The Societal Costs of Anxiety and Mood Disorders: An Epidemiological Perspective
88
. Figure 88-2 Standardized age-of-onset distributions of DSM-IV generalized anxiety disorder, panic disorder, and post-traumatic stress disorder in the WMH surveys (no significant association was found between the position of the curve and either the prevalence of the disorders in the country or the level of economic development of the country. See > Tables 88-1–> 88-3 for a list of the countries included in the figure)
We noted above that AOO distributions can be used to generate projections of lifetime risk: the proportion of the population that will experience a given disorder at some time in their life. These estimates will necessarily be higher than estimates of lifetime prevalence, as they include not only all lifetime-to-date cases but also some number of anticipated future onsets. The issue of interest is the size of this projected number of future onsets. Estimates of projected lifetime risk of any DSM-IV disorder in the WMH data are roughly one-third higher (IQR 28–44%) than the estimates of lifetime prevalence-to-date (Kessler et al., 2008). This means that 3–4 people in the populations of these countries are likely to develop a first mental disorder at some time in the future for every ten people who already had a disorder. The highest risk-to-prevalence ratios (57–69%) were found in countries exposed to sectarian violence (Israel, Nigeria, South Africa). Excluding these three, no strong difference in riskto-prevalence ratios were found of less developed countries (28–41%) versus developed countries (17–49%). Not surprisingly, the highest class-specific proportional increase in projected lifetime risk versus prevalence was associated with mood disorders (IQR 61–98%). This is because of the comparatively late onset of mood disorders compared to most other mental disorders. The proportional increases were comparable for GAD and PTSD, but very low for the other anxiety disorder, again reflecting the typically later ages of onset of the former than the latter. It is noteworthy that most anxiety disorders have considerably earlier AOO distributions than mood disorders or substance use disorders. There are three noteworthy exceptions, though. First, the AOO distribution of GAD looks more like that of major depression than
1517
1518
88
The Societal Costs of Anxiety and Mood Disorders: An Epidemiological Perspective
. Figure 88-3 Standardized age-of-onset distributions of DSM-IV mood disorders in the WMH surveys (no significant association was found between the position of the curve and either the prevalence of the disorders in the country or the level of economic development of the country. See > Tables 88-1–> 88-3 for a list of the countries included in the figure)
other anxiety disorders. Second, the AOO distribution of panic disorder, while similar to that of GAD in being later than most other anxiety disorders, has a more complex shape than GAD and appears in some studies to differ by gender. Third, the AOO distribution of PTSD is most variable due to the fact that traumatic events can occur at any time in the life course. Despite these differences, though, the results in > Figure 88-1 show that the retrospectively reported AOO distribution of any anxiety disorder is remarkably consistent across countries. This early onset, coupled with the fact that significant associations exist between early-onset anxiety disorders and the subsequent first onset of other mental and substance use disorders (Bittner et al., 2004; Zimmermann et al., 2003), has led some commentators to suggest that aggressive treatment of child-adolescent anxiety disorders might be effective in preventing the onset of the secondary mental and substance disorders that are associated with the vast majority of people with serious mental illness (Kendall and Kessler, 2002). It is noteworthy in this regard that despite their generally early ages of onset, epidemiological data show that first treatment of anxiety disorders usually does not occur until adulthood, often more than a decade after onset of the disorder (Christiana et al., 2000).
5
The Course of Anxiety and Mood Disorders
Course of illness, like AOO, has been much less well studied in epidemiological surveys of anxiety and mood disorders than has prevalence. Indeed, few direct questions about course of illness were included in most community epidemiological surveys of mental disorders prior
The Societal Costs of Anxiety and Mood Disorders: An Epidemiological Perspective
88
to the WMH surveys. However, the fact that anxiety and mood disorders are seen as often being quite persistent adds to the judgment that they have such adverse effects. Objective assessment of this persistence can be obtained by comparing estimates of recent prevalence (variously reported for the year, 6 months, or 1 month before interview) with estimates of lifetime prevalence. The 12-month to lifetime prevalence ratios for anxiety and mood disorders are typically in the range .4–.6, with the ratio always somewhat higher for anxiety disorders than mood disorders. This pattern can be seen by comparing the 12-month and lifetime prevalence estimates reported above for any anxiety disorder and any mood disorder. The highest disorder-specific ratios in these surveys are usually found for specific phobia and the lowest for PTSD, and MDD (Bijl et al., 1998; Kringlen et al., 2001). Ratios as high as these strongly imply that anxiety and mood disorders are highly persistent throughout the life course. More detailed analyses of these ratios could be carried out by breaking them down separately for sub-samples defined by age at interview or by time since first onset, but we are unaware of any published research that has reported such analyses. Our own preliminary analyses of this sort in the WMH data suggest, though, that although 12-month to lifetime prevalence ratios decline with increasing age, this decline is fairly modest, suggesting that anxiety and mood disorders are often quite persistent over the entire life course. The few longterm longitudinal studies that exist in representative samples of people with anxiety and mood disorders yield results consistent with this and suggest that this persistence is due to a recurrent-intermittent course that often features waxing and waning of episodes of different comorbid anxiety and mood disorders (Angst and Vollrath, 1991; Bruce et al., 2005). The most detailed information currently available on the course of anxiety disorders comes from the Harvard/Brown Anxiety Disorders Research Program (HARP), a prospective, naturalistic multi-center study of patients who were originally treated for anxiety disorders as adults and were subsequently followed for over a decade to chart the course of their illness (Bruce et al., 2005). Comparable data for mood disorders come from the NIMH Collaborative Depression Study (CDS), a study similar to HARP but for people with depression that has been going on for over two decades (Leon et al., 2003). The HARP and CDS respondents are not representative of all people with anxiety or mood disorders, as they were treated in tertiary care settings and consequently over-represent seriously impaired refractive cases. Nonetheless, the long-term results of these studies are quite consistent with those of community surveys in finding that anxiety and mood disorders have a typically chronic-recurrent course.
6
Comorbidity Among Anxiety and Mood Disorders
Comorbidity among anxiety and mood disorders is quite common, with up to half of people with any lifetime anxiety or mood disorder meeting criteria for two or more such disorders (Kessler, 1995). Factor analytic studies of diagnostic comorbidity consistently document separate internalizing and externalizing factors in which anxiety and mood disorders have high factor loadings on the internalizing dimension (Kendler et al., 2003). The internalizing dimension, furthermore, has secondary dimensions that distinguish between fear disorders (panic, phobia) and distress disorders (depression, dysthymia, GAD) (Watson, 2005). The locations of OCD, PTSD, and BPD in this two-dimensional space are less distinct, as the former appears to be more related to the fear dimension (Watson, 2005), the second more related to the distress dimension (Cox et al., 2002), and the third not strongly related to either dimension (Watson, 2005). Social phobia, furthermore, appears to be somewhat more strongly related
1519
1520
88
The Societal Costs of Anxiety and Mood Disorders: An Epidemiological Perspective
to the distress dimension than are the other phobias. Separation anxiety disorder has not been included in factor analytic studies to date. These results have recently been used by Watson (2005) to call into question the codification of anxiety disorders as a distinct class of disorders in the DSM and ICD systems and to suggest that a more useful organizing scheme in the upcoming DSM-V and ICD-11 revisions would be one that distinguished between fear disorders and distress disorders, with the latter including not only GAD and possibly PTSD but also unipolar depression and dysthymia. The argument for a class of fear disorders has the stronger support of the two in neurobiological research based on extensive investigation of fear brain circuitry (Knight et al., 2005). The possibility also exists that future research might lead to OCD being distinguished from either fear disorders or distress disorders as part of a spectrum of impulse-control disorders based both on evidence of differential comorbidity and differences in brain circuitry (Whiteside et al., 2004).
7
The Adverse Effects of Anxiety and Mood Disorders
Early-onset anxiety and mood disorders are significant predictors of the subsequent onset and persistence of other mental and substance use disorders as well as of a wide range of physical disorders (He et al., 2008; Ormel et al., 2007). It is important to note that these predictive associations are part of a larger pattern of associations that has been documented between anxiety-mood disorders and a much wider array of adverse life course outcomes that might be conceptualized as societal costs of these disorders, including reduced educational attainment, early marriage, marital instability, and low occupational and financial status (Kessler et al., 1995, 1997, 1998). It is unclear if these associations are causal; that is, if interventions to treat early-onset anxiety and mood disorders would prevent the subsequent onset of the adverse outcomes with which they are associated. As a result, it is not possible to state unequivocally that these outcomes are consequences of anxiety and mood disorders. It would be very valuable, though, from a public health perspective to have long-term evidence to evaluate this issue from experimental treatment effectiveness studies. A considerable amount of research has been carried out to quantify the magnitude of the short-term societal costs of anxiety and mood disorders in terms of healthcare expenditures, impaired functioning, and reduced longevity, but most of this work has been done in the US (Greenberg and Birnbaum, 2005; Greenberg et al., 1999). The magnitude of the cost estimates in these studies is staggering. For example, Greenberg et al. (1999) estimated that the annual total societal costs of active anxiety disorders in the US over the decade of the 1990s exceeded $42 billion. This estimate excludes the indirect costs of early-onset anxiety disorders through adverse life course outcomes (e.g., the documented effects of child-adolescent anxiety disorders in predicting low educational attainment and consequent long-term effects on lower income) and through increased risk of other disorders (e.g., anxiety disorders predicting the subsequent onset of cardiovascular disorder). Although comparable studies of the societal costs of anxiety and mood disorders have been carried out in few other countries, a recent study of the comparative impairments in role functioning caused by mental disorders and commonly occurring chronic physical disorders in the WMH surveys documented that anxiety and mood disorders have substantial adverse effects on functioning in many countries around the world (Ormel et al., 2008). This analysis made use of the fact that physical disorders were assessed in the WMH surveys with a standard
88
The Societal Costs of Anxiety and Mood Disorders: An Epidemiological Perspective
chronic disorders checklist. Respondents with the ten most commonly reported such disorders were asked to report the extent to which each such disorder interfered with their ability to carry out their daily activities in both productive roles (i.e., job, school, housework) and social roles (i.e., social and personal life). The same questions about disorder-specific role impairments were also asked of respondents with each of the mental disorders assessed in the surveys, the ten most commonly occurring of which were compared to the ten physical disorders. Of the 100 logically possible pair-wise disorder-specific mental-physical comparisons, mean impairment ratings were higher for the mental than physical disorder in 91 comparisons in developed and also for 91 comparisons in developing countries. Nearly all of these
. Table 88-3 Disorder-specific impairment ratings in the WHO World Mental Health surveysa Proportion rated severely impaired Developed
Developing
nb
%
(SE)
nb
%
(SE)
xb
Arthritis
526
23.3
(1.5)
127
22.8
(3.0)
0.1
Asthma
119
8.2c
(1.4)
44
26.9
(5.4)
9.0c
Back/neck
912
34.6c
(1.5)
305
22.7
(1.8)
27.0c
I. Physical disorders
Cancer Chronic pain Diabetes Headaches Heart disease
60
16.6
(3.2)
8
23.9
(10.3)
0.0
296
40.9c
(3.6)
109
24.8
(3.8)
12.9c
49
13.6
(3.4)
39
23.7
(6.1)
1.4
751
42.1c
(1.9)
401
28.1
(2.1)
15.7c
83
26.5
(3.9)
63
27.8
(5.2)
0.3
(0.9)
144
23.8
(2.6)
50.0c
c
High blood pressure
91
5.3
Ulcer
31
15.3
(3.9)
59
18.3
(3.6)
0.1
ADHD
87
37.6
(3.6)
14
24.3
(7.4)
0.8
Bipolar
419
68.3c
(2.6)
87
52.1
(4.9)
7.9c
1028
65.8c
(1.6)
622
52.0
(1.8)
30.4c
GAD
576
56.3
c
(1.9)
127
42.0
(4.2)
7.9c
IED
136
36.3
(2.8)
106
27.8
(3.6)
2.0
II. Mental disorders
Depression
ODD
29
34.2
(6.0)
12
41.3
(10.3)
1.2
317
48.4c
(2.6)
67
38.8
(4.7)
4.3c
PTSD
329
54.8
c
(2.8)
53
41.2
(7.3)
4.2c
Social phobia
593
35.1
(1.4)
164
41.4
(3.6)
2.6
Specific phobia
537
18.6
(1.1)
144
16.2
(1.6)
1.9
Panic disorder
a
See > Tables 88-1 and > 88-2 for the list of countries included in the analyses. Colombia, Lebanon, Mexico, Nigeria, PRC, South Africa, and Ukraine were classified developing. All other countries were classified developed b Number of respondents rated severely impaired c Significant difference between developed and developing at the .05 level, two-sided test
1521
1522
88
The Societal Costs of Anxiety and Mood Disorders: An Epidemiological Perspective
higher mental-than-physical impairment ratings were statistically significant at the .05 level and held in within-person comparisons (i.e., comparing the reported impairments associated with a particular mental-physical disorder pair in the sub-sample of respondents who had both disorders). Comparable results were obtained for severe disability ratings (> Table 88-3). Furthermore, a similar pattern held when treated physical disorders were compared with all (i.e., treated or not) mental disorders to address the concern that the more superficial assessment of physical than mental disorders might have led to the inclusion of sub-threshold cases of physical disorders that might have low disability. It is noteworthy that seven of the ten mental disorders in this analyses were anxiety or mood disorders.
8
Conclusions
The epidemiological data reviewed here document that anxiety and mood disorders are commonly occurring in the general population, often have an early age of onset and a persistent course, and are often associated with significant adverse societal costs. We also reviewed evidence to suggest that the current definitions of anxiety and mood disorders might under-estimate the proportion of the population with a clinically significant anxiety or mood syndrome, in which case the societal costs of these disorders would be even greater than estimated here. Based on these results, we can safely conclude that anxiety and mood disorders are common and consequential problems. As early-onset conditions, anxiety disorders are of special importance, as they typically begin prior to the vast majority of the other problems with which they are subsequently associated. Yet young people with early-onset anxiety disorders seldom receive treatment. This is a situation that has to change if we are to be effective in addressing the enormous public health burden created by early-onset anxiety disorders. To do this will require a level of political will that has heretofore been lacking in even the most progressive countries.
Summary Points DSM-IV anxiety disorders are usually found to be the most common class of mental
disorders in community epidemiological surveys. Lifetime prevalence estimates of any anxiety disorder are typically in the range 10–17%, while 12-month prevalence estimates are typically in the range 6–12%. DSM-IV mood disorders are usually found to be somewhat less common than anxiety disorders in community epidemiological surveys. Lifetime prevalence estimates of any mood disorder are typically in the range 8–18%, while 12-month prevalence estimates are typically in the range 3–7%. Some DSM-IV anxiety disorders, most notably the phobias and separation anxiety disorder, have retrospectively reported median ages-of-onset in late childhood or early adolescence (ages 7–14).The vast majority of the lifetime cases of these disorders begin within a relatively narrow age band spanning only about a decade. DSM-IV mood disorders and other anxiety disorders have considerably later retrospectively reported median ages-of-onset (ages 25–45) as well as a much wider age range when these disorders typically begin. Although DSM-IV anxiety disorders are typically more persistent than mood disorders, both classes of disorders often have a chronic-recurrent course.
The Societal Costs of Anxiety and Mood Disorders: An Epidemiological Perspective
88
Early-onset anxiety and mood disorders are associated with reduced educational attainment, early marriage, marital instability, and low occupational and financial status.
The short-term societal costs of anxiety and mood disorders are substantial in terms of effects on healthcare expenditures and impaired functioning. Comparative analyses show that the adverse effects of anxiety and mood disorders on role functioning are generally greater than the effects of commonly occurring chronic physical disorders.
Acknowledgments Preparation of this chapter was supported, in part, by grants The National Comorbidity Survey Replication (NCS-R) is supported by NIMH (U01-MH60220) with supplemental support from the National Institute on Drug Abuse (NIDA), the Substance Abuse and Mental Health Services Administration (SAMHSA), the Robert Wood Johnson Foundation (RWJF; Grant 044780), and the John W. Alden Trust. Collaborating NCS-R investigators include Ronald C. Kessler (Principal Investigator, Harvard Medical School), Kathleen Merikangas (Co-Principal Investigator, NIMH), James Anthony (Michigan State University), William Eaton (The Johns Hopkins University), Meyer Glantz (NIDA), Doreen Koretz (Harvard University), Jane McLeod (Indiana University), Mark Olfson (New York State Psychiatric Institute, College of Physicians and Surgeons of Columbia University), Harold Pincus (University of Pittsburgh), Greg Simon (Group Health Cooperative), Michael Von Korff (Group Health Cooperative), Philip Wang (Harvard Medical School), Kenneth Wells (UCLA), Elaine Wethington (Cornell University), and Hans-Ulrich Wittchen (Max Planck Institute of Psychiatry; Technical University of Dresden). The views and opinions expressed in this report are those of the authors and should not be construed to represent the views of any of the sponsoring organizations, agencies, or U.S. Government. A complete list of NCS publications and the full text of all NCS-R instruments can be found at http://www.hcp.med.harvard.edu/ ncs. Send correspondence to [email protected]. The NCS-R is carried out in conjunction with the World Health Organization World Mental Health (WMH) Survey Initiative. We thank the staff of the WMH Data Collection and Data Analysis Coordination Centres for assistance with instrumentation, fieldwork, and consultation on data analysis. These activities were supported by the National Institute of Mental Health (R01 MH070884), the John D. and Catherine T. MacArthur Foundation, the Pfizer Foundation, the US Public Health Service (R13-MH066849, R01-MH069864, and R01 DA016558), the Fogarty International Center (FIRCA R03-TW006481), the Pan American Health Organization, Eli Lilly and Company, Ortho-McNeil Pharmaceutical, Inc., GlaxoSmithKline, and Bristol-Myers Squibb. A complete list of WMH publications can be found at http://www.hcp.med.harvard.edu/wmh/. Portions of this chapter appeared previously in Kessler, R.C., Ruscio, A.M., Shear, K., Wittchen, H.-U. (in press). The epidemiology of anxiety disorders. In M.M. Antony and M.B. Stein (Eds.), Handbook of Anxiety and the Anxiety Disorders. New York: Oxford University Press Oxford University Press; Wang, P.S., Kessler, R.C. (2005). Global burden of mood disorders. In D. Stein, D. Kupfer and A. Schatzberg (Eds.), Textbook of Mood Disorders (pp. 55–67). Washington DC: American Psychiatric Publishing, Inc. American Psychiatric Publishing, Inc. 2005; Kessler, R.C. (2007). The global burden of anxiety and mood disorders: Putting the European Study of the Epidemiology of Mental Disorders (ESEMeD) findings into perspective. Journal of Clinical Psychiatry 68 (suppl. 2), 10–19, Physicians Postgraduate Press, Inc. 2007. All used with permission.
1523
1524
88
The Societal Costs of Anxiety and Mood Disorders: An Epidemiological Perspective
References Angst J, Vollrath M. (1991). Acta Psychiatr Scand. 84: 446–452. Benazzi F. (2007). CNS Drugs. 21: 727–740. Bijl RV, van Zessen G, Ravelli A, de Rijk C, Langendoen Y. (1998). Soc Psychiatry Psychiatr Epidemiol. 33: 581–586. Bittner A, Goodwin RD, Wittchen HU, Beesdo K, Hofler M, Lieb R. (2004). J Clin Psychiatry. 65: 618–626; quiz 730. Brown TA, Barlow DH. (2005). J Abnorm Psychol. 114: 551–556. Bruce SE, Yonkers KA, Otto MW, Eisen JL, Weisberg RB, Pagano M, Shea MT, Keller MB. (2005). Am J Psychiatry. 162: 1179–1187. Christiana JM, Gilman SE, Guardino M, Kessler RC, Mickelson K, Morselli PL, Olfson M. (2000). Psychol Med. 30: 693–703. Cox BJ, Clara IP, Enns MW. (2002). Depress Anxiety. 15: 168–171. Fontenelle LF, Mendlowicz MV, Versiani M. (2006). Prog Neuropsychopharmacol Biol Psychiatry. 30: 327–337. Furmark T. (2002). Acta Psychiatr Scand. 105: 84–93. Greenberg PE, Birnbaum HG. (2005). Expert Opin Pharmacother. 6: 369–376. Greenberg PE, Sisitsky T, Kessler RC, Finkelstein SN, Berndt ER, Davidson JR, Ballenger JC, Fyer AJ. (1999). J Clin Psychiatry. 60: 427–435. Haro JM, Arbabzadeh-Bouchez S, Brugha TS, de Girolamo G, Guyer ME, Jin R, Lepine JP, Mazzi F, Reneses B, Vilagut G, Sampson NA, Kessler RC. (2006). Int J Methods Psychiatr Res. 15: 167–180. He Y, Zhang M, Lin EH, Bruffaerts R, Posada-Villa J, Angermeyer MC, Levinson D, de Girolamo G, Uda H, Mneimneh Z, Benjet C, Graaf RD, Scott KM, Gureje O, Seedat S, Haro JM, Bromet EJ, Alonso J, von Korff M, Kessler R. (2008). Psychol Med. 1–12. Horwath E, Weissman MM. (2000). Psychiatr Clin North Am. 23: 493–507. Judd LL, Akiskal HS. (2000). Pharmacopsychiatry. 33: 3–7. Kendall PC, Kessler RC. (2002). J Consult Clin Psychol. 70: 1303–1306. Kendler KS, Prescott CA, Myers J, Neale MC. (2003). Arch Gen Psychiatry. 60: 929–937. Kessler RC. (1995). In: Tsuang MT, Tohen M, Zahner GEP (eds.) Textbook in Psychiatric Epidemiology. Wiley, New York, pp. 179–197. Kessler RC, Aguilar-Gaxiola S, Alonso J, Angermeyer MC, Anthony JC, Berglund P, Chatterji S, de Girolamo G, de Graaf R, Demyttenaere K, Gasquet I, Gluzman S, Gruber MJ, Gureje O, Haro JM, Heeringa S, Karam A, Kawakami N, Lee S, Levinson D, Medina-Mora ME,
Oakley-Browne MA, Pennell B, Petukhova M, Posada-Villa J, Ruscio R, Stein DJ, Tsang CHA, ¨ stu¨n TB (2008). In: Kessler RC, U ¨ stu¨n TB (eds.) U The WHO World Mental Health Surveys: Global Perspectives on the Epidemiology of Mental Disorders. Cambridge University Press, New York, NY, pp. 511–521. Kessler RC, Amminger GP, Aguilar-Gaxiola S, Alonso J, ¨ stu¨n TB. (2007). Curr Opin Psychiatry. 20: Lee S, U 359–364. Kessler RC, Berglund PA, Foster CL, Saunders WB, Stang PE, Walters EE. (1997). Am J Psychiatry. 154: 1405–1411. Kessler RC, Foster CL, Saunders WB, Stang PE. (1995). Am J Psychiatry. 152: 1026–1032. ¨ stu¨n TB (eds.) (2008). The WHO World Kessler RC, U Mental Health Surveys: Global Perspectives on the Epidemiology of Mental Disorders. Cambridge University Press, New York, NY. Kessler RC, Walters EE, Forthofer MS. (1998). Am J Psychiatry. 155: 1092–1096. Knight DC, Nguyen HT, Bandettini PA. (2005). Neuroimage. 26: 1193–1200. Kringlen E, Torgersen S, Cramer V. (2001). Am J Psychiatry. 158: 1091–1098. Leon AC, Solomon DA, Mueller TI, Endicott J, Rice JP, Maser JD, Coryell W, Keller MB. (2003). Am J Psychiatry. 160: 727–733. Matsunaga H, Seedat S. (2007). CNS Spectr. 12: 392–400. Mylle J, Maes M. (2004). J Affect Disord. 78: 37–48. Nock MK, Borges G, Bromet EJ, Alonso J, Angermeyer M, Beautrais A, Bruffaerts R, Chiu WT, de Girolamo G, Gluzman S, de Graaf R, Gureje O, Haro JM, Huang Y, Karam E, Kessler RC, Lepine JP, Levinson D, Medina-Mora ME, Ono Y, Posada-Villa J, Williams D. (2008). Br J Psychiatry. 192: 98–105. Ormel J, Petukhova M, Chatterji S, Aguilar-Gaxiola S, Alonso J, Angermeyer MC, Bromet EJ, Burger H, Demyttenaere K, de Girolamo G, Haro JM, Hwang I, Karam E, Kawakami N, Lepine JP, Medina-Mora ME, Posada-Villa J, Sampson N, ¨ stun TB, Von Korff M, Williams D, Scott K, U Zhang M, Kessler RC (2008) Br J Psychiatry. 192: 368–375. Ormel J, Von Korff M, Burger H, Scott K, Demyttenaere K, Huang YQ, Posada-Villa J, Lepine JP, Angermeyer MC, Levinson D, de Girolamo G, Kawakami N, Karam E, Medina-Mora ME, Gureje O, Williams D, Haro JM, Bromet EJ, Alonso J, Kessler RC. (2007). Gen Hosp Psychiatry. 29: 325–334. Robins LN, Helzer JE, Croughan JL, Ratcliff KS. (1981). Arch Gen Psychiatry. 38: 381–389.
The Societal Costs of Anxiety and Mood Disorders: An Epidemiological Perspective Robins LN, Regier A. (1991). Psychiatric Disorders in America: The Epidemiologic Catchment Area Study. The Free Press, New York. Robins LN, Wing J, Wittchen HU, Helzer JE, Babor TF, Burke J, Farmer A, Jablenski A, Pickens R, Regier DA. (1988). Arch Gen Psychiatry. 45: 1069–1077. Ruscio AM, Chiu WT, Roy-Byrne P, Stang PE, Stein DJ, Wittchen HU, Kessler RC. (2007). J Anxiety Disord. 21: 662–676. Silverman WK, Moreno J. (2005). Child Adolesc Psychiatr Clin N Am. 14: 819–843; ix–x. Skeppar P, Adolfsson R. (2006). Nord J Psychiatry. 60: 7–26. Somers JM, Goldner EM, Waraich P, Hsu L. (2006). Can J Psychiatry. 51: 100–113. Wang PS, Aguilar-Gaxiola S, Alonso J, Angermeyer MC, Borges G, Bromet EJ, Bruffaerts R, de Girolamo G, de Graaf R, Gureje O, Haro JM, Karam EG, Kessler RC, Kovess V, Lane MC, Lee S, Levinson D, Ono Y, Petukhova M, Posada-Villa J, Seedat S, Wells JE. (2007). Lancet. 370: 841–850.
88
Waraich P, Goldner EM, Somers JM, Hsu L. (2004). Can J Psychiatry. 49: 124–138. Watson D. (2005). J Abnorm Psychol. 114: 522–536. Whiteside SP, Port JD, Abramowitz JS. (2004). Psychiatry Res. 132: 69–79. WHO International Consortium in Psychiatric Epidemiology. (2000). Bull World Health Organ. 78: 413–426. Wittchen HU, Lieb R, Schuster P, Oldehinkel T. (1999). In: Rapoport JL (ed.) Childhood Onset of ‘‘Adult’’ Psychopathology: Clinical and Research Advances. American Psychiatric Press, Washington, DC, pp. 259–302. Wittchen HU. (1994). J Psychiatr Res. 28: 57–84. Wittchen HU, Jacobi F. (2005). Eur Neuropsychopharmacol. 15: 357–376. Zimmermann P, Wittchen HU, Hofler M, Pfister H, Kessler RC, Lieb R. (2003). Psychol Med. 33: 1211–1222.
1525
89 Estimating the Disease Burden of Combat-Related Posttraumatic Stress Disorder in United States Veterans and Military Service Members M. C. Freed . R. K. Goldberg . K. L. Gore . C. C. Engel 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1528 2 What is PTSD? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1529 3 Prevalence in Military and VA Populations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1531 4 Course of Illness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1535 5 Impact on Patients and the Healthcare System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1536 6 Measuring Disease Burden . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1537 7 Efficacy of Primary Prevention and Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1543 8 Other Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1544 9 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1545 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1545
#
Springer Science+Business Media LLC 2010 (USA)
1528
89
Estimating the Disease Burden of Combat-Related Posttraumatic Stress Disorder
Abstract: > Posttraumatic stress disorder (PTSD) is chronic, disabling, but treatable and potentially preventable anxiety disorder characterized by re-experiencing, avoidance, and hyperarousal symptoms following a traumatic event. The World Health Organization (WHO) characterized disease burden in the general population and listed PTSD among the top 20 most ‘‘burdensome’’ diseases. The WHO Global Burden of Disease study estimated a > disability weights for PTSD of 0.11 (0 = perfect health; 1 = dead), but other studies have estimated weights up to 0.66, a figure closer to estimates for severe major depression. Limited research exists on the disease burden of PTSD, and to date, no literature estimated the disease burden of combatrelated PTSD in the US military and Veterans Affairs (VA) healthcare systems. Extant PTSD literature is arguably sufficient in scope and content needed to calculate reasonable estimates of disease burden of PTSD in these populations. We review five key components to estimating the disease burden of PTSD in the US military and VA: disorder > prevalence; course of illness; disability weights; impact on the healthcare system; and prevention and treatment efficacy. Current PTSD prevalence estimates vary, but range from 4.2 to 24.5% in > servicemembers returning from Operation Iraqi and Enduring Freedom, depending on the subpopulation and assessment method. PTSD often develops months to years after traumatic exposure, and 40% of Vietnam veterans with PTSD reported symptom chronicity over 20 years. PTSD is associated with low quality of life, high rates of medical service utilization, interpersonal conflict, co-morbidity, and work impairment compared to persons without PTSD. PTSD is a chronic, disabling, but treatable and perhaps preventable condition. Disease burden estimates of PTSD could be used to better determine the need for and utility of therapeutic interventions designed to alleviate symptoms of PTSD and associated impairment. List of Abbreviations: CS, cross sectional; DALY, disability adjusted life year; DI, diagnostic interview; DoD, department of defense; DSM, diagnostic and statistics manual; GBD, global burden of disease; IOM, Institute of Medicine; L, longitudinal; MOS SF, 36 medical outcome survey short form-36; MRR, medical record review; NRC, National Research Council; OEF, operation enduring freedom; OIF, operation Iraqi freedom; OR, odds ratio; P, Probability; PC-PTSD, primary care- posttraumatic stress disorder (screen); PCL, posttraumatic stress disorder check list; PDHA, post-deployment health assessment; PDHRA, post-deployment health re-assessment; PTO, > person tradeoff; PTSD, posttraumatic stress disorder; QALY, > quality adjusted life year; SF-6D, short form-6 dimensional health state classification; SRS, self-report screen; SUD, substance used disorder; US, United States; VA, veterans affairs; VHA, veterans health administration; WHO, World Health Organization; YLD, year of life with disability; YLL, years of life lost
1
Introduction
In the United States (US), the World Health Organization (WHO) estimated that posttraumatic stress disorder (PTSD) ranked within the top 20 diseases in terms of population disease burden in the general population (Michaud et al., 2006). When ranked in women with an outcome that combines morbidity (i.e., health-related quality of life) and mortality (death), PTSD ranked 17th under diseases like ischemic heart disease (ranked first) and breast cancer (ranked sixth) and above HIV (ranked 18th). When ranked with an outcome that measures only, PTSD falls in the top 20 of most burdensome diseases, when race and sex are collapsed.
Estimating the Disease Burden of Combat-Related Posttraumatic Stress Disorder
89
There has been an increase in the > incidence of PTSD in the US military, since the beginning of the wars in Iraq and Afghanistan (Army Medical Surveillance Activity, 2004, accessed 2 February 2008). Currently, disease burden estimates of PTSD are broad based (Murray and Lopez, 1996), without consideration of combat or other specific traumatic stressors. PTSD and depression are believed to be the two most burdensome psychological disorders (McFarlane, 2004). Extant PTSD literature is sufficient in scope and content to provide the information needed to calculate reasonable estimates of disease burden of PTSD. We review five key components to estimating the disease burden of PTSD in the US military and Veterans Affairs (VA): disorder prevalence; course of illness; disability weights; impact on the healthcare system; and prevention and treatment efficacy. Future researchers can then use this information to estimate the disease burden of combat-related PTSD. In this chapter, we present current prevalence estimates derived from large samples in VA and US military Department of Defense (DoD) settings, with an emphasis on combat-related PTSD from the wars in Iraq and Afghanistan. We focus on current findings because (1) there have been changes in diagnostic criteria from DSM-III (APA, 1980) to DSM-IV-TR (APA, 2000); (2) DoD and VA have recently given increased attention to PTSD; and c) data from the conflicts in Iraq and Afghanistan and related health policy concerns are relevant to both DoD and VA, while > veterans from prior wars are typically only seen in VA medical centers. However, we recognize that other traumas are important issues within both VA and DoD (e.g., 15.9% women and 0.8% men screen positive for military sexual trauma within the Veterans Health Administration; Kimerling et al., 2007) and combat trauma from other eras still afflict many veterans (Magruder et al., 2005) to include 15% current prevalence of PTSD in Vietnam era veterans (Kulka et al., 1990). Although the focus of this chapter is to identify necessary components for disease burden estimates for PTSD resulting from Conflicts in Iraq and Afghanistan, when published data from these datas are lacking, we offer data from non combat PTSD or other era combat PTSD to serve as a proxy.
2
What is PTSD?
PTSD is an anxiety disorder with symptoms divided into three categories: re-experiencing (Criterion B), avoidance (Criterion C), and hyperarousal (Criterion D; DSM-IV-TR; APA, 2000). Symptoms occur following a traumatic event in which ‘‘an individual experienced, witnessed, or was confronted with an event or events that involved actual or threatened death or serious injury, or a threat to the physical integrity of self or others’’ (Criterion A1). And, the individual’s emotional response to the traumatic event involved ‘‘intense fear, helplessness, or horror.’’ (Criterion A2). In order to be diagnosed with PTSD, symptoms must be present for at least 1 month (Criterion E) and be clinically significant (Criterion F). See > Table 89‐1 for a summary of current DSM-IV TR diagnostic criteria for PTSD. PTSD was first introduced as a formal psychiatric disorder in the Diagnostic and Statistical Manual of Mental Disorders, Third Edition (DSM-III; APA, 1980). The Vietnam War was instrumental in the development of PTSD as a formal diagnosis. Veterans of the war were returning with debilitating trauma reactions, and some prominent psychiatrists, who opposed the war, felt that formalizing a diagnoses of PTSD would provided further justification for ending the war (Jones and Wessely, 2007). Thus, PTSD was forged, in part, out of political policy. However, post war difficulties have long been recognized in soldiers following combat, dating back thousands of years to Homer’s, The Illiad (reported in Jones and Wessely, 2007).
1529
1530
89
Estimating the Disease Burden of Combat-Related Posttraumatic Stress Disorder
. Table 89‐1 Summary of DSM-IV-TR (APA, 2000) criteria for PTSD in adults (Blake et al., 1995) A. Defining traumatic exposure (need both) 1. Traumatic event: The event must involve actual or threatened death, serious injury, or a threat to the physical integrity of oneself or another person. A person must personally experience, witness, or learn about the event. 2. The emotional reaction: Emotional response to the event is intense fear, helplessness, or horror. B. Re-experiencing of traumatic event (need one or more) 1. Intrusive thoughts, images, or perceptions. These are unwanted memories, which are difficult to dismiss. 2. Recurring nightmares. That have thematic similarity to the trauma and cause distress (e.g., cause waking, with difficulty returning back to sleep). 3. Acting or feeling as if the experience was occurring again. May include other sensory experiences (e.g., sights, sounds, or smells), and may include some dissociation and a loss of awareness of one’s surroundings. 4. Intense psychological distress (e.g., crying) following reminders of the traumatic event. The psychological distress causes some impairment in functioning. 5. Intense physiological reactivity (e.g., increased heart rate, sweating, or shaking) following reminders of the traumatic event. The physiological reactivity causes some impairment in functioning. C. Avoidant behaviors (need three or more symptoms that were not present prior to the trauma) 1. Avoiding thoughts, feelings, or talking to others about the trauma. Doing so requires significant effort or interferes with one’s life. 2. Avoiding activities, places, or people that serve as reminders about the trauma. Doing so requires significant effort or interferes with one’s life. 3. Difficulty remembering important parts of the traumatic event. The difficulty is not due to the passage of time, a traumatic brain injury, or pharmacological substances. 4. Significant loss of interest, pleasure, or desire to participate in important or pleasurable activities. 5. Feeling distant, cut off, detached, or disconnected from others. These feelings cause impairments in social functioning. 6. Feeling emotionally numb (e.g., unable to feel love or happiness, and similarly, difficulty feeling sadness or grief). 7. Sense that one’s future will be shortened in some way. This symptom differs from suicidal ideation, in that the foreshortened future will not be due to self injury or harm. Rather, one believes he/she may die prematurely from another cause, like an illness. D. Increased arousal (Need two or more symptoms that were not present prior to the trauma) 1. Difficulty with sleep onset or maintenance. 2. Irritability or angry outbursts. 3. Concentration difficulties. 4. Increased watchfulness or alertness for oneself, family, or home (i.e., hypervigilance) even when there is no need to be. 5. Increased startle response, where one is jumpy. E. Symptoms must persist for more a than 1 month
Estimating the Disease Burden of Combat-Related Posttraumatic Stress Disorder
. Table 89‐1 (continued)
89
F. Symptoms must cause significant distress, disturbance, or impairment to domains of life to other life areas (i.e., social or occupational) Specifiers Acute: symptoms less than 3 months Delayed Onset: symptoms appear after 6 months Chronic: symptoms duration 3 months or more This table describes the diagnostic criteria for PTSD. Language is paraphrased from the DSM-IV-TR (APA, 2000), and the Clinician Administered PTSD Scale (CAPS; Blake et al., 1995), considered the ‘‘gold standard’’ of PTSD assessment. PTSD posttraumatic stress disorder
Exposure to a war zone may well satisfy the DSM-IV-TR Criterion A1, but Criterion A2, the emotional response to the exposure, is necessary for a PTSD diagnosis. Breslau and Kessler (2001) reported the conditional probability of endorsing A2 after experiencing military combat is 33.8%, which is relatively low compared to other A1 events. For example, rape or a potentially lifethreatening car accident results in a conditional probability of 93.3% and 73.2% respectively. As discussed in Gore et al. (2006) and Adler et al. (2008), soldiers may experience numbness or intense anger in response to A1 events, emotions not covered by A2, although they may endorse Criteria B through D symptoms. Thus, if future versions of DSM expand the A2 list of emotional reactions, the rates of PTSD diagnosed following exposure to a war zone are likely to be even higher. The inclusion of the precipitating event (i.e., etiology) in the diagnostic criteria is unique to PTSD and adjustment disorders, whereas other mental disorders require only symptom presence to meet diagnostic criteria. The Criterion A definition of a traumatic event has changed considerably with each DSM revision. In DSM-III, a traumatic event was defined as, ‘‘existence of a recognizable stressor that would evoke significant symptoms of distress in almost anyone’’ (APA, 1980). In DSM-III-R, Criterion A1 defined the trauma as, ‘‘an event that is outside the range of usual human experience and that would be markedly distressing to almost anyone’’ (APA, 1987). Here, a distinction was drawn between the event and the emotional reaction to the event. Defining criterion A and developing a universal standard has been difficult. One of the main problems with Criterion A is defining what a usual or unusual human experience is (Davidson and Foa, 1991). For example, how unusual is combat for a soldier in a volunteer Army? In its current form, Criterion A is comprised of two parts, one subjective and one objective. This alleviates the challenge of determining what constitutes an ‘‘unusual human experience’’ and places the importance on the person’s emotional response to that experience (Weathers and Keane, 2007).
3
Prevalence in Military and VA Populations
Rates of PTSD can vary considerably, depending on the sample surveyed and the data collection methods. For example, PTSD can be determined through self-report, diagnostic interview, or by record review. Moreover, patients seen in VA and DoD clinics can include spouses, and family members of those who served in the armed forces, and veterans and active duty servicemembers may or may not have been deployed to a combat zone. See > Table 89‐2 for a summary of these findings. In a study designed to test the operating characteristics of a single item PTSD screening measure, > Gore et al. (2008) screened 3,234 servicemembers, retirees, and family members for PTSD in three National Capital Area military primary care clinics. They then interviewed a
1531
1532
89
Estimating the Disease Burden of Combat-Related Posttraumatic Stress Disorder
. Table 89‐2 Prevalence of PTSD in VA and DoD
Source
Sample
Method CS; DI
Data collection period
Prevalence
2005–2006
9%
Gore et al. 2008
3,234 active duty, retired, and beneficiary (e.g., family members) screened, 213 interviewed, all in a primay care setting
Grieger et al. (2006)
613 (not including L; SRS attrition) soldiers hospitalized due to battle injuries
Hoge et al. (2004)
2,350 soldiers pre-Iraq CS; SRS; deployment and 3,671 liberal soldiers and marines post- criterion Iraq or Afghanistan deployment
2003
Post-deployment (OR compared to pre-deployed); Army Afghanistan: 11.5% (1.25); Army Iraq: 18.0% (2.13); Marine Iraq: 19.9 (2.4)
Hoge et al. (2004)
2,350 soldiers pre-Iraq CS; SRS; deployment and 3,671 conservative soldiers and marines post- criterion Iraq or Afghanistan deployment
2003
Post-deployment (OR compared to pre-deployed); Army Afghanistan: 6.2% (1.26); Army Iraq: 12.9% (2.84); Marine Iraq: 12.2% (2.66)
Hoge et al. (2007)
2,863 soldiers 1-year post- CS; SRS Iraq
Hoge et al. (2008)
2,714 soldiers who returned from Iraq
Hoge et al. (2006)
222,620 returnees from CS; SRS OIF, 16,318 from OEF, and 64,967 from other locations
Magruder et al. (2005)
746 veterans receiving primary care services
CS; DI
1999
11.5% Total; 4.4% No war zone exposure; 19% War zone exposure
Magruder et al. (2005)
746 veterans receiving primary care services
CS; MRR
1999
8.4%
CS; SRS
2003–2004 1-month: 4.2%; 4-months: 12.2%; and 7-months: 12.0%
Not reported 2006
16.6% 14%
2003–3004 OIF: 9.8%; OEF: 4.7%; Other: 2.1%
Milliken et al. 56,356 active duty army (2007) and 31,885 army national guard and reserve
L; SRS; Liberal 2005–2006 Active: PDHA 11.8, PDHRA 16.7; Reserve: PDHA 12.7, Criterion PDHRA 24.5
Milliken et al. 56,356 active duty army (2007) and 31,885 army national guard and reserve
L; SRS; Conservative criterion
Seal et al. (2007)
103,788 OEF/OIF veterans CS; MRR who separated from the military
2005–2006 Active: PDHA 6.2, PDHRA 9.1; Reserve: PDHA 6.6, PDHRA 14.3 2001–2005
13%
Estimating the Disease Burden of Combat-Related Posttraumatic Stress Disorder
89
. Table 89‐2 (continued)
Source Smith et al. (2008)
Sample 50,128 servicemembers
Method L; SRS
Data collection period
Prevalence
2001–2003 New Onset within 2.7 (average) year period; Deployed w/combat exposure 7.6–8.7%; Deployed w/o combat exposure 1.4–2.1%; non deployed 2.3–3.0%; variability reflects screening method
This table presents current prevalence estimates, dates of data collection, and broad assessment method, of combat PTSD in DoD and VA settings, focusing on the conflicts in Iraq and Afghanistan, when possible. CS cross sectional; L longitudinal; SRS self-report screen; DI diagnostic interview; MRR medical record review; OR odds ratio; VA veterans affairs; DoD department of defense; PTSD posttraumatic stress disorder; OEF/OIF operation enduring freedom/operation Iraqi freedom
selected sample of 213 patients with a structured diagnostic interview. Based on the demographics of the larger sample, the estimated prevalence of PTSD was 9%. Preliminary unpublished findings from this dataset suggest that the prevalence of PTSD in servicemembers is approximately 10%, and 39% of servicemembers with PTSD report combat as the index trauma. At the time of this chapter, prevalence stratified by military status was not available. Grieger et al. (2006) assessed PTSD in 613 soldiers hospitalized after medical evacuation from the theater of operations for serious combat injuries at 1- (n = 613), 4- (n = 395), and 7-(n = 301) months following hospitalizations they found that the PTSD prevalence was 4.2%, 12.2%, and 12.0%, respectively. They used the PTSD Checklist (PCL; Blanchard et al., 1996) a 17-item self-report scale. To meet case criteria, respondents needed to endorse one reexperiencing, three avoidance, and two hyperarrousal symptoms of at least moderate distress (a three or more on a 1–5 likert scale), and the sum score of the measure needed to be at least 50. In a landmark study, Hoge et al. (2004) used a serial cross-section research design and anonymously screened Army combat soldiers prior to an Iraq deployment (n = 2,530), Army combat soldiers three-months after Iraq (n = 894) or Afghanistan (n = 1,962) deployment, and combat Marines (n = 815) three-months after Iraq deployment. The researchers used two scoring criteria for PCL (Blanchard et al., 1996). In the pre-deployment group, PTSD prevalence was 5.0 and 9.4%, depending on whether a strict (PCL score 50) or liberal (PCL score 30) criteria was used. In the post-deployment groups, prevalence ranged from 6.2% in soldiers who returned from Afghanistan when a conservative method was used to 19.9%, in marines who returned from Iraq, when a liberal method was employed. The predeployment results were compared with each of the post-deployment groups. The odds of PTSD were 1.26 (95% CI = 0.97–1.64) times greater in the post- than the pre-deployed group, when the prevalence was 6.2%, and the odds of PTSD were 2.40 (95% CI = 1.92–2.99) times greater in the post- than the pre-deployed group when the prevalence was 19.9%. In a similar study, Hoge et al. (2007) evaluated 2,863 soldiers 1-year following their return from Iraq. They reported a prevalence of 16.6% using the same PCL scoring methodology as Grieger et al. (2006; discussed above).
1533
1534
89
Estimating the Disease Burden of Combat-Related Posttraumatic Stress Disorder
Hoge et al. (2008) assessed 2,714 active duty and national guard/reserve soldiers 4 months following their return from Iraq for a history of head injury and a history of mental health problems. They administered the PCL and determined PTSD caseness if a soldier both met DSM symptom criteria of PTSD (one intrusion symptom, three avoidance symptoms, and two hyperarousal symptoms) and reported distress to total at least 50 (out of 85; the conservative method in Hoge et al., 2004). Fourteen percent of their sample met criteria for PTSD. It is unclear the extent that these three studies (Hoge et al., 2004; Hoge et al., 2007; Hoge et al., 2008) rely on independent versus overlapping study samples. Magruder et al. (2005) estimated the prevalence of PTSD to be 11.5% in a random sample of 746 veterans from four hospital-based VA primary care clinics, who were administered a gold standard research diagnostic interview. War zone exposed veterans had a significantly higher prevalence of PTSD (19%) than non-war zone exposed veterans (4.4%). Of note, only 46.5% of veterans with a research diagnosis of PTSD had PTSD documented in their medical record, and 3.4% of veterans who did not receive a research diagnosis of PTSD nonetheless had a medical record diagnosis of PTSD. Milliken et al. (2007) analyzed data from the US military’s population-based postdeployment screening programs to assess the longitudinal mental health of Army soldiers returning from Iraq. The Post-Deployment Health Assessment (PDHA) is administered to all uniformed military personnel upon their return from deployment, and the Post-Deployment Health Re-Assessment (PDHRA) is administered approximately 3 to 6 months later. The PDHRA and PDHA employ a 4-item primary care PTSD screen (PC-PTSD; Prins et al., 2003). Milliken et al. (2007) used these program data to examine the longitudinal course of PTSD screening results in 56,350 Active Component and 31,885 Guard/Reserve Component soldiers. The porportion of soldiers screening positive for PTSD (2 or more on the 4-question screen) increased from 11.8% at baseline to 16.7%, a median of 6-months later in active component soldiers. The increase among national guard and reserve soldiers was even greater, rising from 12.7% to 24.5%. Using PDHA screening results as a baseline, Hoge et al. (2006) prospectively examined data from 222,620 OIF returnees, 13,318 OEF returnees, and 64,967 returnees from other deployment locations for one year. The baseline proportion of returning troops screening positive for PTSD was 9.8%, 4.7%, and 2.1%, repectively. Of note, automated military health sysem utilization data found that 22% of troops accessed mental health service use during the follow-up year, and 35% of returnees from the Iraq War accessed services in the year after return from deployment. However, only about 10% of post-Iraq mental health visits occurred in those who were initially PDHA screen positive. Seal et al. (2007) performed a record review among 103,788 OEF/OIF veterans who separated from the military and received care at VA medical centers. In their sample, 13,205 veterans (13%) received a chart diagnosis of PTSD. Of those 13,205 veterans, 5,844 (44%) were first diagnosed with PTSD by a non-behavioral health provider, like a primary care provider. Of the 5,844 veterans first diagnosed with PTSD by a non-behavioral health provider, 4,198 (72%) veterans followed-up with a behavioral health provider. And, 3925 (94%) of those veterans received a chart diagnoses of PTSD by the behavioral health provider. This high concordance rate may be inflated, as it does not include the 66% of the 13,205 veterans not first diagnosed with PTSD by a behavioral health provider. It is unknown whether those veterans saw a behavioral health provider first, who subsequently made the PTSD diagnosis, or saw a non-behavioral health provider first, who missed the PTSD diagnosis. Thus, if one considers a PTSD from a behavioral health specialist to be more valid than from a primary care provider, the overall 13% prevalence may be an overestimate.
Estimating the Disease Burden of Combat-Related Posttraumatic Stress Disorder
89
Smith et al. (2008) calculated a threefold increase in new onset PTSD among deployed servicemembers (i.e., soldiers, sailors, marines, and airmen) who report combat exposures. In this longitudinal study, data from the first panel of 50,128 servicemembers who participated in the millennium cohort study were used. Participants were evaluated pre- and post-deployment (on average, assessments were 2.7 years apart, SD = 0.5 years) and the researchers used the PTSD Checklist, a 17-item self report questionnaire which assess each of the 17 criterion B, C, and D symptoms of PTSD on a 1–5 Likert scale. Researchers employed two criteria to determine PTSD status: (1) PCL score of 50 or greater, a recommended cutscore or (2) endorsement of at least one reexperiencing, three avoidance, and two hypervigilance symptoms of at least moderate distress (item score of at least three) to match the DSM criteria for PTSD. When using the more conservative cutscore method, they found new onset PTSD in 7.6% of deployed servicemembers who reported combat exposures, 1.4% of servicemembers who were deployed without combat, and 2.3% of servicemembers who were not deployed. The DSM criteria method yielded new onset in 8.7% of deployed servicemembers with combat exposure, 2.1% of deployed without combat exposure, and 3.0% of non deployed servicemembers. Participants with the following characteristics had the highest rates of new onset PTSD: deployed, female, divorced, enlisted, current smoker or problematic drinking at baseline. These prevalence findings demonstrate that PTSD is an increasingly common problem that presents at multiple levels (e.g., primary care and in-patient) of the VA and DoD healthcare system. However a condition can be prevalent without being burdensome.
4
Course of Illness
Exposure to trauma and the development of PTSD. Exposure to traumatic events does not necessarily lead to the development of PTSD symptoms. Rates of traumatic exposure in the general population are quite common and as reported in Hidalgo and Davidson (2000), range from 39% (Breslau et al., 1991) or lower (28%, Hepp et al., 2006) to 90% (Breslau et al., 1998). Of those exposed, relatively few develop PTSD. For example, only 6% of men and 11.3% of women met criteria in Breslau et al. (1991). When stratified by type of trauma (reported in Kessler, 2000), the conditional probability of developing PTSD from combat exposure was 38.8% (in men). Rape had the highest conditional probability (65% men and 45.9% women), and the event with the conditional probability most similar to combat was sexual assault other than rape (12.2% in men and 26.5% in women). More recently, Magruder et al. (2005) reported that the odds of PTSD from serving in a war zone were 9-times greater (unadjusted OR; 19% prevalence) than for those veterans who did not serve in a war zone (4.4% prevalence). Hoge et al. (2006) reported that exposure to combat was correlated with a positive PTSD screen among OIF veterans. For example, compared with 95,894 (47.8%) of OIF returnees who screened negative for PTSD, of the 21,822 (9.8% of OIF sample) veterans who screened positive for PTSD (score of two or more on the 4-item screen; Prins et al., 2003), 79.6% reported ‘‘witnessing persons being wounded or killed or engaging in direct combat during which they discharged their weapon’’ (Hoge et al., 2006, pg 1028). Delayed onset. Of persons who develop PTSD, many experience delayed onset of symptoms (i.e., symptom onset occurs at least 6 months or later following exposure to traumatic event; APA, 2000). In a sample of Vietnam veterans (Schnurr et al., 2003), symptom onset occurred 1.34 years after entry into Vietnam, with 40% of their sample reporting that symptoms first
1535
1536
89
Estimating the Disease Burden of Combat-Related Posttraumatic Stress Disorder
occurred 2 or more years after entering Vietnam. McFarlane (2004) discusses delayed onset PTSD and presents findings from several studies using civilian and veteran populations. Most striking is that some people experience latency periods of decades. More recently, Grieger et al. (2006) reported that the majority (78.8%) of injured soldiers who reported PTSD and depression symptoms at 7 months post hospitalization reported no symptoms at 1 month. Chronicity of symptoms. PTSD is chronic for many veterans, with symptoms lasting more than 50 years (Magruder et al., 2005). In Vietnam veterans, average symptom duration was 18.5 years, with 92.5% of veterans experiencing symptoms for more than 5 years, and 40.2% of veterans reporting a duration of 20 or more years (Schnurr et al., 2003). Kessler et al. (1995) found that 60% of patients with PTSD reached spontaneous remission in 72 months, but type of trauma affected time to remission. In a personal communication with Kessler, McFarlane (2004) reported that time to remission from combat was 85 months (7 years). Even with treatment, approximately 1/3 of respondents did not remit at 120 months (Kessler et al., 1995), suggesting that these treatment refractory people will experience symptoms for their lifetime. In the US burden of disease study (Michaud et al., 2006), researchers estimated PTSD to last 4 years for males and 5 years for females. These estimates did not distinguish combat PTSD from noncombat PTSD. Researchers argue that risk and resilience factors such as social support, intensity, frequency, and duration of trauma, family history of mental illness, ethnicity, and sex can mediate the development and chronicity of PTSD (Friedman, 2006; Hoge et al., 2004; IOM, 2006; Kessler, 2000). Section summary. Although most people exposed to a traumatic event will not develop PTSD, a significant minority will. Of those that develop PTSD, many cases of PTSD will be chronic, and arguably, chronic for the patient’s lifetime. Still, a disease may be chronic and prevalent without being burdensome. Thus, we now examine the psychosocial impact of PTSD on patients and the associated strain on the healthcare system.
5
Impact on Patients and the Healthcare System
PTSD is associated with significant decrements in quality of life. In a sample of Vietnam veterans with chronic PTSD, baseline scores on a health-related quality of life measure (MOS SF-36; Ware and Sherbourne, 1992) were 1–2 standard deviations below the US general population, where lower scores indicate poorer functioning (Schnurr et al., 2006). Magruder et al. (2005) compared SF-36 scores of veterans with and without PTSD, and veterans with PTSD had significantly lower scores than veterans without PTSD. The differences remained even after the researchers controlled for age, race, study site, education, and co-occurring mental disorders. In a review, Seedat et al. (2006) reported that patients with PTSD have more impairments in quality of life than the general public and patients with other chronic physical conditions (e.g., diabetes, multiple sclerosis) and anxiety disorders. They report that PTSD is second only to depression in terms of the quality of life decrements associated with mental disorders. PTSD is also associated with educational failure, work impairment (an estimated $3 billion loss in productivity in the US), unemployment, and marital instability (Kessler, 2000). For example, OEF/OIF returnees reported a 4-fold increase in concerns about interpersonal conflict (Milliken et al., 2007), and have higher rates of attrition from the military (Hoge et al., 2006). OEF/OIF returnees with PTSD were associated with more sick call visits, greater somatic complaints, lower ratings of general health, more physical symptoms, and more missed workdays (Hoge et al., 2007). Most recently, Hoge et al. (2008) concluded that
Estimating the Disease Burden of Combat-Related Posttraumatic Stress Disorder
89
PTSD, and not mild traumatic brain injury, accounted for poor general health, missed workdays, medical visits, and a high number of somatic and post-concussive symptoms (except headache) in a sample of returnees. PTSD is associated with a significant use of healthcare resources. As reviewed by Hidalgo and Davidson (2000) and Kessler (2000), PTSD patients use more medical services than patients without PTSD, and veterans with PTSD were more likely to have dermatologic, gastrointestinal, ophthalmologic, endocrinologic, gynecologic, and cardiovascular problems. PTSD patients seek medical care more often than mental health care. Kessler (2000) summarized reasons why medical services were used more often than mental health services in those with PTSD in a large community study. Males (66.2%) and females (60.0%) believed their problems did not warrant treatment. Males (66.2%) and females (40.4%) believed treatment would be ineffective. Males (42.4%) and females (66.5%) thought the problem would resolve on its own. However, the symptoms resolved in only 4% of males and 32.6% of females. Hoge et al. (2004) reported 86% of returnees who screened positive for a depression, PTSD, or generalized anxiety disorder acknowledged having a problem, but only 45% of those respondents were interested in receiving help, and approximately half of those interested actually received help. Moreover PTSD frequently goes undetected (Hidalgo and Davidson, 2000; Magruder et al., 2005), and therefore untreated, which may contribute to ongoing problems and continued seeking of medical services. War veterans are also eligible to receive disability compensation, called a service-connection, if they meet criteria for PTSD resulting from combat exposure. These benefits are unique to the military and VA healthcare systems, and they pose various assessment, treatment, and health policy challenges because benefits can be denied, reduced, or increased depending on PTSD diagnosis and symptom severity (IOM and NRC, 2007). The number of beneficiaries from fiscal years 1999 to 2004 grew by 79.5%, from 120,265 to 215,871 beneficiaries, and the amount of payment increased 148.8% to $4.28 billion (reported in IOM and NRC, 2007). Veterans can receive over $3,000 per month, depending on their family situation, plus free health care and other benefits. Although there is debate about how compensation for a serviceconnected disability may serve as an incentive for veterans to remain symptomatic (e.g., to improve might mean the loss of substantial benefits; IOM and NRC, 2007), effectively identifying and treating war veterans with PTSD may prevent the need for such lifetime compensation and thereby reduce some of the financial burden to DoD and VA.
6
Measuring Disease Burden
Introducing disease burden and the concept of ‘‘preference.’’ The true burden of a disorder may be defined as ‘‘the burden in the absence of treatment, that is calculated from the burden observed in the population under study plus the burden presently averted by current population coverage and mix of interventions’’ (Andrews et al., 2004, pg. 526). The burden of disease in a population in a given year is the sum of a) years of life lost due to premature deaths (YLL) and b) estimate of the future years of life with disability (YLD) for new cases of disease or injury, weighted for severity (Vos et al., 2001). Disease burden is the confluence of two measurements: morbidity and mortality. Mortality is rarely attributed to mental disorders (Andrews et al., 2004). However, persons with PTSD do attempt suicide at rates of up to 8.2 times of the general population, even after controlling for other disorders, like depression (reported in McFarlane, 2004). Put more simply, disease burden measures
1537
1538
89
Estimating the Disease Burden of Combat-Related Posttraumatic Stress Disorder
pervasiveness (e.g., incidence or prevalence as the number of cases, not the percentage), persistence (that is, course or duration of illness), and impact (that is, death and/or decrement in preference-weighted health status associated with the disease at any given time) in one unit of measurement. Morbidity of a disease is basically measured using one of two scales (both ranging from zero to one), anchored by perfect health and death. The number between 0 and 1 is considered a preference-weight for a particular disease or health state. To estimate the disease burden, the > preference weight is multiplied by a) the amount of time spent in that health state and b) the disease incidence. Preference weights differ from other measures of severity, in that preference-weighted metrics incorporate how a population determines the value or worth of a health condition. For example, Kiberd and Lawson (2000) measured the preference of a Type-I diabetes mellitus (DM) with comorbid blindness and end stage renal disease (ESRD; requiring dialysis) health state from patients with DM and ESRD. In a separate study, Revicki and Wood (1998) measured the preference of a severe depression health state from patients being pharmacologically treated for depression. Preference was measured in units called health utilities (discussed below) by both sets of researchers. Health utilities (and other units of preference) are valued by having a group of raters read a brief vignette describing the health states (depression or DM with ESRD in this example). Raters are administered one or more of econometric instruments (e.g., the person tradeoff; PTO; or > standard gamble; both methods described below) designed to elicit preference. In these two examples, average preference for the severe depression health state was less desirable than was average preference for DM with comorbidities. Thus, a health state like DM with comorbidities, which requires frequent life-saving medical treatment, is more desirable than severe depression, which is theoretically treatable with mediation and/or psychotherapy. Depression is life-threatening only if a patient attempts or completes suicide, unlike DM with ESRD. This example demonstrates that non preference based measures of health status and clinical symptom focused measures are predominantly clinician tools, not health policy tools per se (Bennett and Torrance, 1996; Berzon et al., 1996; Garza and Wyrwich, 2003; Tsevat, 2000). Disability weights associated with PTSD. The Global Burden of Disease study (GBD; Murray and Lopez, 1996) assessed disease burden using the disability adjusted life year (DALY), which equals YLL plus YLD. Here, zero represents perfect health and one represents death, such that a year in perfect health is worth 0 DALYs. DALYs are considered an interval scale, such that they can be summed across individuals and over time. ‘‘The DALY measures the gap between the actual health of a population and a hypothetical norm, namely a life expectancy of 82.5 years for women and 80 years for men. DALYs for a disease or health condition are calculated as the sum of the years of life lost due to premature mortality in the population and the years of life lost due to disability’’ (Michaud et al., 2001, pg. 535). DALYs can also include life expectancy, and adjust for age, such that a young adult is considered to be ‘‘worth’’ more than someone at the beginning or end of the life cycle. A component of YLD is a value called a disability weight, where 0 equals perfect health and 1 equals death. In the GBD study and other studies which assessed DALYs (e.g., Kruijshaar et al., 2005), disability weights for DALYs were obtained for 22 ‘‘indicator conditions,’’ including depression, by administering a series of ‘‘person tradeoff ’’ (PTO) vignettes to a sample of clinicians. Two variants were used. First, clinicians were given a choice between trading quantity of life for healthy individuals and disabled individuals. In the second, they are asked to tradeoff quantity of life for healthy individuals versus improved quality of life for a group of disabled individuals’’ (Murray and Lopez, 1996, pg. 91). As an example, a disability
Estimating the Disease Burden of Combat-Related Posttraumatic Stress Disorder
89
. Figure 89‐1 Person tradeoff method used in Murray and Lopez (1996) and Sanderson and Andrews (2001): ‘‘trading quantity of life for healthy and disabled individuals’’ (pg. 670). This is one variant of the person tradeoff (PTO), a method designed to determine preference and ‘‘worth’’ for a particular disease state. Respondents indicate their preference by valuing quantity of life in a hypothetical sample of healthy and disabled individuals. A low disability weight (close to 0) indicates that the disease is considered less disabling, and conversely, a high disability weight (close to 1) indicates that the disease is considered more disabling
weight would equal 0.5, if a respondent considered 2,000 people with depression to represent the same amount of health as 1,000 people without depression (Sanderson and Andrews, 2001). See > Figures 89‐1 and > 89‐2 for a graphical representation of the PTO. The GBD study valued hundreds of conditions. Because it was not feasible to administer the PTO for each condition, a short-cut estimation procedure was developed. The 22 indicator conditions were used to define seven disability categories. PTSD was one of the conditions valued using this short-cut procedure. In a replication study Sanderson and Andrews (2001) administered the PTO to a panel of physicians for 20 vignettes capturing mental disorders, including PTSD. They also used a rating scale method, on which PTSD is valued from 0 to 100. The differing methodologies led to very different disability weights for PTSD (see > Table 89‐3). Not only is method variance an important concern for the health economics field, but these findings suggest the GBD study grossly underestimated the burden of PTSD. Perhaps, as Kessler (2000) suggests, the burden of PTSD is akin to that of depression, which is projected to rank second (Murray and Lopez, 1996). The GBD study did not evaluate treated versus untreated PTSD, as it did for depression. Andrews et al. (2004) found that treatment of PTSD can account for between 11 and 32% of the disease burden, depending on gross categories of treatment coverage and treatment type.
1539
1540
89
Estimating the Disease Burden of Combat-Related Posttraumatic Stress Disorder
. Figure 89‐2 Second variant of the PTO used in Murray and Lopez (1996) and Sanderson and Andrews (2001): ‘‘trading quantity of life for healthy individuals versus improved quality of life for disabled individuals’’ (pg 670). This is another variant of the person tradeoff (PTO), a method designed to determine preference and ‘‘worth’’ for a particular disease state. Respondents indicate their preference by valuing quantity and quality of life in a hypothetical sample of healthy and disabled individuals. A low disability weight (close to 0) indicates that the disease is considered less disabling, and conversely, a high disability weight (close to 1) indicates that the disease is considered more disabling
Health utilities associated with PTSD. A similar metric is called a quality adjusted life year (QALY). Here, a year in perfect health equals one QALY and death equals zero, the reverse of the aforementioned DALYs. Like DALYs, the number representing disease preference in QALYs is weighted by time spent in a given disease state and by the number of persons with the disease. Unlike DALYs, the number is not adjusted for age. The number is called a > health utility (versus a disability weight). We recognize that the nomenclature used to describe the preference weight needed to calculate a QALY differs according to the method. To minimize jargon, we use the term health utility, independent of method, when referencing the preference based unit of measurement associated with a QALY. Of note, numerous methods can be used to obtain a health utility (e.g., interpolation from a generic health-related quality of life scale or direct valuation of a health state vignette from a community sample). And, as discussed by Sherbourne et al. (2001), each method can produce a wide range of health utilities. Health utilities may or may not be sensitive enough to detect changes in symptom severity, and are therefore questionable indicators of treatment outcomes (Sherbourne et al., 2001). Nevertheless, health utilities (and the eventual calculation of QALYs) are the recommended outcome in costeffectiveness analyses (Gold et al., 1996).
Estimating the Disease Burden of Combat-Related Posttraumatic Stress Disorder
89
. Table 89‐3 Existing disability weights for PTSD and depressive disorders Source Michaud et al. (2006); Murray and Lopez, (1996)
Method Short-cut estimation
Disease
Disability weight
PTSD
0.11
Sanderson and Andrews, (2001) PTO
PTSD
0.66 (SD = 0.25)
Sanderson and Andrews, (2001) Rating scale
PTSD
0.57 (SD = 0.22)
Murray and Lopez, (1996)
Comparator: unipolar depression/ treated depression
PTO
Sanderson and Andrews, (2001) PTO
Comparator: mild/moderate/ severe depression
0.6/0.30 0.09/0.34/ 0.70
This table presents existing disability weights for PTSD and the methods utilized. It also presents a selected sample of utility weights used to calculate the disease burden of depression. PTO person tradeoff; PTSD posttraumatic stress disorder; SD standard deviation
The standard gamble (> Figure 89‐3) is the classic method of eliciting health utilities. Here, respondents choose between (1) a health state with certainty or, (2) a gamble, where treatment could lead to perfect health with some probability of death. The point of indifference is called the health utility. In this paradigm, the health state can be a vignette describing symptoms and disability associated with a specific disease (e.g., depression; Revicki and Wood, 1998). Or, a health state can be a description of generic health-related quality of life impairments (Brazier et al., 2002). Respondents are then administered the standard gamble multiple times, where the researchers vary the value of probability (p), until the respondent is indifferent between Choices 1 and 2. The procedure is repeated for multiple health states within a specific disease (e.g., severe depression, moderate depression, mild depression) or by for multiple severity levels of the domains of health–related quality of life. To our knowledge, the first health utilities for PTSD were only recently calculated (Freed et al., 2007). Here, our team of scientists estimated health utilities by stratifying a sample of veterans based on diagnosis and examining their responses to a preference-weighted measure of health status, the Short-Form 6-dimensional health state classification (SF-6D), an abbreviated version of the Medical Outcome Survey Short-Form-36 (MOS SF-36; Brazier et al., 2002; Ware and Sherbourne, 1992). Magruder et al. (2005) administered questions of the MOS SF-36 (Ware and Sherbourne, 1992) to a large, multi-site primary care sample of veterans. We applied Brazier et al.’s (2002) health utility formula to the SF-36 responses in Magruder et al. (2005). Brazier et al. previously determined the health utilities of SF-6D-defined health states by administering the standard gamble to a community sample. Brazier et al. then intended for these health utilities to be mapped onto the raw scores of an SF-36 dataset without having to re-administer the standard gamble. We then explored the relationship among PTSD, other anxiety disorders, and mood and substance use disorders on health utilities. Mental health disorders were diagnosed using well-validated diagnostic interviews. > Figure 89‐4 presents the
1541
1542
89
Estimating the Disease Burden of Combat-Related Posttraumatic Stress Disorder
. Figure 89‐3 The standard gamble paradigm for calculating a health utility for a particular health state (figure adapted from Freed and Engel, 2006). An example of the standard gamble, a method designed to determine preference and ‘‘worth’’ for a particular disease state. Respondents value a disease by indicating their willingness to accept a risky but disease-curing intervention versus living with the disease. A low health utility (close to 0) indicates that the disease is considered very undesirable, and conversely, a high health utility (close to 1) indicates that the disease is not considered undesirable
mean health utilities and standard error associated with veterans diagnosed with (PTSD+) and without (PTSD-) PTSD and with and without one co-occurring mood, anxiety, and substance use disorder (SUD). The authors concluded that PTSD contributes to decrements in health status, with greater decrements associated with the presence of more than one co-occurring disorder. Furthermore, the health utilities of PTSD were comparable to those for mood, anxiety, and SUD alone. The findings were limited in two important ways. First, the SF-6D preferences were obtained from a British community sample (Brazier et al., 2002), as opposed to veterans or active duty servicemembers. And second, the Magruder et al. (2005) data was collected from veterans before the conflicts in Iraq and Afghanistan, which potentially calls into question the generalizability of these findings to returnees from OEF/OIF. Section summary. Both the disability weights in DALYs and the health utilities in QALYs, are preference-based. Preference-based measures differ form other symptom severity or health-related quality of life measures. Non preference-based measures describe functioning while preference-based measures include a value component, which assesses the relative worth of diseases or health states. Gold et al. (2002) and Sassi (2006) provide detailed discussions of the similarities and differences between QALYs and DALYs.
Estimating the Disease Burden of Combat-Related Posttraumatic Stress Disorder
89
. Figure 89‐4 Estimated mean health utilities with standard error bars for PTSD and co-occurring mental health conditions from Freed et al. (2007). SUD: diagnosed with a substance use disorder; PTSD+: diagnosed with posttraumatic stress disorder; PTSD-: not diagnosed with posttraumatic stress disorder; Mood: diagnosed with one mood disorder; Anxiety: diagnosed with one anxiety disorder. This figure presents the mean estimated health utilities with standard error bars for several combinations of diagnostic groups. PTSD+ veterans had significantly lower health utility scores (indicating greater dysfunction) than veteran without a diagnosed mood, anxiety, or substance use disorder. PTSD+ veterans diagnosed with greater than one co-occurring disorder (e.g., a mood and substance use disorder) had significantly lower health utilities than PTSD-veterans with greater than one co-occurring disorder
7
Efficacy of Primary Prevention and Treatment
The VA/DoD clinical practice guidelines for the management of posttraumatic stress (VHA/ DoD, 2004) suggest that primary prevention strategies should be considered. The US military arguably provides months to years of training in efforts to prepare soldiers for the challenges they may face in combat situations. This training can be seen as primary prevention, and a paucity of research exists which details how these strategies affect a military population. More recently, the military implemented pre- and post-deployment training modules called ‘‘Battlemind’’ (Adler et al., 2007; http://www.battlemind.org/), which can be administered, to soldiers, healthcare professionals, leaders, and family members. The training provides mental health education about the reality of combat situations and readjustment issues upon return. Results of this training have lead to lower PTSD, lower depression, less anger, and less perceived stigma about receiving treatment (Adler et al., 2007). More research needs is needed to determine whether this particular intervention can reduce the incidence of PTSD following combat exposure. In contrast to prevention, much research exists on the efficacy of PTSD treatment. A full review of the efficacy literature is beyond the scope of this chapter, but several recent reviews determined psychotherapeutic (Bisson et al., 2007; Bradley et al., 2005) and pharmacologic (Davidson, 2006; Ipser et al., 2006) interventions for PTSD can be effective. The effectiveness of varying types of therapy or drug treatments depend on the type of intervention and samples studied. In general, treatments were less effective for combat-trauma (Bisson et al., 2007; Bradley et al., 2005; Ipser et al., 2006) relative to non-combat trauma, but the patients in the combat-trauma studies were veterans of other wars (namely Vietnam) and not OEF/OIF. Friedman (2006) discusses specific treatments options for OEF/OIF returnees with PTSD. However, existing barriers to seeking care limit the potential for servicemembers to reap the benefits of treatment.
1543
1544
89 8
Estimating the Disease Burden of Combat-Related Posttraumatic Stress Disorder
Other Considerations
Clearly, the foundation of estimating the disease burden of PTSD rests on multiple factors. So far we have reviewed current prevalence estimates, course of illness, disability weights, and treatment efficacy. In this section, we discuss two more variables that contribute to the estimation of disease burden: techniques used to determine disease prevalence and the consideration of comorbidities. PTSD assessment methods. The VA/DoD clinical practice guidelines (2004) call for a thorough evaluation to accurately diagnose PTSD. It is often impractical to include lengthy, clinician-administered diagnostic evaluations in large epidemiological studies, and therefore self-report measures are used in place of a clinical interview. Although useful and efficient, screening measures such as the 17-item PCL (Blanchard et al., 1996) and the 4-item PC-PTSD (Prins et al., 2003) do not assess Criterion A. For example, Prins et al. (2003); corrigendum (2004) compared the two measures in relation to a structured diagnostic clinician administered PTSD interview. At a cutscore of 3 on the PC-PTSD, sensitivity (proportion of people with PTSD who screen positive) was 0.78 and specificity (proportion of people without PTSD who screen negative) was 0.78. At a cutscore of 48 on the PCL, sensitivity was 0.84 and specificity was 0.90. The cutscores used to indicate presumptive PTSD diagnosis may differ depending on the sample in question, which has a significant impact on the reported prevalence of PTSD across research studies. Since disease prevalence is a component of disease burden, considerations of assessment methods used to derive prevalence estimates are important. Dealing with comorbidities. Disease co-morbidities must also be considered when estimating disease burden. For example, of the veterans diagnosed with PTSD, Magruder et al. (2005) found that 68.6% met criteria for major depression, 73.3% met criteria for another anxiety disorder, and 10.5% met criteria for a substance use disorder, and thus 87.2% of veterans diagnosed with PTSD also met criteria for another mental disorder. In an injured servicemember population (Grieger et al., 2006), the prevalence of PTSD co-occurring with depression was 46% at 1-month, 63% at 4-months, and 53% at 7-months. Smith et al. (2008) found that the odds of new onset PTSD were greater for servicemembers who smoked and who met criteria for problem drinking, although these problems were reported as risk factors, rather than co-morbid conditions. In other reviews (Kessler, 2000), the evidence suggests that co-occurring disorders appear after PTSD, and are due to PTSD and not an underlying vulnerability. Not surprisingly, comorbidity impacts estimates of disease burden, especially when disability weights and health utilities are calculated from disease specific vignettes that do not include the co-occurring condition. Here, one must ask what the impact is of the co-occurring condition on the preference weight. Is the impact additive, for example, such that the preference weight of one disorder is added to the weight of a second disorder for a person’s comorbidity? Comorbidity is a problem in both DALY and QALY calculations, but it has not been thoroughly addressed in disability weighing. Melse et al. (2000) briefly summarize the issue, and state that summing the weights of comorbid conditions is an acceptable solution because the inclusion of comorbidity insignificantly changes disease burden calculations. Using this summing approach, however, permits a person to have a disability weight of greater than one. Thus, if that person is diagnosed with multiple conditions, thereby creating a disability weight greater than one, then theoretically, his disability is worse than death. Freed et al.’s findings (2007; > Figure 89-4) suggest the recommendations of Melse et al. (2000) may not be an acceptable solution because the impact of a second or third co-occurring disorder is not cumulative. Vos et al. (2001) split the number of prevalent cases of cooccurring disorders equally across their sample, thereby allowing for the fact that people
Estimating the Disease Burden of Combat-Related Posttraumatic Stress Disorder
89
with multiple diagnoses would be likely to have more severe disease weightings than those with only one diagnosis. One solution is to create vignettes based on generic health status, establish preference weights for those generic vignettes, and then assess patients with the generic health status measure. Then, it is possible to apply the existing preference weights to the patient responses and examine the relationship between generic health status and specific diseases (e.g., PTSD).
9
Conclusions
Despite an increasing incidence, chronic nature, drain on the healthcare system, and a significant decrement in health status associated with combat PTSD, the major burden of disease is largely unrecognized at a policy level (McFarlane, 2004). To date, Freed et al. (2007) are the only researchers who attempted to estimate health utilities (a morbidity outcome) associated with PTSD in a veteran sample, a morbidity. Freed et al. did not calculate QALYs associated with PTSD, nor did they weight the health utilities by disease prevalence or incidence in the primary care sample. Thus, disease burden estimates from Freed et al. alone are not possible. The other preference-based estimates (> Table 89‐3) did not distinguish combat from noncombat PTSD. The primary method to estimate the disease burden of PTSD is to calculate the years lived YLD. Here, incidence of disease (that is, the number of new cases in one year) is multiplied by the average time spent in the disease state, and the product is then multiplied by the disability weight (Murray and Lopez, 1996). The incidence of PTSD is not widely studied, but the prevalence can be also be used instead of incidence. Melse et al. (2000) multiplied disease prevalence (that is, the number of new cases in one year) by disability weight. If disability weights change over time (e.g., disease worsens with age), then researchers can average the weights and use that average. The choice to use prevalence- or incidence-based estimates is often dependent on the goal of the study (Melse et al., 2000). Incidence-based calculations are more appropriate for disease prevention programs while prevalence-based calculations are presentcentered. Other choices in disease burden estimates include what discounting rate to use and whether to use age weightings. For example, ‘‘discounting is applied because people value their current health more than future health, and age-weighting is applied because most societies would choose to save young to middle-aged adults over the young or very old’’ (Andrews et al., 1998, pg. 123). Ultimately, disease burden estimates are just that, estimates of population health. The estimates are used for comparison purposes to better assist policy and decision makers to reduce that burden and prioritize valuable healthcare resources to better treat and prevent illness. PTSD is a chronic, recurring, but treatable disorder affecting a high percentage of military servicemembers and veterans. Although no disease burden estimates exist to date specific to this population, the breadth of current literature can provide researchers with enough information to reasonably estimate the disease burden of PTSD.
Summary Points PTSD is an anxiety disorder characterized by re-experiencing, avoidance, and hyperarousal symptoms following a traumatic event.
1545
1546
89
Estimating the Disease Burden of Combat-Related Posttraumatic Stress Disorder
Although the WHO characterized the burden to be among the top 20 most burdensome
diseases in the US, others have argued that PTSD is as burdensome as depression, which the WHO predicts will be the second most burdensome disease worldwide. Extant disease burden estimates do not distinguish combat-related from noncombat-related PTSD. Combat exposure is a risk factor for PTSD, and people with combat-related PTSD may respond as well to treatment. PTSD prevalence estimates from OEF/OIF returnees range from 4.2 to 24.5%, depending on the specific subpopulation and assessment method. We review the literature for several key components to estimating the disease burden of combat PTSD in US active duty servicemembers and veteran populations prevalence; course of illness; impact on the healthcare system; preference weights; and intervention effectiveness. PTSD is associated with low quality of life, high rates of medical service utilization, interpersonal conflict, co-morbidity, somatic complaints, lower general health ratings, and work impairment compared to military servicemembers without PTSD. PTSD is chronic, with symptoms lasting over 50 years in some veterans. Disease burden is the confluence of two measurements, morbidity and mortality, although mortality is not typically included when assessing the disease burden of mental health conditions. Unlike symptom severity or generic quality of life, morbidity (measured by a preferenceweighted metric) accounts for the value or worth that society places on a particular disease state, relative to other disease states and anchored by perfect health and death. Once calculated, disease burden for combat PTSD can assist in medical decision making and in determining priorities for interventions.
References Adler AB, Wright KM, Bliese PD, Eckford R, Hoge CW. (2008). J Trauma. 21: 301–308. Adler AB, Castro CA, McGurk D. (2007). US Army Medical Research Unit-Europe Research Report 2007–001. USAMRU-E, Heidelberg, Germany. American Psychiatric Association. (1980). Diagnostic and Statistical Manual of Mental Disorders, 3rd ed. American Psychiatric Association, Washington, DC. American Psychiatric Association. (1987). Diagnostic and Statistical Manual of Mental Disorders, 3rd ed, (text revision). American Psychiatric Association, Washington, DC. American Psychiatric Association. (2000). Diagnostic and Statistical Manual of Mental Disorders, 4th ed, (text revision). American Psychiatric Association, Washington, DC. Andrews G, Issakidis C, Sanderson K, Corry J, Lapsley H. (2004). Br J Psychiatry. 184: 526–533. Andrews G, Sanderson K, Beard J. (1998). Br J. Psychiatry. 173: 123–131.
Army Medical Surveillance Activity. (2004). Defense Medical Epidemiology Database (DMED 3.7) [computer software]. Available at: http://amsa.army.mil/ AMSA/amsa_home.htm. Bennett KJ, Torrance GW. (1996). Measuring health state preferences and utilities: rating scale, time trade-off, and standard gamble techniques. In Spiker B. (ed.), Quality of Life and Pharacoeconomics in Clinical Trials, 2nd ed. Lippincott-Raven Publishers, Philadelphia, PA. Berzon RA, Mauskoph JA, Simeon GP. (1996). Choosing a health profile (descriptive) and/or a patientpreference (utility) measure for a clinical trial. In B. Spiker (ed), Quality of Life and Pharacoeconomics in Clinical Trials, 2nd ed. Lippincott-Raven Publishers, Philadelphia, PA. Bisson JI, Ehlers A, Matthews R, Pilling S, Richards D, Turner S. (2007). Br J Psychiatry. 190: 97–104. Blake DD, Weathers FW, Nagy LM, Kaloupek DG, Gusman FD, CharneyDS, Keane TM. (1995). J Trauma Stress. 8: 75–90.
Estimating the Disease Burden of Combat-Related Posttraumatic Stress Disorder Blanchard EB, Jones-Alexander J, Buckley TC, Forneris CA. (1996). Behav Res Ther. 34: 669–673. Bradley R, Greene J, Russ E, Dutra L, Westen D. (2005). Am J Psychiatry. 162: 214–227. Brazier J, Roberts J, Deverill M. (2002). J Health Econ. 21: 271–292. Breslau N, Davis GC, Andreski P, Peterson E. (1991). Arch Gen Psychiatry. 48: 216–222. Breslau N, Kessler RC. (2001). Biol Psychiatry. 50: 699–704. Breslau N, Kessler RC, Chilcoat HD, Schultz LR, Davis GC, Andreski P. (1998). Arch Gen Psychiatry. 55: 626–632. Davidson JR. (2006). J Clin Psychiatry. 67: 34–39. Davidson JRT, Foa EB. (1991). J Abnorm Psychology. 100: 346–355. Freed MC, Engel CC. (2006). Behavioral Health Track at the U.S. Army Center for Health Promotion and Preventive Medicine-hosted Force Health Protection Conference, Albuquerque, NM. Freed MC, Yeager D, Liu X, Gore KL, Engel CC, Magruder KM. (2007). Behavioral Health Track at the U.S. Army Center for Health Promotion and Preventive Medicine-hosted Force Health Protection Conference, Louisville, KY. Friedman MJ. (2006). Am J Psychiatry. 163: 586–593. Garza AG, Wyrwich KW. (2003). Acad Emerg Med. 10 (4): 360–363. Gold MR, Siegel JE, Russell LB, Weinstein MC. (1996). Cost-Effectiveness in Health and Medicine. Oxford University Press, New York. Gold MR, Stevenson D, Fryback DG. (2002). Annu Rev Public Health. 23: 115–134. Gore KG, Engel CC, Freed MF, Liu XL, Armstrong DW. (2008). Gen Hosp Psychiatry. 30: 391–397. Gore KG, Freed MC, Messer SC, Engel CC. (2006). Behavioral Health Track at the U.S. Army Center for Health Promotion and Preventive Medicinehosted Force Health Protection Conference, Albuquerque, NM. Grieger TA, Cozza SJ, Ursano RJ, Hoge C, Martinez PE, Engel CC, Wain HJ. (2006). Am J Psychiatry. 163: 1777–1783. Hepp U, Gamma A, Milos G, Eich E, Adjacic-Gross V, Rossler W, Angst J, Schnyder U. (2006). Eur Arch Psychiatry Clin Neurosci. 256: 151–158. Hidalgo RB, Davidson JR. (2000). J Clin Psychiatry. 61: 5–13. Hoge CW, Auchterlonie JL, Miliken CS. (2006). JAMA. 295: 1023–1032. Hoge CW, Castro CA, Messer SC, McGurk D, Cotting DI, Koffman RL. (2004). N Engl J Med. 351: 13–22. Hoge CW, McGurk D, Thomas JL, Cox AL, Engel CC, Castro CA. (2008). N Engl J Med. 358: 453–463. Hoge CW, Terhakopian A, Castro CA, Messer SC, Engel CC. (2007). Am J Psychiatry. 164(1): 150–153.
89
Institute of Medicine (IOM), National Research Council (NRC), Committee on Veterans Compensation for Posttraumatic Stress Disorder. (2007). PTSD Compensation and Military Service. National Academies Press, Washington, DC. Available at: http://www. nap.edu/catalog/11870.html. Institute of Medicine (IOM), Subcommittee on Posttraumatic Stress Disorder of the Committee of Gulf War and Health. (2006). Posttraumatic Stress Disorder: Diagnosis and Assessment. National Academies Press, Washington, DC. Available at: www.nap.edu/ catalog/11674.html. Ipser J, Seedat S, Stein DJ. (2006). S Afr Med J. 96: 1088–1096. Jones E, Wessely S. (2007). J Anxiety Disord. 21: 164–175. Kessler RC. (2000). J Clin Psychiatry. 61: 4–12. Kessler RC, Sonnega A, Bromet E, Hughes M, Nelson CB. (1995). Arch Gen Psychiatry. 52: 1048–1060. Kiberd BA, Lawson T. (2000). Transplantation. 70(7): 1121–1127. Kimerling R, Gima K, Smith MW, Street A, Frayne S. (2007). Am J Public Health. 97: 2160–2166. Kruijshaar ME, Hoeymans N, Spijker J, Stouthard ME, Essink-Bot ML. (2005). Bull World Health Organ. 83: 443–448. Kulka RA, Schlenger WE, Fairbank JA, Hough RL, Jordan BK, Marmar CR, Weiss DS. (1990). Trauma and the Vietnam War Generation: Report of Findings from the National Vietnam Veterans Readjustment Study. Brunner/Mazel, New York. Magruder KM, Frueh BC, Knapp RG, Davis L, Hamner MB, Martin RH, Gold PB, Arana GW. (2005). Gen Hosp Psychiatry. 27: 169–179. McFarlane A. (2004). Soc Psychiatry Psychiatr Epidemiol. 39: 874–882. Melse JM, Essink-Bot ML, Kramers PG, Hoeymans N. (2000). Am J Public Health. 90: 1241–1247. Michaud CM, McKenna MT, Begg S, Tomijima N, Majmudar M, Bulzacchelli MT, Ebrahim S, Ezzati M, Salomon JA, Kreiser JC, Hogan M, Murray CJ. (2006). Popul Health Metr. 4:11. Michaud CM, Murray C, Bloom B. (2001). JAMA. 285: 535–539. Milliken CS, Auchterlonie JL, Hoge CW. (2007). JAMA. 298: 2141–2148. Murray CJL, Lopez AD (eds). (1996). The Global Burden of Disease: A Comprehensive Assessment of Mortality and Disability from Diseases, Injuries, and Risk Factors in 1990 Projected to 2020. Harvard School of Public Health, Cambridge, MA. Prins A, Ouimette P, Kimerling R, Cameron RP, Hugeishofer DS, Shaw-Hegwer J, Thraikill A, Gusman FD, Sheikh JI. (2003). Primary Care Psychiaty. 9: 9–14; Corrigendum (2004) Primary Care Psychiatry. 9(4): 151. Revicki DA, Wood M. (1998). J Affect Disord. 48: 25–36.
1547
1548
89
Estimating the Disease Burden of Combat-Related Posttraumatic Stress Disorder
Sanderson K, Andrews G. (2001). Aust N Z J Psychiatry. 35: 668–676. Sassi F. (2006). Health Policy Plan. 21: 402–408. Schnurr PP, Hayes AF, Lunney CA, McFall M, Uddo M. (2006). J Consult Clin Psychol. 74: 707–713. Schnurr PP, Lunney CA, Sengupta A, Waelde LC. (2003). J Trauma Stress. 16: 545–553. Seal KH, Bertenthal D, Miner CR, Sen S, Marmar C. (2007). Arch Intern Med. 167(5): 476–482. Seedat S, Lochner C, Vythilingum B, Stein DJ. (2006). Pharmacoeconomics. 24: 989–998. Sherbourne C, Unutzer J, Schoenbaum M, Duan N, Lenert LA, Sturm R, Wells KB. (2001). Med Care. 39: 1246–1259.
Smith TC, Ryan MA, Wingard DL, Slymen DJ, Sallis JF, Kritz-Silverstein D. (2008). BMJ. 336: 366–371. Tsevat J. (2000). Med Care. 38(9) II160–II164. Veterans Health Administration (VHA), Department of Defense (DoD). (2004). VA/DoD0 Clinical Practice Guideline for the Management of Post-Traumatic Stress. Veterans Health Administration, Department of Defense, Washington, DC. Vos T, Mathers C, Herrman H, Harvey C, Gureje O, Bui D, Watson N, Begg S. (2001). Soc Psychiatry Psychiatr Epidemiol. 36: 53–62. Ware JE, Sherbourne CD. (1992). Med Care. 30: 473–483. Weathers FW, Keane TM. (2007). J Trauma Stress. 20: 107–121.
90 Estimating the Disease Burden of Seasonal Affective Disorder M. C. Freed . R. L. Osborn . K. J. Rohan 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1550
2 2.1 2.2 2.3 2.4 2.5 2.6
What Is SAD? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1551 Prevalence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1553 Treatment Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1554 Measuring Disease Burden . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1556 Course of Illness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1560 Other Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1564 Bringing It All Together . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1565 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1565
Disclaimer: The opinions and assertions expressed herein are those of the authors and are not to be construed as expressing the views or policies of the Deployment Health Clinical Center, United States Army, Uniformed Services University of the Health Sciences, Department of Defense, United States Government, or The University of Vermont. #
Springer Science+Business Media LLC 2010 (USA)
1550
90
Estimating the Disease Burden of Seasonal Affective Disorder
Abstract: > Seasonal Affective Disorder (SAD) is defined as a recurrent pattern of Major Depressive Episodes with a temporal relationship between the episodes and time of year (e.g., onset during fall/winter with remission during spring/summer) that affects 10–20% of major depression cases. Because SAD is a subtype of recurrent major depression and shares many symptoms, it also presumably shares the same disability during symptomatic months. In an alarming trend, the WHO predicts that by 2020, unipolar depression will rank second only to ischemic heart disease in disease burden. Although no SAD disease burden estimates exist to date, extant SAD literature is sufficient in scope and content to provide the information needed to estimate the disease burden, using the major depression literature as a proxy if necessary. We summarize and present literature that addresses key components to estimating disease burden: > prevalence; treatment efficacy, existing > preference weights, and course of illness. Current SAD prevalence estimates vary between 0.4 and 16% in the general adult population, depending on latitude, age, gender, and measurement method. Longitudinal data suggest that SAD remits in up to 27% of patients, becomes more complicated in presentation in up to 44% of patients, and remains in 42% of patients. Efficacy data suggests that treatment can lead to remission in up to 83% of patients, depending on treatment type. We present preference weights from the major depression literature, because they do not exist for SAD. > Disability weights range from 0.09 (minor depression) to 0.83 (severe depression). Similarly, health utilities range from 0.83 (mild depression) to 0.15 (severe depression). SAD is a chronic, disabling, but treatable condition. Disease burden estimates of SAD could be used to better determine the need for and utility of therapeutic interventions designed to alleviate symptoms of SAD and associated impairment. List of Abbreviations: APA, American Psychiatric Association; CBT, > cognitive behavioral therapy; CI, confidence interval; DALY, disability adjusted life year; DSM IV-TR, diagnostic and statistical manual of mental disorders IV – text revision; GBD, global burden of disease; ICD-10, international classification of disease-10; LT, > light therapy; NDA, no data available; QALY, > quality adjusted life year; SAD, seasonal affective disorder; SHQ, Seasonal Health Questionnaire; SIGH-SAD, structured interview guide for the Hamilton depression rating scale-seasonal affective disorder version; SPAQ, seasonal pattern assessment questionnaire; SSRIs, selective serotonin reuptake inhibitors; WHO, World Health Organization; YLD, years of life with disability; YLL, years of life lost
1
Introduction
As of 2000, the World Health Organization (WHO) estimated that unipolar (major) depressive disorders were the fourth most common cause global disease burden in women and the seventh in men (Ustu¨n et al., 2004). In an alarming trend, the WHO predicts that by 2020, unipolar depression will rank second only to ischemic heart disease in disease burden (Michaud et al., 2001). This rank was supported by new GBD data and remains constant in 2,030 projections (Mathers and Loncar, 2006). Because seasonal affective disorder (SAD) is a subtype of recurrent major depression and shares many symptoms (APA, 2000), it also presumably shares the same disability during symptomatic months (Freed et al., 2007). Despite these symptom-based and possibly functional commonalities, SAD prevalence
Estimating the Disease Burden of Seasonal Affective Disorder
90
(Magnusson, 2000), course (APA, 2000), etiology (Sohn and Lam, 2005), and treatment choice and efficacy (Westrin and Lam, 2007b) differ from that of nonseasonal major depression. No SAD disease burden estimates exist to date, and the disease burden estimates for major depression may not generalize to SAD. That said, extant SAD literature is arguably sufficient in scope and content to provide the information needed to calculate reasonable estimates of disease burden of SAD. We review four key components to estimating the disease burden of SAD: disorder prevalence to include demographic and environmental influences and measurement variance; disease burden measurement techniques; course of illness; and treatment efficacy.
2
What Is SAD?
SAD, as defined by DSM IV-TR (APA, 2000), is a recurrent pattern of Major Depressive Episodes with a temporal relationship between the episodes and time of year (e.g., onset during fall and winter with remission during spring and summer). SAD affects 10–20% of recurrent depression cases (Magnusson, 2000). Individuals diagnosed with SAD frequently endorse ‘‘atypical’’ symptoms of depression, including carbohydrate cravings, hypersomnia, fatigue, and weight gain. However, cognitive symptoms are often similar to nonseasonal depression (Sullivan and Payne, 2007), and may include feelings of worthlessness, excessive or inappropriate guilt, indecisiveness, diminished ability to concentrate, and/or thoughts of death. Left untreated, SAD episodes often recur annually (Sakamoto et al., 1995). See > Table 90-1 for the diagnostic criteria and associated features of SAD. Etiological models of SAD propose biological mechanisms to link light availability to SAD onset, based on the observation that SAD prevalence increases with latitude in the United States (Mersch et al., 1999). Given that photoperiod (i.e., day length) is completely determined by latitude and day of the year, light availability may be implicated in SAD etiology through a number of mechanisms: a) phase-delayed circadian rhythms with respect to objective time or to the timing of sleep, b) a dose of light (photons) to the retina that falls below a critical threshold, c) an abnormally prolonged duration of nocturnal melatonin release during winter, and d) reduced serotonergic activity (Magnusson and Boivin, 2003). The aforementioned hypothesized light availability-based models do not necessarily hold true for summer-type SAD (i.e., depressive episodes which begin in spring and remit in the fall and winter; Schwartz and Schwartz, 1993). DSM IV-TR (APA, 2000) does not distinguish between summer- and winter-type SAD. Summer SAD appears much less common than winter SAD (Magnusson, 2000) and includes the ‘‘typical’’ depressive symptoms common in nonseasonal depression such as weight loss, insomnia, agitation, and loss of appetite (Schwartz and Schwartz, 1993). Some persons diagnosed with summer SAD, especially in tropical climates, find relief in the dark or in air conditioned spaces (Schwartz and Schwartz, 1993). These observations suggest that temperature, and not necessarily light availability, mediates SAD symptoms. Because the DSM-IV-TR and ICD-10 classification systems do not distinguish between summer- and winter-type SAD, we report findings from both the summer and winter SAD literature; but unless explicitly named, SAD will refer to winter-type SAD in this chapter. Young (1999) and Young et al. (1991, 1997) were the first to suggest that there may be a ‘‘dual’’ vulnerability to SAD, consisting of a physiological vulnerability and a psychological
1551
1552
90
Estimating the Disease Burden of Seasonal Affective Disorder
. Table 90-1 Diagnostic criteria and associated features or SAD (APA, 2000) Must meet criteria for a major depressive episode as part of major depressive disorder, recurrent; bipolar I disorder; or bipolar II disorder In adults, a major depressive episode is characterized by these clinically significant symptoms a) Depressed mood or loss of interest in activities and b) Four or more of these symptoms 5% unintentional change in weight or significant change in appetite Daily and significant increased or decreased sleep Observable and significant restlessness or slowing Fatigue or loss of energy Daily and significant feelings of worthlessness or guilt, which is considered excessive or inappropriate Difficulty thinking, concentrating, or making decisions Recurrent thoughts of death or dying (not just fear of dying) with or without a plan for suicide or a suicide attempt Here, symptoms for a major depressive episode cannot be due to events such as bereavement or a medical condition SAD is a specifier, not a stand-alone disorder, and is characterized by the following Regular temporal relationship between onset of episode and time of year for at least 2 years (with no nonseasonal episodes in that timeframe) Full remission or change to mania/hypomania has a regular temporal relationship Seasonal episodes occur more frequently than nonseasonal episodes over a patient’s lifetime Persons diagnosed with SAD episodes are often report these ‘‘atypical’’ symptoms of depression. However, these symptoms are not required for a SAD diagnosis Prominent anergy Increased sleep Overeating Weight gain and Carbohydrate cravings This table summarizes the diagnostic criteria and associated features of SAD as reported in the DSM-IV-TR (APA, 2000)
vulnerability. In an expansion of the concept of a dual-vulnerability to SAD, Rohan et al. (2003, 2004a) specified content for the psychological component of SAD vulnerability; specifically, the psychological component to the etiology of SAD is hypothesized to involve activation of negative cognitions (e.g., maladaptive schemas, attitudes, and automatic thoughts) and behavioral disengagement from potential sources of positive reinforcement during the winter months. Rohan et al. further proposed that the psychological vulnerability interacts with the physiological vulnerability (consisting of the factors mentioned above) in a maintaining cycle. Rohan’s model is termed the integrative, cognitive-behavioral model. Rohan’s cognitive-behavioral factors might also interact with an underlying physiological vulnerability to explain onset of SAD symptoms during the debilitating summer heat in tropical climates.
Estimating the Disease Burden of Seasonal Affective Disorder
2.1
90
Prevalence
We discuss several factors affecting prevalence of SAD: latitude, age, gender, and measurement method. Although evidence suggests that one or more of these factors is related to prevalence of SAD, reported prevalence rates vary widely across geographic regions and have been estimated to range from 0.4 to 16% in the general adult population (Blazer et al., 1998; Booker et al., 1991; Magnusson, 2000). As previously discussed, length of photoperiod (i.e., the light/dark cycle) is thought to be related to winter SAD onset (Mersch et al., 1999; Young et al., 1997) through one or more proposed mechanisms. Prevalence may be higher, and severity may be greater at northern latitudes, where sunlight is reduced (Lurie et al., 2006). Rosen et al. (1990) reported a nearly linear relationship between latitude within the United States and prevalence of SAD. Specifically, they reported SAD prevalence of just over 1% in Florida compared to 9.7% in New Hampshire. Mersch et al. (1999) investigated the relationship between the prevalence of seasonal affective disorder and latitude and found the mean prevalence of SAD to be two times higher in North America compared to Europe. Moreover, a significant positive correlation was found between prevalence and latitude in North America. However, the correlation was low, calling into question the clinical utility of such information. Haggerty et al.’s (2001) meta-analysis of the literature also reported significant correlations between latitude and SAD, although similar to Mersch et al. (1999), the predictive power of such correlations was extremely low. Importantly, Magnusson has reminded us that ‘‘latitude is only an indirect measure of how much light people receive in winter’’ (Magnusson, 2000, pg. 177). Sakamoto et al. (1993) reported prevalence of SAD at different latitudes in Japan, but it was total hours of sunshine that appeared to be related to prevalence, not latitude. Still others report no relationship between SAD onset and hours of sunshine (Young et al., 1997). Because SAD is thought to be primarily caused by lack of light in winter (Magnusson, 2000), this factor and others (e.g., climate, genetics, and socio-cultural factors) may be expected to play a role in seasonal depression as researchers continue to delineate these relationships. In conclusion, a review of the literature reveals a small to moderate correlation between latitude and SAD; but these results should be considered with caution until a more systematic method of evaluating this relationship is established. Age has repeatedly been described as a factor related to prevalence of SAD. Rosenthal et al. (1984) reported that there may be a developmental component to SAD, such that children and adolescents may develop SAD and may respond to light treatment (Giedd et al., 1998; for a discussion of treatments, see below). As described in Magnusson’s (2000) review, Carskadon and Acebo (1993) estimated that 4.2% of children ages 9–12 met their criteria for SAD, with lower rates among the younger children. Using a self-report questionnaire, Sonis (1989) reported 6.9% of adolescents aged 14–18 years had high seasonal variation in depressive symptoms. Swedo et al. (1995) reported estimates of 1.7–5.5% of students from an American sample were identified as probable SAD cases. Rastad et al. (2006) reported that among Swedish students of similar ages (17–18 years), 20% self-reported depressed mood during the winter season. However, these data are not descriptive of diagnosed SAD but rather of selfreported seasonal variation in mood. Rohan and Sigmon (2000) reported that prevalence of SAD among college students in Maine, United States, was similar to that of an adult sample from nearby latitudes. They reported that over half of the participants reported a winter mood pattern, but only 5% met criteria for SAD, suggesting that studies only reporting seasonal patterns of mood should not
1553
1554
90
Estimating the Disease Burden of Seasonal Affective Disorder
be taken as a reliable estimate of true SAD cases. In fact, Soriano et al. (2007) reported no relationship between age and SAD among a sample of postgraduate students (medical students) in Romania with an average age of 23.2 years. Mersch et al. (1999) measured SAD in a large sample of the general population in The Netherlands and found that young women were more sensitive to seasonal variation than older women and that this sensitivity decreased with age for both men and women. Overall, it seems that younger people may be at equal, or nominally higher, risk for SAD than older individuals (Mangusson, 2000), however the problems plaguing prevalence rate estimations (discussed later) limit the generalizability of the current findings. Finally, women have generally been reported to represent a larger proportion of SAD cases than men (Chotai et al., 2004; Magnusson, 2000; Rastad et al., 2005), which may be a reflection of the gender difference in depressive disorders in general. Kasper (1989) found a 3.5:1 (women: men) sex ratio of seasonal mood patterns in the general population. However, others report no particular relationship between gender and SAD. For example, Lucht and Kasper (1999) reported no differences on the Seasonal Pattern Assessment Questionnaire (SPAQ) between men and women, but indicated that more women than men referred themselves to the outpatient clinic, highlighting the importance of considering bias when drawing conclusions. As with other mental health conditions, data on differences between men and women may be influenced by several factors, including response bias, symptom report differences, or other factors; and, therefore, readers are encouraged to consider such factors when drawing conclusions. Although current data seem to suggest that SAD may be less common among men overall, Goel et al. (2002) reported that men and women with SAD did not differ in the overall severity of the depression. Problematic for the studies reviewed here and other places (e.g., Magnusson, 2000) is the fact that many epidemiological reports rely on widely varying methodologies when reporting SAD prevalence. Many have reported prevalence of SAD assessed by the SPAQ (e.g., Soriano et al., 2007; Srisurapanont and Intaprasert, 1999; Tonetti et al., 2007). However, others have used a modified version of the SPAQ (e.g., Rastad et al. 2005), the Structured Interview Guide for the Hamilton Depression Rating Scale – Seasonal Affective Disorder version (SIGH-SAD; e.g., Goel et al., 2002; Williams et al., 1992), the Seasonal Health Questionnaire (SHQ; e.g., Thompson et al., 2004), or DSM criteria (Blazer et al., 1998) to name a few. Studies using self-report questionnaires to assess seasonality may or may not give accurate rates of seasonal depression meeting diagnostic criteria for SAD. Larger scale community samples using clinical, diagnostic interviews to determine DSM criteria have reported lower prevalence (e.g., Blazer et al., 1998), suggesting that the SPAQ may tend to overestimate true prevalence of the disorder. In fact, Thompson and Cowan (2001) suggest that the SPAQ should not be used to determine prevalence of SAD as it ‘‘gives misleadingly high estimates of prevalence.’’ Interestingly, the Blazer et al. (1998) study reported no relationship between latitude and prevalence of SAD when using DSM criteria, supporting the questions we have raised previously.
2.2
Treatment Options
Broadly, three treatment options exist for SAD: (1) light therapy (LT), considered the goldstandard; (2) pharmacotherapy; and (3) behavioral interventions. Rosenthal et al. (1984) were the first researchers to demonstrate that bright artificial light had an antidepressant
Estimating the Disease Burden of Seasonal Affective Disorder
90
effect. In a recent summary, Westrin and Lam (2007b) describe LT as the most widely used and preferred treatment for SAD. It includes daily exposure to 10,000 lux of cool-white or fullspectrum fluorescent lamps with ultraviolet rays filtered out, for at least 30-min during symptomatic months. The light sources must be positioned at the appropriate distance from the patient to ensure that the full 10,000 lux reaches the retina, as lux exponentially decreases as distance from light source increases. Indoor evening room light is 100 lux, and a brightly lit office is 500 lux. Outdoor light ranges from 4,000 lux (cloudy grey winter day) to 100,000 lux (bright sunny day). Side effects are considered mild and transient and can include headache, nausea, eyestrain, insomnia, agitation, and mania. Variants of LT exist, but we focus on the more traditional light box in this chapter. For example, dawn simulators (i.e., a light that gradually brightens as patients wake) and portable head-mounted light therapy devices are available. Golden et al. (2005) reviewed eight light therapy trials that met their selection criteria (i.e., inclusion of an adequate placebo control group and SAD diagnosis according to established criteria). The combined effect size was 0.84 with a 95% CI of 0.60–1.08. Six of the eight studies reported individual effect sizes where the lower end of the CI around the effect size was greater than 0. Golden et al. also reported an odds ratio for remission of 2.9 (95% = 1.6–5.4) at the end of a trial of light therapy from the four studies where the number of remitted subjects was known. In an older review, where study inclusion was less restrictive than Golden et al. (2005), Terman et al. (1989) concluded that LT is most likely to lead to remission if patients begin with mild SAD symptoms. Broken down by light dosing schedule, Terman et al. reported remission rates of 51%, 53%, 32%, and 38% in patients who received morning plus evening light, morning light alone, midday light alone, and evening light alone, respectively. Thus, up to 53% of individuals with SAD, and an even smaller percentage (43%) of individuals with moderate to severe SAD showed clinically significant improvement with LT (Terman et al. 1989). In contrast, he reported remission rates of 31% for brief exposure to light (a control condition) and 11% for dim light exposure (a control condition). Fewer trials have evaluated the efficacy of medication in the treatment of SAD. Lam et al. (2006) summarized the main findings of two key medication trials. To date, selective serotonin reuptake inhibitors (SSRIs) have demonstrated the best treatment efficacy. Relative to placebo, patients taking fluoxetine and sertraline (in two different studies) demonstrated improvement in symptom severity. However, improvement from baseline was not statistically significant in the fluoxetine trial, and percentages of score reduction (50% was considered a clinically significant response though not necessarily indicative of full remission) between the active treatment and placebo were strikingly similar: 59% (fluoxetine) versus 34% (placebo) and 63% (sertraline) versus 46% (placebo). Lam et al. state that these clinical response rates are similar to trials for nonseasonal depression. Westrin and Lam (2007a) summarized other findings that bupropion was helpful to prevent a SAD episode recurrence when initiated before symptoms began (16% bupropion versus 28% placebo); but the rate of recurrence overall was remarkably low, even in the placebo condition. Less rigorous examinations of other medications have been completed and suggest potentially promising acute treatments that require further research. Lam et al. (2006) state that both antidepressants and LT are considered first-line SAD treatments. In head-to-head comparisons of LT and medication, current findings are equivocal, with descriptive evidence suggesting superiority for LT. The findings from Lam et al., who compared fluoxetine and LT in longest and largest trial to date (8-weeks; N = 96), supported
1555
1556
90
Estimating the Disease Burden of Seasonal Affective Disorder
this conclusion. Specifically, the reported a 50% remission for LT compared to 54% for fluoxetine, and remission rates were similar in patients with more severe symptoms (48% for LT and 50% for fluoxetine). As an alternative to LT, Rohan et al. (2004b) developed a group cognitive-behavioral therapy (CBT) for SAD. Preliminary data suggest that Rohan’s SAD-tailored CBT may be comparably efficacious to the ‘‘best available’’ treatment (i.e., LT) in the acute treatment phase and substantially more effective than LT in symptom severity and relapse rates 1-year later (Rohan et al., 2004b). The initial study, a feasibility randomized clinical trial (N = 23) compared Rohan’s 6-week SAD-tailored group CBT, LT, and the combination of SAD-tailored group CBT + LT. Participants were evaluated at pre-treatment, post-treatment, and a 1-year follow-up. Although all treatment groups were associated with significant improvement in SAD symptoms from pre- to post-treatment on both measures, CBT demonstrated superior treatment durability at 1-year follow-up (i.e., the winter following treatment completion). No participant who received CBT (either alone or in conjunction with LT) experienced a full episode recurrence at the 1-year follow-up. However, 62.5% of participants who received LT alone endorsed severity of symptoms sufficient to qualify for a full-blown SAD episode recurrence. Moreover, at the 1-year follow-up, participants in the CBT alone and CBT + LT conditions reported significantly lower SAD symptoms. 1-year remission rates are as follows (variations reflect differences in remission criteria): (1) 43–57% for CBT (2) 25–38% for LT, 67–83% for CBT + LT. In a larger sample (N = 61), Rohan et al. (2007) generally replicated her follow-up findings from the pilot study (1-year data is not yet published). In the intent-to-treat sample, remission rates at post treatment were as follows (again, variations reflect differences in remission criteria): (1) 44–50% for LT, (2) 40–47% for CBT, (3) 53–73% for CBT + LT, and (4) 7–20% for a minimal contact delayed treatment control group.
2.3
Measuring Disease Burden
The true burden of a disorder may be defined as ‘‘the burden in the absence of treatment, that is calculated from the burden observed in the population under study plus the burden presently averted by current population coverage and mix of interventions’’ (Andrews et al., 2004, pg. 526). The burden of disease in a population in a given year is the sum of (1) years of life lost due to premature deaths (YLL) and (2) an estimate of the future years of life with disability (YLD) for new cases of disease or injury, weighted for severity (Vos et al., 2001). Disease burden allows disease pervasiveness (measured by incidence or prevalence), persistence (that is, length of time someone has the disease), and morbidity to be combined onto one scale. In the Global Burden of Disease study (Murray and Lopez, 1996), depression was ranked fourth in disease burden and expected to rank second by 2020. In ages 14–44, unipolar depression was ranked between first and third in disability, depending on sex and country development. Disease burden is the confluence of two measurements: morbidity and mortality. However, mortality is rarely attributed to mental disorders (Andrews et al., 2004), and persons with SAD experience less suicidal ideation than in nonseasonal depression (reported in Magnusson and Partonen, 2005). However, mortality is associated with depression (Kinder et al., 2008). Although patients with SAD rarely commit suicide, researchers have considered mortality associated with depression to be negligible (Melse et al., 2000; Stouthard et al., 2000). We focus on morbidity in this chapter.
Estimating the Disease Burden of Seasonal Affective Disorder
90
Morbidity of a disease is basically measured using one of two scales (both ranging from zero to one) and anchored by perfect health and death. The number between 0 and 1 is considered a preference weight for a particular disease or health state. To estimate the disease burden, the preference weight is multiplied by (1) the amount of time spent in that health state and (2) the disease prevalence. Preference weights differ from other measures of severity, in that preference-weighted metrics incorporate how a health condition is valued by a given population. For example, Kiberd and Lawson (2000) measured the preference of a Type-I diabetes mellitus (DM) with comorbid blindness and end stage renal disease (ESRD; requiring dialysis) health state from patients with DM and ESRD. In a separate study, Revicki and Wood (1998) measured the preference of a severe depression health state from patients being pharmacologically treated for depression. Preference was measured in units called health utilities (discussed below) by both sets of researchers. Health utilities (and other units of preference) are valued by having a group of raters read a brief vignette describing the health states (depression or DM with ESRD in this example). Raters are administered one or more of econometric instruments (e.g., the person trade-off; PTO, described below) designed to elicit preference. In these two examples, average preference for the severe depression health state was less desirable than was average preference for DM with comorbidities. Thus, a health state like DM with comorbidities, which requires frequent life-saving medical treatment, is more desirable than severe depression, which is theoretically treatable with medication and/or psychotherapy. Depression is life-threatening only if a patient chooses to attempt suicide, unlike DM with ESRD. This example demonstrates that non preference-based measures of health status and clinical symptom focused measures are predominantly clinician tools, not health policy tools per se (Bennett and Torrance, 1996; Berzon et al., 1996; Garza and Wyrwich, 2003; Tsevat, 2000). The GBD study (Murray and Lopez, 1996) assessed disease burden with the disability adjusted life year (DALY), which equals YLL plus YLD. Here, 0 represents perfect health and 1 represents death, such that a year in perfect health is worth 0 DALYs. DALYs are considered an interval scale, such that they can be summed across individuals and over time. ‘‘The DALY measures the gap between the actual health of a population and a hypothetical norm, namely a life expectancy of 82.5 years for women and 80 years for men. DALYs for a disease or health condition are calculated as the sum of the years of life lost due to premature mortality in the population and the years of life lost due to disability’’ (Michaud et al., 2001, pg. 535). DALYs also include life expectancy, and can adjust the weighting based on age, such that a young adult is ‘‘worth’’ more than someone at the beginning or end of the life cycle. A component of YLD is a value called a disability weight, where 0 equals perfect health and 1 equals death. In the GBD study and other studies which assessed DALYs (e.g., Kruijshaar et al., 2005; Sanderson and Andrews, 2001), disability weights for DALYs were obtained for specific diseases by administering a series of ‘‘> person tradeoff ’’ (PTO) vignettes to a sample of clinicians. Two variants were used. First, clinicians were given a choice between trading ‘‘quantity of life for healthy individuals and disabled individuals.’’ In the second, they are asked to tradeoff ‘‘quantity of life for healthy individuals versus improved quality of life for a group of disabled individuals’’ (Murray and Lopez, 1996, pg. 91). As an example, a disability weight would equal 0.5, if a respondent considered 2,000 people with depression to represent the same amount of health as 1,000 people without depression (Sanderson and Andrews, 2001). See > Figures 90-1 and > 90-2 for graphical representation of the PTO. A similar metric is called a quality adjusted life year (QALY). Here, a year in perfect health equals one QALY and death equals zero, the reverse of the aforementioned DALY. Like DALYs,
1557
1558
90
Estimating the Disease Burden of Seasonal Affective Disorder
. Figure 90-1 Person tradeoff method used in Murray and Lopez (1996) and Sanderson and Andrews (2001): ‘‘trading quantity of life for healthy and disabled individuals’’ (pg. 670). This is one variant of the PTO, a method designed to determine preference and ‘‘worth’’ for a particular disease state. Respondents indicate their preference by valuing quantity of life in a hypothetical sample of healthy and disabled individuals. A low disability weight (close to 0) indicates that the disease is considered less disabling, and conversely, a high disability weight (close to 1) indicates that the disease is considered more disabling
the number representing disease preference in QALYs is weighted by time spent in a given disease state and disease prevalence. Unlike DALYs, the number is not adjusted for age. The number is called a > health utility (versus a disability weight). We recognize that the nomenclature used to describe the preference weight needed to calculate a QALY differs according to the method. To minimize jargon, we use the term health utility in this chapter, independent of method. Of note, numerous methods can be used to obtain a health utility (e.g., interpolation from a generic health-related quality of life scale or direct valuation of a health state vignette from a community sample). And, as discussed by Donald Sherbourne et al. (2001), each method can produce a wide range of health utilities. Health utilities may or may not be sensitive enough to detect changes in symptom severity, and are therefore questionable indicators of treatment outcomes. Nevertheless, health utilities (and the eventual calculation of QALYs) are the recommended outcome in cost-effectiveness analyses Gold et al. (1996). The > standard gamble (> Figure 90-3) is the classic method to elicit health utilities. Here, respondents choose between (1) a health state with certainty or (2) a gamble, where treatment could lead to perfect health with some probability of death. The point of indifference is called the health utility. In this paradigm, the health state can be a vignette describing symptoms and
Estimating the Disease Burden of Seasonal Affective Disorder
90
. Figure 90-2 Second variant of the PTO used in Murray and Lopez (1996) and Sanderson and Andrews (2001): ‘‘trading quantity of life for healthy individuals versus improved quality of life for disabled individuals’’ (pg. 670). This is another variant of the PTO, a method designed to determine preference and ‘‘worth’’ for a particular disease state. Respondents indicate their preference by valuing quantity and quality of life in a hypothetical sample of healthy and disabled individuals. A low disability weight (close to 0) indicates that the disease is considered less disabling, and conversely, a high disability weight (close to 1) indicates that the disease is considered more disabling
disability associated with a specific disease (e.g., depression; Revicki and Wood, 1998). Respondents are then administered the standard gamble multiple times, where the researchers vary the value of p, until the respondent is indifferent between Choices 1 and 2. The procedure is repeated for multiple health states within a specific disease (e.g., severe depression, moderate depression, mild depression). Freed et al. (2007) successfully assessed the only published study that reported health utilities in the SAD literature. Both the disability weight in DALYs and the health utility in QALYs are preference-based. Preference-based measures differ form other symptom severity or health-related quality of life measures. Where as non preference-based measures describe functioning, preference-based measures include a value component, which assess the relative worth of diseases or health states. Gold et al. (2002) and Sassi (2006) provide detailed discussions of the similarities and differences between QALYs and DALYs. In > Tables 90-2 and > 90-3, we report the disability weights and health utilities associated with non seasonal depression. Preference-based assessments of SAD do not yet exist in the literature. As such, preferences for non seasonal depression can serve as a proxy for SAD, as was argued and demonstrated in Freed et al. (2007). Here, the researchers estimated QALYs of SAD patients who received one of
1559
1560
90
Estimating the Disease Burden of Seasonal Affective Disorder
. Figure 90-3 The standard gamble paradigm for calculating a health utility for a particular health state (figure adapted from Freed et al., 2006). An example of the standard gamble, a method designed to determine preference and ‘‘worth’’ for a particular disease state. Respondents value a disease by indicating their willingness to accept a risky but disease-curing intervention versus living with the disease. A low health utility (close to 0) indicates that the disease is considered very undesirable, and conversely, a high health utility (close to 1) indicates that the disease is not considered undesirable
three treatments: light therapy, group cognitive-behavioral therapy, a combination of light and group cognitive-behavioral therapies. The researchers coarsely accounted for seasonality by assuming all patients would spontaneously remit in the spring and summer months. Because of this recurring seasonal remission, an episode of SAD would be less burdensome than an episode of nonseasonal depression, if symptom severity during the episode was equal.
2.4
Course of Illness
As reported in Magnusson and Partonen (2005) and Tonetti et al. (2007), the average age of SAD onset is between 20 and 30 years old, and the average patient age is 40. Schwartz et al. (1996) reported age of onset between 35.9 and 38.6 years (SD ranges 8.0–9.9) and Sakamoto et al. (1995) reported 31.5 years (SD = 13.3). SAD onset has been reported in childhood and adolescence. Tonetti et al. (2007) surveyed age groups in Italy: 10–17 years old (n = 1,709) and 18–25 years (n = 1,867). As reported above in this chapter, the younger group had a lower prevalence (9.52% females and 7.95% males) than the older sample (12.74% females and 8.93% males; results of significance tests between the older and younger groups were not reported).
Estimating the Disease Burden of Seasonal Affective Disorder
90
. Table 90-2 Existing disability weights for nonseasonal depression Source
Method
Depression disability weights
Kruijshaar et al. (2005)
Extrapolation from a preferenceweighted self-report health-related quality of life measure
Mild: 0.19 (95% CI = 0.16–0.22) Moderate to severe: 0.51 (95% CI = 0.46–0.55) Severe w/psychotic features: 0.84 (95% CI = 0.80–0.88)
Murray and Lopez (1996)
Person trade-off
Overall: 0.6 Treated: 0.3
Sanderson and Andrews (2001)
Person trade-off
Mild: 0.09 (SD = 0.07) Moderate: 0.34 (SD = 0.21) Severe: 0.70 (SD = 0.26) Overall: 0.42
Sanderson and Andrews (2001)
Rating scale
Mild: 0.19, SD = (0.11) Moderate: 0.32, SD = (0.15) Severe: 0.61, SD = (0.17)
Schwarzinger et al. (2003)
Visual analog scale
Severe: 0.78 (SD = 0.17)
Schwarzinger et al. (2003)
Person trade-off
Severe: 0.34 (SD = 0.40)
Schwarzinger et al. (2003)
Time trade-off
Severe: 0.67 (SD = 0.26)
Stouthard et al. (2000)
Person trade-off
Mild: 0.14 (95% CI = 0.09–0.19) Moderate: 0.36 (95% CI = 0.27–0.42) Severe: 0.76 (95% CI = 0.56–0.97) Severe w/psychotic features: 0.83 (95% CI = 0.75–0.92) Overall males: 0.41 Overall females: 0.37
Because no disability weights for SAD exist, we report the disability weights for nonseasonal depression. Different measures exist for determining disability weights. There is a relationship between depression severity and the disability weighting, such that more severe depression is considered more disabling
The 10-year differential between SAD onset and average patient age suggests that persons with SAD either are misdiagnosed or cope with SAD symptoms for many years before seeking treatment. We argue that the former is more likely because of the high use of health care services by undiagnosed persons with SAD. Eagles et al. (2002) screened primary care patients for SAD and examined health care usage over a 5-year period. They identified 145 previously diagnosed SAD suffers, matched them to 246 non-SAD controls, and compared the two groups along a variety of health services variables. On average (median) SAD patients had 43% more primary care consultations, 50% more tests/investigations, 56% more prescriptions, and 200% more specialist referrals than the non-SAD controls over the 5-year period. Moreover, SAD patients presented with a greater diversity of both psychiatric and nonpsychiatric symptoms. The authors suggest that SAD possibly goes undiagnosed and untreated because patients present with other symptoms, many of which are further assessed and treated with limited success. Magnusson and Partonen (2005) summarize the very limited long-term and aggregated follow-up data, which suggests that 22–42% of persons with SAD still experience symptoms
1561
1562
90
Estimating the Disease Burden of Seasonal Affective Disorder
. Table 90-3 Existing health utilities for nonseasonal depression Source Hatziandreu et al. (1994)
Method Direct General practitioners and psychiatrists administered the standard gamble
Kamlet et al. Indirect Determined utilities from a (1995) previously preference-weighted selfreport health-related quality of life measure administered to research participants
Depression health utilities Treatment with SSRI: 0.72 Treatment with TCA: 0.69 Remission, with SSRI maintenance: 0.93 Remission, off drug therapy: 0.95 Overall: 0.45 Worst Case: 0.3 Best Case = 0.7
Revicki et al. Direct valuation from patients using the Overall w/pharmacotherapy: 0.67–0.82 (1997) standard gamble Overall no treatment: 0.31 Revicki and Direct valuation from patients using the Mild (with pharmacotherapy): 0.64–0.73, Wood (1998) standard gamble (SD = 0.20–0.21) Moderate: (with pharmacotherapy): 0.55–0.63 (SD = 0.03– 0.23) Severe (no medication): 0.30 (SD = 0.28) Sapin et al. (2004)
Indirect Determined utilities from a previously preference-weighted selfreport health-related quality of life measure administered to research participants
Slight to Moderate: 0.45 (SD = 0.22) Marked: 0.33 (SD = 0.24) Serious: 0.15 (SD = 0.21)
Sobocki et al. (2007)
Indirect Determined utilities from a previously preference-weighted selfreport health-related quality of life measure administered to research participants
Mild: 0.60 (95% CI = 0.54–0.65) Moderate: 0.46 (95% CI = 0.30–0.48) Severe: 0.27 (95% CI = 0.21–0.34)
Donald Sherbourne et al. (2001)
Various Methods Indirect from Symptoms (no diagnosis): 0.56–0.94 preference-weighted health-related Lifetime: 0.47–0.92 Current: 0.44–0.90 quality of life measures; direct valuation Double depression: 0.38–0.88 with time-tradeoff and standard gamble
Because no health utilities for SAD exist, we report the health utilities for nonseasonal depression. Different measures exist for determining health utilities. There is a relationship between depression severity and the health utility, such that more severe depression is considered more disabling. SSRI selective serotonin reuptake inhibitor; TCA tricyclic antidepressant
5–11 years later. In that same timeframe, 33–44% of persons with SAD developed a non-seasonal pattern in subsequent episodes; approximately 6% of persons experienced sub-syndromal SAD, and SAD resolved in only 14–18% of patients. The authors state that the data may be artificial, and not reflect the natural course of SAD because the data were obtained from patients who sought treatment for SAD. Course of SAD from several published studies is presented in > Table 90-4. These studies report that SAD is often characterized by chronic SAD and often complicated with other affective symptoms, even with treatment.
26
NDA
41
59
93
Leonhardt et al. (1994)
Magnussson and Partonen (2005)
Sakamoto et al. (1995)
Schwartz et al. (1996)
Thompson et al. (1995)
18
14
NDA
14–18
27
Remission (%)
37
44
42
33–44
38
38
42
22
22–42
31
Remain with pure SAD (%)
11
NDA
NDA
6
NDA
Sub-clinical SAD (%)
NDA
42% (LT) 58% (Antidepressants)
NDA
NDA
54% (LT) 19% (Antidepressants)
Reported treatment usea
5–8
8.8 (SD = 1.3)
10.4 (SD = 4.8)
5–11
4
Mean follow-up period (years)
Interview
Case record and interview
Case record and interview
Literature reviewb
Interview
Data collection method
This table reports on the course of SAD in patients who participated in long-term follow-up studies. Over many years, SAD can remit, become more complicated (e.g., significant depression symptoms or episodes occur during what were asymptomatic months), not develop into complicated SAD, or improve (but not to remission). NDA no data available a Not mutually exclusive b Review article summarized existing literature
N
Study
Develop complicated SAD (%)
. Table 90-4 Course of SAD reported in long-term follow-up studies Estimating the Disease Burden of Seasonal Affective Disorder
90 1563
1564
90 2.5
Estimating the Disease Burden of Seasonal Affective Disorder
Other Considerations
Clearly, the foundation of estimating the disease burden of SAD rests on an accurate and consistent diagnosis. However, method variance exists in three arenas: diagnostic criteria, measures used, and the inclusion of co-occurring disorders. SAD is accurately diagnosed when all criteria are met according to DSM-IV-TR (APA, 2000). To prospectively assess SAD, researchers rely on self-report screening and/or structured diagnostic interviews. A structured interview is preferred, but this type of assessment is time consuming and may not be practical for larger epidemiological studies. When assessing the efficacy of SAD treatment from a clinical trial, interviews are frequently used, the most common of which is SIGH-SAD (Williams et al., 1992). The most widely used screening tool is the SPAQ (Rosenthal et al., 1987). As discussed in Magnusson (2001) and Lurie et al. (2006), the SPAQ overestimates prevalence of SAD. For example, of the 274 identified as SAD positive on the SPAQ, only 127 (46%) met DSM-IV criteria for SAD, when interviewed (Eagles et al., 2002). SAD screening instruments are not recommended for routine use in clinical practice because they are not sensitive enough (Lurie et al, 2006). Co-occurring conditions (i.e., comorbidities), common in clinical settings, can compromise the identification and treatment of SAD (Magnusson and Partonen, 2005). Common cooccurring conditions are similar to those of nonseasonal depression and include bulimia nervosa, panic disorder, posttraumatic stress disorder, attention deficit hyperactivity disorder, generalized anxiety disorder, other depressive disorders (e.g., bipolar II and premenstrual depressive disorder) adjustment disorder, alcohol use disorders, and Axis II disorders, as reviewed in Lurie et al. (2006), Magnusson (2000), Magnusson and Partonen (2005), and Westrin and Lam (2007b). Not surprisingly, co-morbidity clouds estimates of disease burden, especially when disability weights and health utilities are calculated from disease specific vignettes that do not include the co-occurring condition. Here, one must ask what the impact is of the co-occurring condition on the preference weight? Is the impact additive, for example, such that the preference weight of one disorder is added to the weight of a second disorder for a person’s comorbidity? A solution is to create vignettes based on generic health status, value those generic vignettes, and then assess patients with a generic measure that includes domains in the vignettes. But researchers then must partial out what disorder is driving the decrements in health status, an arduous task when there is a paucity of literature which assesses the quality of life in patients with SAD (Michalak et al., 2005; Michalak et al., 2007). Also, researchers may not have the resources to assess co-occurring conditions in large epidemiological studies of SAD. Comorbidity is a problem in both DALY and QALY calculations, but it has not been thoroughly addressed in disability weighing. Melse et al. (2000) briefly summarize the issue, and state that summing the weights of comorbid conditions is an acceptable solution because the inclusion of comorbidity insignificantly changes disease burden calculations. Using this summing approach, however, permits a person to have a disability weight of greater than 1. Thus, if that person is diagnosed with multiple conditions creating a disability weight greater than 1, then theoretically, his disability is worse than death. Vos et al. (2001) split the number or prevalent cases of co-occurring disorders equally across their sample, thereby allowing for the fact that people with multiple diagnoses would be likely to have more severe disease weightings than those with only one diagnosis.
Estimating the Disease Burden of Seasonal Affective Disorder
2.6
90
Bringing It All Together
To date, Freed et al. (2007) is the only published article that attempted to synthesize current literature and estimate the QALYs gained from SAD treatment. Their research combined health utilities derived from patients with nonseasonal depression (Revicki and Wood, 1998), and participant responses from a small pilot study (Rohan et al., 2004b). Freed et al. (2007) did not include any estimates of disease prevalence, and the participants were heavily screened before they were enrolled in the study (e.g., participants with another Axis I disorder were excluded). Their time horizon was 1-year, and they rationally but arbitrarily and simplistically decided when SAD symptoms would remit for the spring and summer months. Although arguably sufficient enough in scope to incorporate treatment costs and publish a paper on the cost-effectiveness of SAD treatments, Freed et al.’s article is not sufficient enough to estimate the disease burden of SAD. The primary method to estimate the disease burden of SAD is to calculate the years lived with disability (YLD). Here, > incidence of disease is multiplied by the average time spent in the disease state, and the product is then weighted by the disability weight (Murray and Lopez, 1996). The incidence of SAD is not widely studied, but the prevalence can be also be used instead of incidence. Melse et al. (2000) multiplied disease prevalence by disability weight. If disability weights changed over time (e.g., disease worsens with age), then researchers can average the weights and use that average. The choice to use prevalence- or incidence-based estimates is often dependent on the goal of the study (Melse et al., 2000). Incidence-based calculations are more appropriate for disease prevention programs while prevalence-based calculations are present-centered. Other choices in disease burden estimates include what discounting rate to use and whether to use age weightings. For example, ‘‘discounting is applied because people value their current health more than future health, and age-weighting is applied because most societies would choose to save young to middle-aged adults over the young or very old’’ (Andrews et al., 1998, pg. 123). Ultimately, disease burden estimates are just that, estimates. The estimates are used for comparison purposes to better assist policy and decision makers to reduce that burden and prioritize valuable healthcare resources to better treat and prevent illness. SAD is a chronic, recurring, but treatable subtype of major depression affecting people worldwide. Although no disease burden estimates exist to date, the breadth of current literature can provide researchers with enough information to reasonably estimate the disease burden of SAD worldwide, and across strata of age, gender, and latitude.
Summary Points SAD is a subtype of major depression that recurs annually and may be related to available light, negative cognitions, or the combination.
SAD prevalence estimates vary from 0.4 to 16% depending on latitude, age, gender, and measurement method.
SAD is a chronic and disabling condition that includes symptoms such as fatigue, weight gain, feelings of guilt, and loss of concentration with average age of onset 20–30 years.
1565
1566
90
Estimating the Disease Burden of Seasonal Affective Disorder
Treatment for SAD, such as light therapy, cognitive-behavioral therapy, medication, or
some combination thereof is efficacious to varying degrees, and under optimal conditions, can lead to remission in 83% of patients. Despite the chronicity, worldwide prevalence, and disability, no disease burden estimates exist for SAD to date. However, extant literature is sufficient in scope and practice to reasonably estimate the disease burden of SAD, borrowing from the major depression literature, when necessary. Disease burden is the confluence of two measurements, morbidity and mortality, although mortality is not typically included when assessing the disease burden of mental health conditions. Unlike symptom severity or generic quality of life, morbidity (measured by a preferenceweighted metric) accounts for the value or worth that society places on a particular disease state, relative to other disease states and anchored by perfect health and death. Once calculated, disease burden estimates for SAD can assist in medical decision making and in determining priorities for interventions. The present chapter reviews the key components necessary to estimate the disease burden of SAD: prevalence, course of illness, morbidity, and treatment efficacy.
References American Psychiatric Association. (2000). Diagnostic and Statistical Manual of Mental Disorders, 4th ed, Text Revision. > American Psychiatric Association, Washington, DC. Andrews G, Issakidis C, Sanderson K, Corry J, Lapsley H. (2004). Br J Psychiatry. 184: 526–533. Andrews G, Sanderson K, Beard J. (1998). Br J Psychiatry. 173: 123–131. Bennett KJ, Torrance GW. (1996) Measuring health state preferences and utilities: rating scale, time trade-off, and standard gamble techniques. In: Spiker B (ed.) Quality of Life and Pharacoeconomics in Clinical Trials, 2nd ed. Lippincott-Raven Publishers, Philadelphia, PA. Berzon RA, Mauskoph JA, Simeon GP. (1996). Choosing a health profile (descriptive) and/or a patientpreference (utility) measure for a clinical trial. In: Spiker B (ed.) Quality of Life and Pharacoeconomics in Clinical Trials, 2nd ed. Lippincott-Raven Publishers, Philadelphia, PA. Blazer DG, Kessler RC, Swartz MS. (1998). Br J Psychiatry. 172: 164–167. Booker JM, Hellekson CJ, Putilov AA, Danilenko KV. (1991). Arctic Med Res. Suppl: 281–284. Carskadon MA, Acebo C. (1993). J Am Acad Child Adolesc Psychiatry. 32: 264–269. Chotai J, Smedh K, Johansson C, Nilsson LG, Adolfsson R. (2004). Nord J Psychiatry. 58(6): 429–437.
Donald Sherbourne C, Unu¨tzer J, Schoenbaum M, Duan N, Lenert LA, Sturm R, Wells KB. (2001). Med Care. 39(11): 1246–1259. Eagles JM, Howie FL, Cameron IM, Wileman SM, Andrew JE, Robertson C, Naji SA. (2002). Br J Psychiatry. 180: 449–454. Freed MC, Rohan KJ, Yates BT. (2007). J Affect Disord. 100(1–3): 83–89. Garza AG, Wyrwich KW. (2003). Acad Emerg Med. 10 (4): 360–363. Giedd JN, Swedo SE, Lowe CH, Rosenthal NE. (1998). J Am Acad Child Adolesc Psychiatry. 37(2): 218–220. Goel N, Terman M, Terman JS. (2002). Depress Anxiety. 15(1): 34–41. Gold MR, Siegel JE, Russell, LB, Weinstein MC. (1996). Cost-Effectiveness in Health and Medicine. Oxford University Press, New York. Gold MR, Stevenson D, Fryback DG. (2002). Annu Rev Public Health. 23: 115–134. Golden RN, Gaynes BN, Ekstrom RD, Hamer RM, Jacobsen FM, Suppes T, Wisner KL, Nemeroff CB. (2005). Am J Psychiatry. 162: 656–662. Haggerty J, Hons B, Cernovsky Z, Husni M. (2001). J Nerv Ment Dis. 189(7): 482–484. Hatziandreu EJ, Brown RE, Revicki DA, Turner R, Martindale J, Levine S, Siegel JE. (1994). PharmacoEcon. 5(3): 249–268
Estimating the Disease Burden of Seasonal Affective Disorder Kamlet MS, Paul N, Greenhouse J, Kupfer D, Frank E, Wade M. (1995). Control Clin Trials. 16: 17–40. Kasper S, Wehr TA, Bartko JJ, Gaist PA, Rosenthal NE. (1989). Arch Gen Psychiatry. 46(9): 823–833. Kiberd BA, Lawson T. (2000). Transplantation. 70(7): 1121–1127. Kinder LS, Bradley KA, Katon WJ, Ludman E, McDonell MB, Bryson CL. (2008). Psychosom Med. 70(1): 20–26. Kruijshaar ME, Hoeymans N, Spijker J, Stouthard MEA, Essink-Bot ML. (2005). Bull World Health Organ. 83(6): 443–448. Lam RW, Levitt AJ, Levitan RD, Enns MW, Morehouse R, Michalak EE, Tam EM. (2006). Am J Psychiatry. 163(5): 805–812. Leonhardt G, Wirz-Justice A, Kraeuchi K, Graw P, Wunder D, Haug H-J. (1994). Compr Psychiatry. 35: 457–464. Lucht MJ, Kasper S. (1999). Women Ment Health. 2(2): 83–89. Lurie S, Gawinski B, Pierce D, Rouseeau SJ. (2006). Am Fam Physician. 74: 1521–1524. Magnusson A. (2000). Acta Psychiatr Scand. 101(3): 176–184. Magnusson A, Boivin, D. (2003). Chronobiol Int. 20: 189–207. Magnusson A, Partonen T. (2005). CNS Spectr. 10: 625–634. Mathers CD, Loncar D. (2006). PLoS Med. 3(11): e442. Melse JM, Essink-Bot ML, Kramers PG, Hoeymans N. (2000). Am J Public Health. 90(8): 1241–1247. Mersch PPA, Middendorp HM, Bouhuys AL, Beersma DGM, Van Den Hoofdakker FH. (1999). J Affective Disord. 53: 35–48. Michalak EE, Murray J, Levitt AJ, Levitan RD, Enns MW, Morehouse R, Tam EM, Cheung A, Lam RW. (2007). Psychol Med. 37: 727–736. Michalak EE, Tam EM, Manjunath, Levitt AJ, Levitan RD, Lam RW. (2005). Can J Psychiatry. 50(5): 292–295. Michaud CM, Murray C, Bloom B. (2001). JAMA. 285: 535–539. Murray CJL, Lopez AD. (1996). The global burden of disease. Harvard University Press Cambridge, MA. Rastad C, Sjoden PO, Ulfberg J. (2005). Psychiatry Clin Neurosci. 59: 666–675. Rastad C, Ulfberg J, Sjo¨de´n PO. (2006). J Am Acad Child Adolesc Psychiatry. 45(2): 231–238. Revicki DA, Brown RE, Keller MB, Gonzales J, Culpepper L, Hales RE. (1997). J Clin Psychiatry. 58(2): 47–58. Revicki DA, Wood M. (1998). J Affect Disorders. 48: 25–36. Rohan KJ, Roecklein KA, Tierney Lindsey K, Johnson LG, Lippy RD, Lacy TJ, Barton FB. (2007). J Consult Clin Psychol. 75(3): 489–500.
90
Rohan KJ, Sigmon ST. (2000). J Affect Disorders. 59: 85–96. Rohan KJ, Sigmon ST, Dorhofer DM. (2003). J Consult Clin Psychol. 71: 22–30. Rohan KJ, Sigmon ST, Dorhofer DM, Boulard NE. (2004a). Cognit Ther Res. 28(1): 39–56. Rohan KJ, Tierney Lindsey K, Roecklein KA, Lacy TJ. (2004b). J Affect Disorders. 80: 273–283. Rosen LN, Targum SD, Terman M, Bryant MJ. (1990). Psychiat Res. 31(2): 131–144. Rosenthal NE, Genhart M, Sack FM, Skewerer RG, Wehr TA. (1987). In Hudson and Pope (eds) The Psychobiology of Bulimia. American psychiatric Press, Washington, DC. Rosenthal NE, Sack DA, Gillin JC, Lewy AJ, Goodwin FK, Davenport Y, Mueller PS, Newsome DA, Wehr TA. (1984). Arch Gen Psychiatry. 41:72–80. Sakamoto K, Nakadaira S, Kamo K, Kamo T, Takahashi K. (1995). Am J Psychiatry. 152(6): 862–868. Sakamoto K, Kamo T, Nakadaira S, Tamura A, Takahashi K. (1993). Acta Psychiatr Scand. 87: 258–265. Sanderson K, Andrews G. (2001). Aust N Z J Psychiatry. 35: 668–676. Sapin C, Fantino B, Nowicki ML, Kind P. (2004). Health Qual Life Outcomes. 2: 20. Sassi F. (2006). Health Policy Plan. 21(5): 402–408. Schwartz OJ, Brown C, Wehr TA, Rosenthal NE. (1996). Am J Psychiatry. 153: 1028–1036. Schwartz A, Schwartz RM. (1993). Depression Theories and Treatments: Psychological, Biological and Social Perspectives. Columbia University Press, New York. Schwarzinger M, Stouthard ME, Burstro¨m K, Nord E. (2003). Popul Health Metr. 1(1): 9. Sobocki P, Ekman M, Agren H, Krakau I, Runeson B, Ma˚rtensson B, Jonsson B. (2007). Value Health. 10 (2): 153–160. Sohn C-H, Lam RW. (2005). CNS Spectr. 10: 635–646. Sonis WA. (1989). Rosenthal NE, Blehar MC (eds.) Seasonal affective disorders and phototherapy. The Guilford Press, New York/London. Soriano JJ, Ciupagea C, Rohan KJ, Neculai DB, Yousufi SM, Guzman A, Postolache TT. (2007). Scientific World Journal. 7: 870–879. Srisurapanont M, Intaprasert S. (1999). J Affect Dis. 54: 97–99. Stouthard MEA, Essink-Bot M-L, Bonsel GJ. (2000). Eur J Pub Health. 10: 24–30. Sullivan B, Payne TW. (2007). Am J Psychiatry. 164(11): 1663–1667. Swedo SE, Pleeter JD, Richter DM, Hoffman CL, Allen AJ, Hamberger SD, Turner EH, Yamada EM, Rosenthal NE. (1995). Am J Psychiatry. 152: 1016–1019.
1567
1568
90
Estimating the Disease Burden of Seasonal Affective Disorder
Terman M, Terman JS, Quitkin FM, McGrath PJ, Stewart JW, Rafferty B. (1989). Neuropsychopharmacology 2: 1–22. Thompson C, Cowan A. (2001). J Affect Disorders. 64 (1): 89–98. Thompson C, Raheja SJ, King E. (1995). Br J Psychiatry. 167: 380–384. Thompson C, Thompson S, Smith R. (2004). J Affect Disord. 78: 219–226. Tonetti L, Barbato G, Fabbri M, Adan A, Natale V. (2007). J Affect Disorders. 97(1–3): 155–160. Tsevat J. (2000). Med Care. 38(9) Suppl II:II160–II164. Ustu¨n TB, Ayuso-Mateos JL, Chatterji S, Mathers C, Murray CJ. (2004). Br J Psychiatry. 184: 386–392. Vos T, Mathers C, Herrman H, Harvey C, Gureje O, Bui D, Watson N, Begg S. (2001). Soc Psychiatry Psychiatr Epidemiol. 36(2): 53–62.
Westrin A, Lam RW. (2007a). CNS Drugs. 21(11): 901–909. Westrin A, Lam RW. (2007b). Ann Clin Psychiatry. 19(4): 239–246. Williams JBW, Link MJ, Rosenthal NE, Amira L, Terman M. (1992). Structured Interview Guide for the Hamilton Depression Rating Scale—Seasonal Affective Disorder Version (SIGH-SAD). New York State Psychiatric Institute, New York. Young MA. (1999). Biol Rhythms Bull. 1: 4–6. Young MA, Meaden PM, Fogg LF, Cherin EA, Eastman CI. (1997). J Abnorm Psychol. 106(4): 554–562. Young MA, Watel LG, Lahmeyer HW, Eastman CI. (1991). J Affect Disord. 22(4): 191–197.
91 The Disease Burden of Suicide: Conceptual Challenges and Measurement Standards C. A. Claassen . R. M. Bossarte . S. M. Stewart . E. Guzman . P. S. F. Yip 1 Conceptual Challenges and Measurement Standards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1570 2 Population-Based Health Summary Measures: The Quality-Adjusted Life Year (QALY) and the Disability-Adjusted Life Year (DALY) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1571 3 Is the Burden Associated with Suicidal States Acute Only, Episodic, Progressive, Injury-Related, or Disease-Based? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1577 4 Methods and Measurement Standards for Calculating the Burden of Suicidal Acts via the DALY: Step One – Years of Life Lost to Suicide . . . . . . . . . . . . . . . . . . . . . . 1580 5 Step Two – Estimating the Years of Life Lived with Disability Due to Nonfatal Suicidal Behaviors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1582 6 The Importance of Suicide Burden Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1587 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1588
#
Springer Science+Business Media LLC 2010 (USA)
1570
91
The Disease Burden of Suicide: Conceptual Challenges and Measurement Standards
Abstract: Suicide was among the first public health issues to be studied at a population level, and the field has a research tradition that now spans 110 years. Yet, despite 8 decades of research documenting their inherent error, suicide rates remain the most widely used population-based suicide metric today. Health expectancy and > health gap measures have begun to replace disease rates in population health studies, but a number of conceptual and methodological challenges still need to be addressed when adapting these new tools to suicide research. A prerequisite to the use of any suicide measure is the need to define clearly what is being measured. The International Classification of Disease Manuals categorize suicidal acts as one type of injury event, suggesting that, like other injuries, these events are most often singular, largely unpredictable events. However, suicidal behavior is often preceded by functional decline and often follows a chronic, recurrent course – bearing more resemblance to a disease process than to an injury event. Next, both case definitions and procedures to correct for the error inherent in suicide data vary widely, highlighting the need to establish measurement standards. Because any burden study is only as valid as the disease concepts and measurement approaches upon which it is based, refinement of methodology will be necessary in order to achieve maximal benefit when applying the newer population summary statistics to suicide. List of Abbreviations: CDC, Centers for Disease Control and Prevention; COPD, Chronic obstructive pulmonary disease; DALY, Disability-Adjusted Life Years; DHHS, Department of Health and Human Services; DISMOD, Disease Modeling software program; E-codes, External Cause of Injury Codes from the ICD Manuals (see below); ED, Emergency Department; GBD, Global Burden of Disease; GDP, Gross Domestic Product; GBD1990, Global Burden of Disease Study for 1990; GBD2000, Global Burden of Disease Study for 2000; GBD2002, Global Burden of Disease Study for 2002; GBD2005, Global Burden of Disease Study for 2005; HIV/AIDS, Human Immunodeficiency Virus/Acquired Immune Deficiency Syndrome; ICD, International Classification of Disease Manual; ICD-9, International Classification of Disease Manual, 9th Version; ICD-10, International Classification of Disease Manual, 10th Version; ICE, International Collaborative Effort for Injury Statistics; MMWR, Morbidity and Mortality Weekly Review; NEISS-AIP, National Electronic Injury Surveillance System – All Injury Program; NHAMCS, National Hospital Ambulatory Medical Care Survey; QALY, QualityAdjusted Life Years; US, United States; WHO, World Health Organization; YLD, Years of Life Lived with Disability; YLL, Years of Life Lost
1 "
Conceptual Challenges and Measurement Standards ‘‘It is both extraordinary and unfortunate that, at the end of the twentieth century, the international public health community does not routinely quantify or project the health problems of populations. There are no standardized compilations of information on the extent of morbidity, disability and death in different populations of the world. Information at a global or regional level on behaviors and exposures that are important risk factors for death and disability is also extremely limited. Although the demographic community routinely publishes projections of fertility and population, future trends have been projected for only a very limited number of causes of death. . ..’’ –Murray & Lopez, 1996
The development of population-level health descriptors has not kept pace with the need for such measures, and suicide measures are no exception. Suicide was among the first health problems to be studied at a population level (Durkheim, 1951; 1897), and the field has a
The Disease Burden of Suicide: Conceptual Challenges and Measurement Standards
91
research tradition that now spans 110 years. A century ago, the first population-based studies compared national suicide rates, but – despite eight decades of research documenting their inherent error (Zilborg, 1935; Mohler and Earls, 2001) – suicide rates remain the most widely used population-based suicide measure in use today. National suicide rates are not designed to monitor shifts in suicidal behavior found in particular subgroups (e.g., adolescents, females, or ethnic minorities – Gunnell and Middleton, 2003; Yip et al., 2005), nor do they incorporate information on nonfatal suicidal behavior – despite the significant public health burden represented by this class of self harming behaviors (Corso et al., 2007). A more meaningful indicator might include a description of additional types of suicide phenomena, rather than just report the number of suicides occurring annually within a nation. In lieu of disease rates, two important population > health summary measures are being used with increasing frequency in health impact studies. Health expectancy measurement tools monitor progress toward improved health status, while health gap measurement tools estimate and compare the burdens imposed by > disease states and injury conditions. (Murray and Lopez, 1996) In light of the World Health Organization’s (WHO) 1948 definition of health as a ‘‘state of complete physical, mental and social well-being, and not merely the absence of disease or infirmity,’’ the emergence of these new measures signals movement toward adoption of a global public health agenda that addresses not only the length, but also the quality of life. The application of these new approaches to population-level suicide phenomena poses both conceptual dilemmas and methodological challenges. A prerequisite to modeling suicide summary indicators is the need to define clearly what is to be measured – are suicidal acts really best understood as injury events or as part of a > suicidal process? Next, measurement approaches for suicide-related phenomena across studies varies widely, both in terms of case definitions and of the methods used to estimate and correct for error. Given the substantial issues associated with the measurement of suicide burden, this chapter first describes currently published applications of health expectancy and gap summary measures to suicide phenomena. The implications of defining suicide as an injury event rather than a disease process are examined, and the technical standards for calculation of the primary suicide burden metric, called the Disability Adjusted Life Year (DALY), are reviewed. A schematic placing various suicide indicators within an established public health framework is presented within this discussion.
2
Population-Based Health Summary Measures: The Quality-Adjusted Life Year (QALY) and the Disability-Adjusted Life Year (DALY)
The newer classes of population-based health summary measures were introduced when, in 1968, a health expectancy measure was used to evaluate treatments for renal failure. Health gap measures were developed later, and represent a further development of the health expectancy measure methodology (Salomon and Lopez, 2002). The conceptual relationship between health expectancy and health gap measures is illustrated in > Figure 91-1. Both classes of indicators address the decline in health associated with a disease state or injury condition (A). However, health expectancy measures demonstrate the potential gain toward > ‘‘full’’ health status that would occur with the development of new treatment and/or prevention methods (B). In contrast, health gap measures illustrate the difference between ‘‘full’’ health status (C) and > actual health status (A) as it stands after adjustment for lives lost and disability sustained as a result of a disease process or injury condition.
1571
1572
91
The Disease Burden of Suicide: Conceptual Challenges and Measurement Standards
. Figure 91-1 Relationship between health expectancy and health gap summary measures Formulas: Health Expectancy = A + ƒ(B) Health Gap = C + g(B) Notes: Health expectancies measure the potential gain in health ƒ(B) resulting from the reduction of disability for a given condition over present status (A), while health gap measures illustrate the difference between ‘‘full’’ health status (C) and actual health status after adjustment for lives lost and disability sustained g(B) due to a given condition. The health expectancy increase over baseline identified above as f(B) is often expressed in terms of cost per QALY gained from a new treatment or prevention strategy. For any given health condition, the health gap ‘‘burden,’’ expressed as actual versus ‘‘full’’ population health status [identified above as g(B)] is often measured in terms of DALY rate per 100,000 people. (Adapted with permission from: Mathers CD, Lopez AD, Murray CLJ, 2006. ‘‘The burden of disease and mortality by condition: Data, methods and results for 2001’’)
Health expectancy measures are used in cost utilization analyses to evaluate specific treatments or interventions, while health gap measures are used to compare the relative ‘‘burden’’ placed on a population by the presence of one or more disease or injury conditions (Garcia-Altes, 2006). Health expectancy measure increase over baseline is often expressed in terms of cost per quality-adjusted life year (QALY) gained from a new treatment or prevention strategy. For any given health condition, the health gap ‘‘burden’’ metric, expressed as actual versus ‘‘full’’ population health status, is often measured in terms of the rate of disability-adjusted life year per 100,000 people. A further comparison of the health expectancy measure known as the ‘‘Quality Adjusted Life Year’’ (QALY), and the DALY is found in > Table 91-1. To date, the QALY has been used in a variety of cost utility studies, while the DALY has been used primarily in global burden of disease and injury (GBD) studies. Both the QALY and the DALY metrics are constructed on a ratio scale where the health decrement associated with ill health is on a continuum, between ‘‘perfect’’ or ‘‘full’’ health and death (Mathers et al., 2006). Using suicide as an example, calculation of both measures begins by establishing the
Definitions: A QALY is a year of life adjusted for its “quality” or its value. It is a composite health expectancy indicator of the relative improvement in quality and quantity of life associated with a specific treatment of a particular disease or injury condition. It addresses condition-specific mortality; mortality, prevalence, incidence, duration of disability; degree to which health status is affected
A DALY is a year of life adjusted for the average level of disability associated with a particular condition. It is a composite health gap measure of the relative burden placed on a population by the presence of a particular disease or injury condition for the year under study. It addresses condition-specific mortality; mortality, prevalence, incidence, duration of disability; degree to which health status is affected The relative value of lessthan-perfect nonfatal health states was originally established through expert panel or population-based survey, using methods similar to those described above. The “burden” of nonfatal health states are generally expressed on a numerical scale ranging from 0 to 1, in which 0 represents no burden/ “full” health, and 1 the burden of death due to the condition under study.
DisabilityAdjusted Life Years (DALY)
Definitions and Data Source
QualityAdjusted Life-Year (QALY)
Name of Measure Type of Weighting
A year in perfect health is considered equal to 0 DALYs. The value of a year in a disabled condition is discounted according to weights established via a multi-method surveillance system.
A year in perfect health is considered equal to 1.0 QALY. The value of a year in ill health is discounted. For example, a year bedridden might have a value equal to 0.5 QALY.
. Table 91-1 Summary measures of population health that integrate information
DALYS provide a straightforward partitioning of total burden by an exhaustive set of disease and injury categories, and are additive across disease categories Burden methodologies also lend themselves to comparison of the public health impact of a large number of conditions, and to crosscultural comparisons, attributable risk, cost effectiveness, and forecasting analyses.
The primary use of QALYs is within the framework of cost-effectiveness analysis, to assess the improvement in quality-adjusted life expectancy obtained through a specific health intervention relative to a situation in which either no intervention or a standard alternative intervention is provided. QALYs can therefore provide an indication of the benefits gained from a variety of medical procedures in terms of quality and life and survival for the patient.
Primary Applications
The Disease Burden of Suicide: Conceptual Challenges and Measurement Standards
91 1573
1574
91
The Disease Burden of Suicide: Conceptual Challenges and Measurement Standards
levels of mortality and disability (morbidity) associated with suicidal acts within a given population. Both health expectancy and gap measures establish level of disability as a function of: (1) the total estimated time the population spent in this state during the index year, and (2) the level of discomfort created by living in this disabled state relative to other disabling states (e.g., motor vehicle accident, cancer, blindness, etc.) For QALYs the degree of ‘‘gain’’ in health at the population level associated with a specific treatment or prevention strategy is then estimated and the ‘‘cost utility’’ of the proposed treatment relative to baseline is established. For the DALY, mortality and disability totals are added together to get a population-based measure of burden associated with suicidal behavior. Although neither the DALY nor the QALY has been used extensively in suicide research to date, a few such analyses have been conducted. For instance, QALYs were used to retrospectively establish the cost effectiveness of an area-based suicide prevention program (primarily a form of > ‘‘gatekeeper’’ training – see definitions) conducted within the Western Athabaskan Native American Tribe living in New Mexico (in Platt et al., 2006). The study estimated that the value of suicides prevented at $1.7 million, at a cost per Quality Adjusted Life Year gained of $419. The cost effectiveness of a brief cognitive behavioral intervention for reducing repeated suicide attempts among urban patients with high levels of comorbid depression and substance use was also evaluated with the QALY (Rothbard and Koizumi, 2006). The results of that analysis showed an incremental cost effectiveness of $14,126 per QALY at 18 month follow-up, which was regarded as ‘‘within the acceptable range’’ of economic costs, compared to other mental health interventions. In contrast to the QALY, the DALY has primarily been used to assess the national, regional and global burden of suicide. > Table 91-2 ranks countries with the highest suicide DALYS as calculated for the 2002 Global Burden of Disease and Injury (GBD) Update. Case Definitions and Data Sources: No globally accepted, operationalized definition of the term ‘‘suicide’’ or ‘‘suicide attempt’’ exists, and the case definition adopted by a given study is often derived from whatever data source is available to address the question under study. In public health studies, suicide data are most commonly taken from vital registration (mortality) and medical (morbidity) records. By 2003, the diagnostic framework used with both of these data types in 116 nations was some version of the World Health Organization’s (WHO) International Classification of Disease and Injury Manual (ICD – WHO, 1992). Under the ninth revision coding system, suicidal behaviors are among those acts classified as ‘‘intentional self injuries,’’ a classification that falls within the external cause of injury (E-codes) coding matrix in ICD-9. To facilitate standardized ICDbased analyses, the International Collaborative Effort for Injury (ICE) External Cause Committee has published a recommended list of ICD diagnostic codes that should be included in ICD-based suicide case definitions (Fingerhut, 2004). A comprehensive list of these ICE-recommended suicide codes can be found in > Table 91-3 by mechanism and ICD version. It is important to note that ICD-based case definitions of suicidal acts do not actually distinguish between self injurious behaviors in which the intent was to kill oneself via the act, and those in which this ‘‘suicidal intent’’ was absent. Therefore, the nonfatal condition identified by these codes more closely resembles the European category of ‘‘deliberate self harm’’ (also called ‘‘self-harm’’ or ‘‘self-injury’’ in this chapter) than a condition that could be exclusively labeled as ‘‘suicidal.’’ Despite this important caveat, these are the codes
The Disease Burden of Suicide: Conceptual Challenges and Measurement Standards
91
. Table 91-2 Highest thirty-five country-level suicide DALYs in 2002 Level of evidence Suicides per 100,000
DALY rate per 100,000
Age-standardized DALY per 100,000
Total mortality♦
Russian Federation
41.0
900
855
Level I or II Level III
Kazakhstan
37.1
892
869
Level I or II Level III
Lithuania
45.5
885
854
Level I or II Level III
Belarus
38.2
787
746
Level I or II Level III
Sri Lanka
31.9
734
694
Level III
YLD♦
Ukraine
35.8
694
658
Level I or II Level III
Lao People’s Democratic Republic
21.2
645
692
Level IV
Latvia
30.5
572
544
Level I or II Level III
Estonia
28.7
571
560
Level I or II Level III
Guyana
20.4
559
524
Level III
Slovenia
29.5
527
476
Level I or II Level III
India
17.4
512
504
Level III
Suriname
18.1
508
462
Level III
Finland
23.4
487
488
Level I or II Level III
Hungary
28.2
485
446
Level I or II Level III
China
20.9
429
403
Level III
Japan
24.6
427
399
Level I or II Level III
Jordan
17.2
413
464
Level III
Belgium
20.9
409
407
Level I or II Level III
Bhutan
13.5
405
433
Level IV
Bangladesh
12.2
387
384
Level IV
Kyrgyzstan
14.8
383
388
Level I or II Level III
Nicaragua
11.9
379
372
Level I or II Level III
Republic of Moldova
18.3
378
358
Level I or II Level III
Trinidad and Tobago
14.8
357
334
Level I or II Level III
Turkmenistan
12.5
357
360
Level I or II Level III
Poland
17.3
356
333
Level I or II Level III
Republic of Korea
18.2
351
315
Level I or II Level III
Nepal
10.3
349
355
Level IV
Papua New Guinea
10.0
347
350
Level IV
Myanmar
10.6
344
326
Level IV
Croatia
19.9
339
317
Level I or II Level III
Mongolia
12.3
323
330
Level I or II Level III
1575
1576
91
The Disease Burden of Suicide: Conceptual Challenges and Measurement Standards
. Table 91-2 (continued) Level of evidence Suicides per 100,000
DALY rate per 100,000
Age-standardized DALY per 100,000
Total mortality♦
Austria
18.2
323
303
Level I or II Level III
Mauritius
11.9
317
300
Level I or II Level III
YLD♦
Level of Evidence Code: (1) (Mortality Levels I & II) and (YLD Level III) = Death registration data, complete, or incomplete, containing useable information on causes of death is available and used to adjust regional YLD distributions for causes with significant case fatality. Partial country-specific information on incidence or prevalence of nonfatal causes available. (2) (Mortality Level III; Missing YLD) = Other forms of information on child and adult mortality or causes of death (e.g., verbal autopsy) available. Country-specific information on mortality for specific causes available. Partial country-specific information on incidence or prevalence of nonfatal causes available. (3) (Mortality Level IV; Missing YLD) = Country information on level of adult mortality not available and it was predicted from child mortality level OR cause of death information for most causes not available, and cause pattern predicted using cause-of-death models. Partial country-specific information on incidence or prevalence of nonfatal causes available Legend: DALYs – Disabilty Adjusted Life Years (World Health Organization, 2002 Global Burden of Disease and Injury Study. Available at: www.who.int/entity/ healthinfo/statistics/bodgbddeathdalyestimates/, Reprinted with permission)
used in most studies of suicide epidemiology. One of the only studies to date to assess the prevalence of suicidal acts among injuries diagnosed with these codes has estimated that between 60% and 70% of all medically-diagnosed nonfatal intentional self-harm events are suicide attempts (Ikeda and Mahendra, 2002). Definitions of Inputs for Suicide Summary Measures: A number of epidemiological parameters are required to calculate both QALY and DALY suicide summary measures. > Figure 91-2 depicts a version of the natural history of suicidal states showing the relationships between ten public health parameters that are important to the calculation of suicide health summary measures. While all ten parameters are important to the complete characterization of health-threatening suicidal experiences within a given population, for calculation of DALYs the most important parameters are: 1. 2. 3. 4. 5.
Population size by age group, Incidence of intentional self-harm and age of onset, Lifetime attempt rate by age group, Case fatality ratios (counts of suicides within the population), and Case complication ratios (disability incidence, duration, and distribution by severity class and age group (Michaud and McKenna, 2006), 6. Suicide rate. From these inputs, other GBD variables can also be derived. GBD researchers have developed a software program, called the DISMOD software program, which uses the expected causal relationships inherent in disease incidence – prevalence – mortality relationships to create a mathematical model that can impute missing values and correct error in eventbased datasets (Kruijshaar and Barendregt, 2002). The program establishes consistent matches between estimates of prevalence, duration, remission, and mortality for any given health condition, and it is these estimates that are used to calculate GBD for specific conditions.
91
The Disease Burden of Suicide: Conceptual Challenges and Measurement Standards
. Table 91-3 ICE External Cause of Injury Case Definitions for Suicide and Injuries of Undetermined Intent: ICD-9 and ICD-10 Equivalent Codes ICD-103
ICD-9
Mechanism/Cause
Self-inflicted
Undetermined
Selfinflicted
Undetermined
Cut/pierce
E956
E986
X78
Y28
Drowning/submersion
E954
E984
X71
Y21
Fall
E957.0-.9
E987.0-.9
X80
Y30
Fire/burn
E958.1,.2,.7
E988.1,.2,.7
X76–X77
Y26–Y27
E958.1
E988.1
X76
Y26
Fire/flame
E958.2,.7
E988.2,.7
X77
Y27
Firearm
Hot object/substance
E955.0–.4
E985.0–.4
X72–X74
Y22–Y24
All Transporta
E958.5–6
E988.5–6
X82
Y32
Poisoning
E950.0–E952.9
E980.0–E982.9
X60–X69
Y10–Y19
Suffocation
E953.0–.9
E983.0–.9
X70
Y20
Other specified and classifiableb,c,d
E955.5,.6,.7,.9 E958.0,.4
E985.5,.6,.7 E988.0,.4
X75, X81, * U03.0
Y25, Y31
Other specified, not elsewhere classifiable
E958.8, E959
E988.8, E989
X83, Y87.0
Y33, Y87.2
Unspecified
E958.9
E988.9
X84, *U03.9
Y34, Y89.9
All injury
E950–E959
E980–E989
X60–X84, Y87.0 *U03
Y10–Y34, Y87.2, Y89.9
a
Three 4th-digit codes (.4[occupant of streetcar], .5 [rider of animal], .8[other specified person]) are not presented separately because of small numbers. b Codes in bold are for morbidity coding only, and do not apply to mortality c Drowning is the one external cause that has been redefined in this matrix. Codes for water transportation-related drowning, V90 and V92, are included in the transportation codes rather than with the drowning codes. In the ICD-9 version of the matrix, the comparable codes, E830 and E832, were included with drowning. This change was made to be consistent with other mechanisms involved with water transport-related injuries. d This table contains the new ICD-10 codes for terrorism. The codes are preceded with*** Legend: ICE – International Collaborative Effort for Injury Statistics ICD – 9 International Classification of Diseases and Injuries, 9th Edition ICD – 10 International Classification of Diseases and Injuries, 10th Edition Source: (Fingerhut, 2004) Reprinted with permission
3
Is the Burden Associated with Suicidal States Acute Only, Episodic, Progressive, Injury-Related, or Disease-Based?
Assumptions about the nature and course of the suicidal state underlies any method used to analyze and calculate its associated burden. Because the datasets used to calculate suicide burden are most often mortality or medical data coded according to some version of the ICD diagnostic system (Begg and Tomijima, 2000), the suicidal state described in burden methodologies is defined by assumptions associated with the ICD diagnostic system. The ICD regards suicidal acts as one type of injury event. Even intentional injuries are traditionally understood to be
1577
1578
91
The Disease Burden of Suicide: Conceptual Challenges and Measurement Standards
. Figure 91-2 Definitions of common suicide-related public health indicators Legend: ICD-9 International Classification of Diseases and Injuries, 9th Edition (Adapted from the template ‘‘Diagram of Dynamic Disease Model’’ in Mathers, Sabate, and Lopez 2001Template used and adapted with permission)
predictable only in the ‘‘epidemiologic sense’’ (Institute of Medicine, 1999), and are regarded as occurring at an identifiable point in time and producing some measure of post-event disability in a portion of the affected population. Stated another way, those who sustain injury events are assumed to be in a state approximating full health until the point at which the event occurs. The fact that the overwhelming majority of suicides are the result of sudden traumatic insult related to blunt force, suffocation, or firearm reinforces this view, as does the absence of
The Disease Burden of Suicide: Conceptual Challenges and Measurement Standards
91
uniform, identifiable symptoms signaling the presence of an increasingly serious suicidal state (Rudd et al., 2006). However, suicidal states are often preceded by declines in cognitive and behavioral functioning (i.e., disability – Kaplan et al., 2007), which are not commonly associated with injury events. The other class of medical events included in the ICD lexicon is the disease state, and this type of ill-health experience can include symptoms that emerge across the course of the illness episode – some of which may be present before diagnosis. Kashner et al. (2000) suggest that, while suicidal behaviors do not represent a disease entity, they are ‘‘the most serious natural end point of a disabling mood or alcohol-related neuropsychiatric disease process.’’ The fact that a high proportion of self-harming individuals are diagnosed with a psychiatric disorder suggests that these disorders are indeed among the most salient of risk factors for suicidal behavior. However, multiple, biologically diverse, neuropsychiatric disease processes are known to precede suicidal acts (Harris and Barraclough, 1997), and a significant percentage of these acts are preceded by no identifiable disorder at all (Yang and Phillips, 2005) – highlighting the incomplete nature of existing evidence in support of the notion that all suicidal phenomena represent sequelae of neuropsychiatric disease processes. On death certificates, ICD nosologists have historically ‘‘resolved’’ problems created by the discrepant clinical presentations associated with suicidal behavior through the use of a set of coding rules where suicidal acts can be listed as either an injury event or a symptom of a neuropsychiatric disease state. If, for instance, a psychiatric disorder is prominent in the time period preceding a suicide and the death certifier believes that treatment of this disorder would have averted the suicide, until 2003, ICD nosologists ‘‘coded’’ the ‘‘underlying cause’’ of the act as that psychiatric disorder. This is one reason that depression or other psychiatric disorders were sometimes found among cause-of-death listings for suicide. However, where no preinjury psychiatric disorder could be identified, the suicidal act itself became the underlying cause of death. Classification of suicidal events as injuries has profound implications for the calculation of suicide burden. When coded in this manner, measurement of the burden associated with nonfatal self injury involves only disability occurring from the point of the injury forward in time. Pre-injury burden is assumed to be either absent or not causative. While the ICD coding conventions partially resolve problems related to diverse suicidal ICD coding, they do not produce a coherent understanding of the suicidal state. Once a serious suicidal state emerges, its course is often more like a chronic disease process with intermittent ‘‘illness episodes’’ than like an injury ‘‘event.’’ For one thing, the experience of nonfatal suicidal behavior itself generates a lifetime vulnerability to additional self-harming episodes. The 13-country, 5-year WHO/EURO multicenter study on Suicidal Behavior found that over 50% of self-injurers experienced more than one self harm-related injury, with nearly 20% of second episodes occurring within 12 months of the first (Schmidtke et al., 1996). Presumably, some portion of these events can be understood as part of the ‘‘illness episode’’ signaled by the first event. A British cohort of over 11,000 intentional self-injurers followed for over a decade likewise demonstrated a 39% lifetime recurrence rate (Hawton et al., 2006). In short, as is the case for many diseases, the evidence suggests the existence of a suicidal process that varies both in course and outcome, but that can carry substantial within-episode pre- and post-disability, as well as a marked vulnerability for recurrence. If defined as a process, however, separation of pre-injury suicide burden-related disability from common psychiatric and medical disorder sequelae will require substantial additional study.
1579
1580
91 4
The Disease Burden of Suicide: Conceptual Challenges and Measurement Standards
Methods and Measurement Standards for Calculating the Burden of Suicidal Acts via the DALY: Step One – Years of Life Lost to Suicide
As above, the DALY is a population-based, cross-sectional, health gap indicator that incorporates both fatal and nonfatal outcomes, taking into account the severity and duration of common disabling sequelae (Murray and Lopez, 1997). It has been used in a series of WHOsponsored global burden studies, including the 1990 study calculating global disease rates (GBD1990), the subsequent update, and comparative risk assessment studies (GBD2000, GBD2002), the currently underway 2005 methods revision and update study (GBD2005), and over 30 national and local burden studies. Applied to intentional self-harm, the DALY is calculated as the sum of all life-years lost (YLL) plus the sum of all life years lived in a disabled state (YLD) as a result of an episode of nonfatal intentional self-harm: DALY ¼ YLL þ YLD Calculation of the mortality portion of the suicide DALY (years of life lost to suicide) is fairly straightforward. The difference between life expectancy and age at suicide is summed for each suicide, represented as: YLL ¼
X
dx ðL xÞ
where: L_x is the population-based life expectancy at a given age at which suicide occurs – L_x is available from WHO in the life tables for each respective country. dx is the number of suicides occurring in the population at each age x Depending on the research question under study, this formula can be adjusted to account for population growth, unstable age structure, and change in death rates over time (Gunnell and Middleton, 2003). ‘‘Full’’ life expectancy is an upper limit age, based on convenience or appropriateness to a given research question. GBD1990 set the standardized life expectancy at 82.5 years for females, and the corresponding GBD male life expectancy at 80 years, to account for ‘‘probable biological differences’’ in longevity (Murray and Lopez, 1996). However, considerable debate exists about how to establish this presumed ‘‘full’’ life expectancy (MMWR, 1986). In light of recent advances in the understanding of gender-based risk factors for premature death, GBD2005 is likely to drop gender differences in full life expectancy (Murray, 2008). The answers to two related questions are important in the development of YLL measures: 1. (Age weighting) whether there are some age groups for which premature death is regarded as more ‘‘tragic’’ than for others and therefore should be weighted more heavily in calculating a population’s YLL (i.e., when the individual dies in the ‘‘prime’’ of life). 2. (Time Discounting) whether a nation’s citizens are more willing to expend public health dollars on symptom alleviation now or on research that may cure the disease or injury condition in 30 years (e.g., ‘‘health now’’ vs. ‘‘health later’’). The GBD age weighting paradigm gives more weight to deaths occurring in adolescence and early adulthood over those occurring during other life states, using methods that approximate the values illustrated in > Figure 91-3. For GBD2005, age weighting, but not
The Disease Burden of Suicide: Conceptual Challenges and Measurement Standards
91
. Figure 91-3 Approximate relative age weighting applied to GBD YLL calculations across the age span Legend: GBD – Global Burden of Disease and Injury Studies YLL – Years of Life Lost
discounting, is likely to be applied to baseline burden calculations (Murray, 2008). Because the frequency of suicide is highest at the ages weighted more heavily in the GBD age weighting paradigm in most parts of the world, suicide consistently ranks as one of the most burdensome of all conditions analyzed in GBD studies. (Methods for calculating GBD age weights are found in Murray, 1996.) Once these calculations are completed – or a decision is made not to apply age weighting or discounting to suicide data – a life table can be constructed, and the relative contributions to the YLL posed by each suicide occurring within the population for the year under study can be calculated and summed together. Challenges and Issues when Using Suicide Data to Calculate YLLs: Suicide data are the end product of a reporting chain that includes new potential sources of error at each level (Claassen Yip et al., in press). During cause of death determination, factors such as death circumstance, occupation, and training of the death investigator, death office caseload, and social stigma contribute to underreporting of suicides. In addition, completion of the death certificate is frequently inaccurate, and the stringency of the follow-back procedures used by the mortality reporting agency to correct obvious certificate errors affects data completeness and reliability. Among the annual mortality datasets provided to the WHO for 107 countries in the world, evidence of at least some degree of poor death reporting practice is identifiable (Mathers and Ma Fat, 2005). To correct for some mortality classification/death certificate coding errors, GBD researchers have developed specific algorithms that redistribute poorly-coded or obviously miscoded deaths to other cause of death categories (Michaud and McKenna, 2006). ‘‘Proportional reallocation’’ of deaths to new cause codes is done within morality ‘‘envelopes,’’ created by breaking mortality data into 5-year age groups, where the total number of deaths after reallocation is required to equal the number prior to this process. For example, the approximately 4,500 injury deaths of undetermined intent occurring in the US annually (CDC, 2008) are proportionally reclassified to other injury-related causes within > age-based mortality envelopes, with about 50% of these deaths reassigned as suicides (Michaud and McKenna, 2006). Ill-Defined Cause Deaths: Under current GBD procedures, deaths classified as ‘‘Signs, Symptoms and Ill-Defined Cause’’ are reallocated among candidate disease states but not to
1581
1582
91
The Disease Burden of Suicide: Conceptual Challenges and Measurement Standards
injury conditions (e.g., none of these deaths gets reassigned as a suicide) (Michaud and McKenna, 2006). However, in countries where a death certificate has to be filled out before burial, the ill-defined cause category is often used as a ‘‘placeholder’’ cause on death certificates until an investigation can be concluded. If cause-of-death is not routinely updated in mortality datasets at the conclusion of investigation, this placeholder becomes the final underlying cause used for official reporting purposes. Hawton et al., 2006 examination of all-cause mortality among 11,000 prior suicide attempters found 3.9 times more ill-defined cause deaths than would have been expected based on all-cause proportions (Hawton et al., 2006), and a case review study in France found that 25% of all ill-defined cause deaths were probable suicides (Jougla and Pequignot, 2002). Intentional Deaths among Infants and Young Children: One additional common source of error in suicide death classification occurs as a result of coding suicide as the cause of death among the very young. The capacity to understand the consequences of deliberate, serious self injury is believed by many child development experts to be beyond the cognitive abilities of infants and young children (Grollman, 1967). Therefore, intentional self-injury is never identified as the cause of death in children under the age of five in the death registries of many countries (Miller et al., 2004), and a number of researchers refuse to list suicide as the cause of death for any individual under the age of 10.
5
Step Two – Estimating the Years of Life Lived with Disability Due to Nonfatal Suicidal Behaviors
Estimating the years lived with a disability (YLD) is regarded as the more complex step in burden calculations, requiring systematic assessment of available evidence on incidence, prevalence, duration and severity of disability for the condition under study (Murray and Lopez, 1997). For some aspects of this process, adequate population-based data is not often available. Complicating the process further is the range of possible future disability scenarios for each condition studied, which varies both by severity and length of time lived in the disabled state. The basic approach to establishing YLD levels involves calculating, for each incident, the degree of disability times the duration of the disability as follows (Murray and Lopez, 1996): YLD ¼ x:D:L where: x is the number of incident cases D is the disability ‘‘weight’’ for the condition (0 = perfect health; 1 = dead) L is the average duration of disability in years This approach quantifies, for each disabling condition, the relative distance between a state of ‘‘absolute’’ or ‘‘full’’ health and the state of reduced health due to nonfatal disability, represented as the severity and duration of the injury episode. Ratings for each disabling condition are developed as relatively constant, homogeneous experiences, symbolizing the total health loss associated with the injury event in the absence of any comorbid condition (Begg and Tomijima, 2000; Murray and Lopez, 1996). The data needed to calculate the suicide-related YLD for each 5-year age group include:
The incidence of suicide attempts with nonfatal outcomes by mechanism and nature of injury;
The Disease Burden of Suicide: Conceptual Challenges and Measurement Standards
91
The duration of injury-related sequelae, classified by severity level; and Age at onset of the disability for each event occurring in the population for the year under study.
Case Definitions: Nonfatal injury events included in GBD calculations are injury episodes severe enough to require either emergency care or hospitalization (Begg and Tomijima, 2000), regardless of whether appropriate care is accessible. To understand the DALY’s notion of nonfatal suicide burden, it is necessary to distinguish between condition-associated impairment, disability, and handicap (> Figure 91-4). For instance, a serious intentional overdose in which the agent is an analgesic such as paracetamol (acetaminophen) might result in hepatotoxicity (impairment), followed by liver failure (disability), and unemployment due to the need for ongoing medical care (handicap). It is the disability associated with injury conditions that is measured in the YLD portion of the DALY, rather than either the impairment or the handicap. The rationale for this approach is that disability is the focus of those rehabilitation services designed to minimize functional limitations, which vary widely as a result of any given condition. Disability Weights: YLD disability weights range from 0 (ideal health) to 1 (death), and are intended to be a clear measure of relative health decrement (Bickenback, 2008). These weights reflect preferences for living in one state of reduced health versus another one, and are quantified according to the ‘‘severity’’ of disability associated with different disease states and injury conditions. The original disability weightings were developed for GBD 1990 by an expert panel for twenty-two different disability conditions, using a method called the Person-Trade-Off exercise (Michaud et al., 2006). After these weights were developed, the same group of experts used a modified Delphi approach to weight over 400 other health states, using these twenty-two indicator conditions as benchmarks (Murray and Lopez, 1996). The range of disability weights established during GBD1990 included the following: severe sore throat/anemia 0.021–0.120, radius fracture in a stiff case/angina 0.12–0.240, below-knee amputation/deafness 0.241–0.360, rectrovaginal fistula/mental retardation/Downs’ syndrome 0.361–0.500, unipolar depression/ blinness/paraplegia 0.5010–00.700, and active psychosis/dementia (with memory impairment, aphasia, apraxia)/quadriplegia 0.7010–01.000 (Murray and Lopez, 1997). Subsequent work suggested that a reasonably high level of agreement on disability weights across condition was present across a number of countries (Schwarzinger and Stouthard, 2003). Nature of Injury versus External Cause: An important element in the process of establishing disability weights is defining the nature of the injury. In contrast to the external cause of injury (i.e., codes for unintentional injury, suicide, homicide, undetermined intent injuries), ICD coding rules define the nature of the injury according to the actual bodily harm sustained. Severity of injury and duration of disability are more easily established using classifications based on nature of injury than on external cause of injury, although they . Figure 91-4 Relationship between impairment, disability, and handicap (From Murray & Lopez, 1996) Legend: (Murray and Lopez, 1996. Reprinted with permission)
1583
1584
91
The Disease Burden of Suicide: Conceptual Challenges and Measurement Standards
are then attributed back to cause-specific injury classes. These codes are broken down into 33 categories for GBD analyses (> Table 91-4). Using nature of injury codes, external cause conditions are classified by expected severity and expected duration of disability (both short- and long-term), whether treated or untreated. Some nature of injury categories involve only short-term disability (e.g., open wounds, fractured arm) and others involve only cases of permanent disability (e.g., amputations, spinal cord lesions). A few categories, however, contain both short- and long-term disability (e.g., head injuries). Sensitivity analyses conducted as part of GBD1990 revealed that short-term injuries had little impact on overall YLD calculations, so essentially only long-term disability significantly adds to the YLD for a given injury class (Begg and Tomijima, 2000; Murray and Lopez, 1996). . Table 91-4 GBD nature of injury case definitions Nature of injury
ICD-9 CODE
ICD-10 CODE
FRACTURES Skull – short-term
800–801
S02.0/1/7/9, T90.2
Skull – long-term
800–801
S02.0/1/7/9, T90.2
Face bones
802
S02.2/6/8
Vertebral column
805
S12, S22.0/1, S32.0/7, T91.1
Rib or sternum
807
S22.2–9
Pelvis
808
S32.1–5/8, T91.2
Clavicle, scapula, or humerus
810–812
S42, S49.7
Radius or ulna
813
S52, S59.7, T10, T92.1
Hand bones
814–817
S62, S69.7, T92.2
Femur – short-term
820–821
S72, S79.7
Femur – long-term
820–821
S72, S79.7
Patella, tibia, or fibula
822–823
S82.0–4, S82.7/9, S89.7, T12
Ankle
824
S82.5–6/8
Foot bones
825–826
S92, S99.7
INJURED SPINAL CORD
806, 952
S14, S24, S34, T06.0–1, T08, T91.3
Shoulder, elbow, or hip
831, 832, 835
S03.0–3, S13, S23, S33, S53, S63.0–1, S83.1–3, S93.1–3, T03, T11.2, T13.2, T14.3, T92.3, T93.3
Other dislocation
830, 833–834, 836–839
S03.0–3, S13, S23, S33, S53, S63.0/1, S83.1–3, S93.1–3, T03, T11.2, T13.2, T14.6, T92.5, T93.5
SPRAINS
840–848
S0.34/5, S16, S29.0, S39.0, S46, S56, S63.5–7, S66, S76, S83.4/7, S86, S93.4/6, S96, TY06.4, T11.5, T13.5, T14.6, T92.5, T93.5
Short-term
850–854
S06, T90.5
Long-term
850–854
S06, T90.5
DISLOCATIONS
S43, S73
INTRACRANIAL INJURIES
The Disease Burden of Suicide: Conceptual Challenges and Measurement Standards
. Table 91-4 (continued) Nature of injury
ICD-9 CODE
91
ICD-10 CODE
INTERNAL INJURIES
860–869
S25-S27, S35-S37, S39.6, T06.4, T91.4/5
OPEN WOUND
870, 872–884, 890–894
S01, S08, S11, S15, S21, S31, S41, S45, S51, S55, S61, S65, S71, S75, S81, S85, S91, S95, T01, T11.1/4, T13.5, T14.6, T90.1, T92.5, T93.5
Short-term
871, 950
S05, T90.4
Long-term
871, 950
S05, T90.4
Thumb
885
S680
Finger
886
S68.1/2
Arm
887
S48, S58, S68.3–9, T05.0/2, T11.6
Toe
895
S98.1/2
Foot
896, 897.0–1
S98.0/3/4, T05.3
Leg
897.2–3
S78, S88, T05.4/6, T13.6
CRUSHING
925–929
S07, S17, S28, S38, S47, S57, S67, S77, S87, S97, T04, T14.7, T92.6, T93.6
Less than 20% – short-term
940–947, 948.0–1
T31.0/1
Less than 20% – Long-term
940–947, 948.0–1
T31.0/1
20 – 60% – short-term
948.2–5
T331.2/5
INJURY TO EYES
AMPUTATIONS
BURNS
20 – 60% – long-term
948.2–5
T331.2/5
Greater than 60% – short-term
948.6–9
T31.6/9
Greater than 60% – long-term
948.6–9
T31.6/9
Short-term
951, 953–957
S04, S44, S54, S64, S74, S84, S94, T06.2, T11.3, T13.3, T14.4
Long-term
951, 953–957
S04, S44, S54, S64, S74, S84, S94, T06.2, T11.3, T13.3, T14.4
POISONING
960–979, 980–989 T36–T65, T96–T97
INJURED NERVES
Legend: GBD – Global Burden of Disease and Injury Studies ICD-9 International Classification of Diseases and Injuries, 9th Edition ICD-10 International Classification of Diseases and Injuries, 10th Edition ((Murray and Lopez, 1996) Reprinted with permission)
Establishing Expected Severity and Weights for Disability Classes: Data from epidemiological reviews, expert panels and population-based surveys are used to estimate the proportion of injuries in each class that result in long-term, severe disability, as well as the severity of the disability. The process of establishing initial YLD estimates for self-injuries, therefore, involves estimating the percentage of these injuries falling into each severity cell by mechanism within age-based mortality envelopes. To date, similar disability weights have been applied to injuries within the same nature of injury cell, regardless of external cause (Begg and Tomijima,
1585
1586
91
The Disease Burden of Suicide: Conceptual Challenges and Measurement Standards
2000), although some preliminary evidence suggests that intentional self harm events with the same injury severity scores result in greater disability (David et al., 2007). Challenges and Issues when using Intentional Self-Harm Data to Calculate YLDs: Where population-based morbidity data are available, data source profoundly affects the rate of nonfatal suicidal behavior found. There tends to be a stable, significant difference in the incidence of self-reported nonfatal ‘‘suicide attempt’’ found via > population-based surveillance and that identified within a medical registry. In the US, Kessler et al., 1990–1992 National Comorbidity Study and its 2001–2003 replication suggested a stable, population-based 12-month incidence of suicidal behavior of 0.6–0.9% within the US population (600–900/ 100,000 – Kessler et al., 2005). In contrast, while no national medical registry of suicidal events exists for the US, the National Electronic Injury Surveillance System – All Injury Program’s annual, estimated incidence of nonfatal intentional self-harm events treated in US emergency departments (EDs) between 2001 and 2003 was 122.2/100,000 US population (CDC, 2008). It is assumed that the definition of suicidal behavior as used in the general population varies widely by individual, and that definitions of self-reported suicidal act are equally as variable. Presumably, the medical diagnosis of intentional self-harm involves uniform training in the use of standardized diagnostic criteria (ICD codes) and standardized application of diagnoses across cases and treatment providers. Medical records, therefore, are believed to provide more reliable information on incidence of nonfatal suicidal behavior. Emergency department-based counts of new cases of injuries – where they represent all cases treated within a given geographic area – serve as a primary source of data for GBD YLD injury calculations. In de-identified medical registries, the incidence of nonfatal intentional self-harm events is often mistakenly equated with the number of ED visits. However, at least some patients likely make more than one visit to a medical facility for treatment of the same injury. A brief comparison of two surveillance systems currently operating in the US provides some information about the number of self-harm episodes treated more than once in US EDs. The National Hospital Ambulatory Medical Care Survey (NCHAMCS) estimates the number of visits to US emergency departments, counting each visit, regardless of whether the injury being treated has also been treated during a prior visit (McCaig, 2004). In contrast, the National Electronic Injury Surveillance System – All Injury Program estimates the number of first visits to US emergency departments for treatment of a new injury condition (CDC, 2008). Assuming that both systems provide reasonably accurate estimates of US emergency department visit frequency, the 2004–2005 average annual number of intentional self-injury events occurring within the country as estimated by the NHAMCS was 161.83 visits per 100,000 person years (McCaig and Nawar, 2006; Nawar and Niska, 2007). The corresponding 2004–2005 NEISS-AIP first-visit estimate was 135.28 per 100,000 person years (CDC, 2008), suggesting that approximately 16.4% of all ED intentional self harm visits for those 2 years were made for follow-up care of a previously-treated nonfatal intentional self-harm event. Mathers et al. (2006) reports that when the 2000–2002 GBD studies attempted to locate datasets from which to estimate YLD for 107 disease and injury conditions, nearly 6,600 were found to inform estimates of disability due to disease states, while only 18 adequate datasets world-wide were located to inform YLD estimates of long-term injury sequelae. Among these 18 datasets, one quarter involved populations in sub-Saharan Africa, and one-fifth were from high-income countries, requiring disability estimates to be imputed for most of the world. Because of the paucity of data on nonfatal injury, the incidence of disabling events is frequently
The Disease Burden of Suicide: Conceptual Challenges and Measurement Standards
91
imputed in GBD studies from age-specific case fatality ratios, under the assumption that incident-to-death ratios for many injury events are relatively constant across time and geography after taking into account available emergency medical care (Begg and Tomijima, 2000; Murray and Lopez, 1996). Standardized GBD estimating models use mortality rates and per capita gross domestic product (GDP) as rough proxies for access to health care in these models. This practice is somewhat less relevant to suicide analyses because the majority of nonfatal suicidal acts are self-poisonings which do not result in inpatient treatment or longterm physical disability (Fleishmann and Bertolote, 2005). Therefore in countries with reasonable access to emergency care, suicide ‘‘burden,’’ becomes largely a function of the years of life lost to suicide, while the impact of nonfatal suicidal acts on health status, contributes little to overall suicide burden (Begg and Tomijima, 2000). However, without sap access, nonfatal burden contributes more substantially to overall burden. Applications to Other Conditions: The misclassification of deaths is a common problem among a handful of injury death categories besides suicide (e.g., certain falls, single-vehicle accidents, certain poisonings). For these injury categories, many of the same burden-relates issues arise as for suicides.
6
The Importance of Suicide Burden Calculations
As the time this chapter was written, suicide metrics were being selected for a number of highprofile local, national, and global public health projects. For instance, suicide measurement tools were being developed to accompany the suicide prevention goals in national public health agendas for the decade beginning in 2010 (DHHS, 2008). In addition, mentioned earlier, GBD2005 – which involves over 800 public health researchers worldwide – is currently developing comparative estimates of the health burdens associated with 175 different disease states and injury conditions across 21 regions of the world. Knowledge of the suicidal process is far from complete, and appropriate data, error correction methods, and measurement strategies are limited. While this deficit should not stop attempt to the measure population-based leveis of suicided behavior, it is imperative that the best, most current knowledge available be employed during each successive iteration of these measurement processes. In the case of suicide burden, this will require ongoing developmental the metrics used to measure it. > Table 91-5 shows GBD2002 projected rankings for the disease states and injury conditions with the greatest global burdens. These DALY-based projections suggest that, the global burden of suicide will rise from 17th to 14th by the year 2030 (Mathers and Loncar, 2006). However, the GBD2005 decision to drop gender differences in full life expectancy (Murray, 2008) means that the number of YLLs lost to suicide will increase comparatively for approximately three out of every four suicides worldwide, thereby increasing the overall estimated suicide burden substantially. Population analysts, advocacy groups, and policy makers at all levels will need to acquire the knowledge base necessary to interpret such changes in burden methodologies, and understand the strengths and limitations of the summary health measures used to inform decision-making about investment of public health resources. Such literacy is likely to become increasingly important as methods are refined and measures change over the years to come.
1587
1588
91
The Disease Burden of Suicide: Conceptual Challenges and Measurement Standards
. Table 91-5 2002 ranked burden as measured in DALYs of selected conditions, and projected change for the year 2030 Category Within top 15
Condition Perinatal conditions
2002 Rank
2030 Projected Rank
Change in Rank
1
5
4
Lower respiratory infections
2
8
6
HIV/AIDS
3
1
+2
Unipolar depressive disorders
4
2
+2
Diarrheal diseases
5
12
7
Ischemic heart disease
6
3
+3
Cerebrovascular disease
7
6
+1
Road traffic accidents
8
4
+4
Malaria
9
15
6
Tuberculosis
10
25
15
COPD
11
7
+4
Congenital anomalies
12
20
8
Hearing loss, adult onset
13
9
+4
Cataracts
14
10
+4
Violence
15
13
+2
Outside top 15 Self-inflicted injuries Diabetes
17
14
+3
20
11
+9
Legend: DALYs – Disability Adjusted Life Years HIV/AIDS – Human Immunodeficiency Virus/ Acquired Immune Deficiency Syndrome COPD – Chronic Obstructive Pulmonary Disease (Mathers CD, Loncar D (2006) Projections of global mortality and burden of disease from 2002 to 2030. PloS Med 3 (11): 3442. doi:10.1271/journal.ped.0030442 Reprinted with permission)
Summary Points Despite 8 decades of research documenting their inherent error, suicide rates remain the most widely used population-based suicide metric today.
Suicidal behavior is classified as an injury event for the purposes of burden analysis; however, it is often preceded by a decline in functioning and often follows a chronic, recurrent course – bearing more resemblance to a disease process than to an injury event. No globally accepted, operationalized definition of the term ‘‘suicide’’ or ‘‘suicide attempt’’ exists, and the case definition adopted by a given study is often derived from whatever data source is available to address the question under study. Although no death classified as ‘‘Signs, Symptoms and Ill-Defined Cause’’ is reassigned as a suicide under current burden methods, recent evidence suggests that a significant number of these deaths may in fact be suicides.
The Disease Burden of Suicide: Conceptual Challenges and Measurement Standards
91
In countries with reasonable access to emergency care, the majority of nonfatal suicidal
acts do not result in long-term physical disability, so suicide ‘‘burden’’ is largely a function of mortality-related years of life lost. The recent GBD2005 decision to drop gender differences in full life expectancy calculations means that the level associated with number of years of life lost to suicide will likely rise beyond the currently projected suicide ranking of 14th most burdensome condition by the year 2020. It is imperative that the best, most current knowledge available be employed during each successive iteration of the burden measurement process. In the case of suicide, this will require ongoing conceptual and method logizal retirements.
References Begg S, Tomijima N. (1986). MMWR. 25: 1–11. Begg S, Tomijima N. (2000). In: Global Program on Evidence for Health Policy. Global Burden of Disease 2000. World Health Organization, Geneva. Begg S, Tomijima N. (2008). In: Public Health Agency of Canada Initiatives [online] Accessed 2008 Jan 15. Available at: http://www.phac-aspc.gc.ca/ph-sp/phdd/ initiative/index.html Bickenback J. (2008). In: Ethical Issues in the Measurement of Health, Boston, MA, Harvard University [online] Accessed 2008 May 1. Available at: peh. harvard.edu/events/2008/global_burden_disease/ CDC. (2008). In: Web-based Injury Statistics Query and Reporting System (WISQARS). [Online Database]. National Center for Injury Prevention and Control. Accessed 2008 1 Jan. Available at: www.cdc.gov/ ncipc/wisqars Claassen CA, Yip SFP, Bossarte RM, Corcoran P in press, Suicide and L.h-Threatening Behavior Corso PS, Mercy JA, Simon TR, Finkelstein EA, Miller TR. (2007). Am J Prev Med. 32: 474–482. David JS, Gelas-Dore B, Inaba K, Levrat A, Riou B, Gueugniaud PY, Schott AM. (2007). J Trauma. 62: 1495–1500. DHHS. (2008). In: Office of Disease Prevention and Health Promotion. Department of Health and Human Services [online] Accessed 2008 Feb 15. Available at: www.healthypeople.gov Durkheim E. (1951 [1897]). Suicide. Free Press. New York, Fingerhut L. (2004). WHO Family of International Classifications Network Meeting. Reykajavik, Iceland, pp. 1–5. Fleischmann A, Bertolote JM, De Leo D, Botega N, Phillips M, Sisask M, Vijayakumar L, Malakouti K, Schlebusch L, De Silva D, Nguyen VT, Wasserman D. (2005). Psychol Med. 35: 1467–1474.
Garcia-Altes A. (2006). CAHTA Newsletter. Barcelona, Spain, Catalan Agency for Health Technology Assessment 38. Grollman E. (1967). Explaining Death to Children. Beacon Press, Boston. Gunnell D, Middleton N. (2003). Lancet. 362: 961–962. Harris EC, Barraclough B. (1997). Brit J Psychol. 170: 205–228. Hawton K, Harriss L, Zahl D. (2006). Psychol Med. 36: 397–405. Ikeda R, Mahendra R. (2002). MMWR, 51: 436–438. Institute of Medicine. (1999). Reducing the Burden of Injury: Advancing Prevention and Treatment. National Academy Press, Washington, DC. Jougla EF, Pequignot E. (2002). Rev Epidemiol Sante. 50: 49–62. Kaplan MS, McFarland BH, Huguet N, Newsom, JT. (2007). Am J Orthopsychiat. 77: 56–60. Kashner MT, Shoaf T, Rush AJ. (2000). Econ of Neurosci. 2: 44–48. Kessler RC, Berglund P, Borges G, Nock M, Wang P. (2005). JAMA, 293: 2487–2495. Kruijshaar M, Barendregt J. (2002). Bull World Health Organ. 80: 622–628. Mathers CD, Loncar D. (2006). Plos Med. 3: 3442. Mathers C, Lopez A, Murray CJL. (2006). In: Lopez A, Mathers C, Ezzati M, Murray C, and Jamison D (ed.) Global Burden of Disease and Risk Factors, Oxford University Press, New York. Mathers C, Ma Fat D. (2005). Bull World Health Organ. 83: 171–177. McCaig L. (2004). 2004 User’s Data Conference [online] Accessed 2004 September 27, 2004. Available at: http:// www.cdc.gov/nchs/about/major/ahcd/ahcd1.htm. McCaig L, Nawar E. (2006). National Hospital Ambulatory Medical Care Survey: 2004 Emergency Department Summary. Adv Data, No. 372 Hyattsville, MD, National Center for Health Statistics.
1589
1590
91
The Disease Burden of Suicide: Conceptual Challenges and Measurement Standards
Michaud CM, McKenna MT. (2006). Popul Health Metr. 4: 1–49. Miller M, Azrael D, Hemenway, D. (2004). Ann Emerg Med. 43: 723–730. Mohler B, Earls F. (2001). Am J Public Health. 91: 150–153. Murray CJL. (2008). In: Harvard University Program in Ethics and Health [online] Accessed 2008 May 10. Available at: http://peh.harvard.edu/events/2008/ global_burden_disease/day2.html. Murray CJL, Lopez AD. (1996). Science. 274: 740–743. Murray CJL, Lopez AD. (1996). The Global Burden of Disease: A Comprehensive Assessment of Mortality and Disability from Diseases, Injuries and Risk Factors in 1990 and Projected to 2020. Harvard University Press, Cambridge, MA. Murray CJL, Lopez AD. (1997). Lancet, 349: 1347–52. Murray CJL, Lopez AD. (1997). Bull World Health Organ. 75: 377–381. Murray CJL, Lopez AD. (2000). Health Econ. 9: 69–82. Nawar E, Niska R. (2007). Adv Data, No. 386. Hyattsville, MD, National Center for Health Statistics. Platt S, McLean J, McCollam A, Blamey A, Mackenzie M, McDavid D, Maxwell M, Halliday E, Woodhous A. (2006). Evaluation of the First Phase of Choose Life: The National Strategy and Action Plan to Prevent Suicide in Scotland. Department of Scottish
Ministers, Edinburgh. [online] Accessed 2008 Jan 15. Available at: http://www.scotland.gov.uk/Publications/2006/09/06094756/1. Rothbard A, Koizumi N. (2006). Public Health and Human Rights. American Public Health Association, Boston, MA. Rudd MD, Berman AL, Joiner TE Jr, Nock MK, Silverman MM, Mandrusiak M, VanOrden K, Witte T. (2006). Suicide Life-Threat. 36: 255–262. Salomon JD, Lopez AD. (2002). Summary Measures of Population Health: Concepts, Ethics, Measurement and Application. World Health Organization, Geneva. Schmidtke A, Bille-Brahe U, DeLeo D, Kerkhof A. (1996). Acta Psychiat Scand. 93: 327–338. Schwarzinger M, Stouthard M. (2003). Popul Health Metr. 1: 15. WHO. (1948). World Health Organization Constitution. World Health Organization, Geneva. WHO. (1992). International Statistical Classification of Diseases and Related Health Problems, 10th rev. Geneva, Switzerland, World Health Organization. Yang G, Phillips MR. (2005). Biomed Environ Sci. 18: 379–89. Yip PSF, Liu KY, Law CK, Law YW. (2005). Crisis. 26: 156–159. Zilborg G. (1935). Am J Psychol. 92: 34–52.
92 The Disease Burden Due to Epilepsy in Rural China D. Ding . W. Z. Wang . Z. Hong 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1592 2 Study Areas and Target Population . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1593 3 Prevalence and Treatment Gap . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1593 4 Disease Burden by DALY Measure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1595 5 Applications of the Model to Other Diseases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1597 6 Premature Mortality in People with Epilepsy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1598 7 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1600 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1600
#
Springer Science+Business Media LLC 2010 (USA)
1592
92
The Disease Burden Due to Epilepsy in Rural China
Abstract: The World Health Organization (WHO), in cooperation with the International League Against Epilepsy (ILAE) and the International Bureau for Epilepsy (IBE) launched the Global Campaign Against Epilepsy (GCAE) in an attempt to bring epilepsy “out of the shadows” and to improve the acceptability, treatment, services and prevention of epilepsy world-wide. As part of this campaign and under the auspices of the WHO and the Ministry of Health of China, a demonstration project was implemented in rural areas in six non-contiguous provinces of the People’s Republic of China (PRC). In the sample areas, the lifetime prevalence and prevalence of > active epilepsy in rural China were 6.0–7.0/1,000 and 3.8–4.6/1,000 respectively. Sixty-three percent of people with active epilepsy had not received antiepileptic medication in the week before the survey (i.e., the > treatment gap was 63%). The disease burden of epilepsy was measured as 2.08 Disability Adjusted Life Years (> DALYs) lost per 1,000 population. The > standardized mortality ratio (SMR) of people with epilepsy was 3.9 indicating that people with epilepsy are three to four times more likely to die prematurely than those in the general Chinese population. The risk in young people with epilepsy aged 15–29 years in China is particularly high. List of Abbreviations: AED, anti-epileptic drug; CFR, > case fatality rate; DALY, disability adjusted life year; EMPHL, epilepsy management at primary health level; GBD, global burden of disease; GCAE, global campaign against epilepsy; IBE, International Bureau for Epilepsy; ICBERG, International community-based epilepsy research group; ILAE, International League Against Epilepsy; PMR, > proportional mortality rate; PRC, People’s Republic of China; SMR, standardized mortality ratio; WHO, World Health Organization; YLD, years lived with disability; YLL, years of life lost
1
Introduction
Epilepsy is one of the most common serious neurological disorders, affecting approximately 50 million people worldwide; approximately 2 million people develop epilepsy each year. With the correct diagnosis and treatment, many people with epilepsy will have a significant reduction in seizure frequency or be seizure free. In resource-poor countries, 60–90% of people with epilepsy receive no treatment or are inadequately treated due to deficiencies in health-care resources and delivery, and to social stigma (Meinardi et al., 2001; Scott et al., 2001; World Health Organization, 2000). Prior to 2000 there were few studies to investigate the prevalence of epilepsy in the People’s Republic of China (PRC) and those that were conducted showed widely divergent values for the prevalence and incidence of epilepsy (Ding et al., 2006). The differences may be due to differing study objectives and methodologies or to different economic levels or causes of epilepsy. The > lifetime prevalence of epilepsy is between three and five per 1,000 population, and the incidence of epilepsy is between 30 and 40 per 100,000 population per year and epilepsy-related mortality is between 3 and 7.9 per 100,000 population (Ding et al., 2006). In 1990, the Global Burden of Disease (GBD) study estimated the DALYs lost due to epilepsy as 0.81 per 1,000 population (Murray and Lopez, 1996). Large numbers of people with epilepsy are at risk of morbidity and mortality, mainly because of difficulties with the treatment infrastructure and the availability of suitable drugs (Scott et al., 2001). In rural China about half of patients have not been treated with standard antiepileptic treatment (Kleinman et al., 1995). Some patients have no treatment at all for their
The Disease Burden Due to Epilepsy in Rural China
92
condition, whereas others are treated inappropriately with high-cost technologies that incur unnecessarily high expenses. Mortality associated with a specific disease is assessed with particular parameters: mortality rate, case fatality rate (CFR), proportional mortality rate (PMR), and standardized mortality ratio (SMR). In China, only a few studies have described mortality rates of epilepsy in the general population, providing a range of 3.0–7.9 per 100,000 people per year. The CFR estimated from national health statistics is 0.7% of in-patient cases (Ding et al., 2006). In 1997, the World Health Organization (WHO), in cooperation with the International League against Epilepsy (ILAE) and the International Bureau for Epilepsy (IBE) launched the Global Campaign Against Epilepsy in an attempt to bring epilepsy “out of the shadows” and to improve the treatment of people with epilepsy in resource-poor countries (Kale, 1997). Demonstration projects were set up aiming to reduce the treatment gap and morbidity of people with epilepsy using community level interventions, train and educate health professionals, dispel stigma, identify potential for prevention and develop models of integration of epilepsy control into the health systems of participating countries (Li, 1989). One such demonstration project, “Epilepsy Management at Primary Health Level” (EMPHL), under the auspices of the WHO and the Ministry of Health of China, was implemented in rural areas in six non-contiguous provinces of the PRC.
2
Study Areas and Target Population
The demonstration project was implemented in rural areas in six non-contiguous provinces of the PRC, with a total population of over three million people. These areas represented different scenarios within PRC. The study sites were: Mulin and Dongning counties in Heilongjiang province; Wuzhong and Qingtongxia counties in the Ningxia Hui Autonomous Region; Wuxhi county in Henan Province; Zezhou county in Shanxi province; Hanjiang county in Jiangsu province; and Jinshan county in Shanghai Municipality (> Figure 92-1). The study areas were representative not only of the different geographical areas but also of the different economic levels in China.
3
Prevalence and Treatment Gap
The prevalence of epilepsy and the treatment gap were estimated among 66,000 of the total population of 70,000 in the six areas. 66,393 of the total population of 70,462 (94%) in these six rural areas were surveyed (Ding et al., 2004; Wang et al., 2002; Wang et al., 2003). The screening questionnaire was based on the WHO screening questionnaires previously used in PRC and on the ICBERG screening instrument (Wang et al., 2003) and validated at the Beijing Neurological Institute for specificity (78.5%) and sensitivity (100%). Participating physicians and health workers were trained to conduct the questionnaire with a standardized technique. After the questionnaire had been completed, each person who had a positive response to any of the questions was examined by a neurologist to determine the diagnosis. The minimum lifetime prevalence of epilepsy, based on the number of people identified as having a definitive history of epilepsy, was 6.8/1,000. The prevalence of active epilepsy was 4.5/1,000 (> Table 92-1). Thus, the number of people with epilepsy in PRC was estimated as almost nine million, and approximately six million people in PRC have active epilepsy.
1593
1594
92
The Disease Burden Due to Epilepsy in Rural China
. Figure 92-1 Map of China, showing the intervention sites
. Table 92-1 Prevalence of epilepsy in six study areas in rural China Life-time epilepsy Province
Active epilepsy
Cases identified
Prevalence (%)
Cases identified
Prevalence (%)
Heilongjiang
82
8.1
49
4.8
Henan
59
4.7
43
3.5
Jiangsu
87
7.8
53
4.8
Ningxia
99
8.5
78
6.7
Shanghai
65
6.0
41
3.8
60
5.8
34
3.3
452
6.8
298
4.5
Shanxi Total
This table demonstrated the prevalence of life-time epilepsy and active epilepsy in different study areas in China. The highest prevalence of epilepsy is in Ningxia
The Disease Burden Due to Epilepsy in Rural China
92
The prevalence of epilepsy is presented by sex, age group, and study area in > Table 92-2. The prevalence of epilepsy ranged from 3.5 (Shanxi) to 6.4 (Ningxia) per 1,000 population within six areas. Except in Shanxi, the prevalence of epilepsy was higher for females than for males. The age-specific epilepsy prevalence was different in six areas. In Jiangsu and Ningxia, the highest prevalence of both sexes was in the “0–4” year (11.49/1,000) and “5–14” year (8.64/ 1,000) age groups respectively. In other three areas, however, the peaks of prevalence of both sexes were in the “45–59” or “60–69” year age groups. Two fifths of people had not received any treatment for epilepsy, one third had received irregular treatment, and only a quarter had received reasonable, regular doses of AEDs in the week before the survey. Of those with active epilepsy, two thirds did not receive reasonable treatment in the week before the survey (> Table 92-3). The reason for the treatment gap is unclear, but it may be the result of the perceptions of epilepsy in PRC. If epilepsy is not viewed as a manageable condition, it is instead stigmatized, and people are unlikely to admit that they have epilepsy or to know that treatment exists. This situation has been seen across many different cultures and is probably also evident in PRC (Scott et al., 2001).
4
Disease Burden by DALY Measure
The Disability Adjusted life year (DALY) is an indicator of the time lived with a disability (Years Lived with Disability, YLD) plus the time lost due to premature mortality (Years of Life lost, YLL). The general formula and corresponding parameters of the GBD study were used to calculate the number of YLDs and YLLs lost by one individual (Murray and Lopez, 1996). YLD was calculated based on the prevalence data from the epidemiological survey. The average mortality of 6 per 100,000 population reported by Chinese literature was used for the calculation of YLLs due to epilepsy according to sex and age group in each of the study areas. The population of China in 2000 was used to adjust for the sex and age structures of the population of each study area (Ding et al., 2006). > Table 92-4 shows that epilepsy caused 1.31–1.52 YLLs per 1,000 population in the six study areas in 2000. The YLDs caused by epilepsy ranged from 0.46 to 1.01 per 1,000 population. There were 1.83 and 2.48 DALYs per 1,000 population caused by epilepsy in Henan and Ningxia, which had the lowest and highest DALY lost respectively in six areas. Overall, the YLLs lost for males were higher than those for females, while the YLDs lost for males were lower than those for females except in Shanxi province. By adjusting to sex and age structures of the population of China in 2000, epilepsy caused 1.41 YLLs and 0.67 YLDs per 1,000 population, thus the DALYs due to epilepsy was 2.07 per 1,000 population, representing the epilepsy disease burden in rural China. The distribution of disease burden within age groups is shown in > Table 92-5. The YLLs lost per 1,000 population varied from 0.03 (“0–4” year age group) to 2.53 (“15–29” year age group), whereas the YLDs lost per 1,000 population varied from 0.16 (“70+” year age group) to 0.91 (“45–59” year age group). Over 3.2 DALYs lost per 1,000 population in people with 15–44 years old, indicating that epilepsy caused large disease burden especially in young population with physical labor. In the classification of the WHO sub-regions, China belongs to the sub-region of WPRO B. The DALY lost due to epilepsy in rural China, based on the combined data of six study areas, was 2.08 per 1,000 population, which is 2.89 times higher than that of the WPRO B sub-region (0.72/1,000) and 1.73 times higher than the worldwide burden (1.20/1,000) as assessed by the GBD study (GBD, 2000).
1595
0.00
0.00
23.26
4.32
0.00
2.56
Heilongjiang
Henan
Jiangsu
Ningxia
Shanghai
Shanxi
Males in age group (years)
Females in age group (years)
3.15
2.29
8.89
2.84
0.65
1.21
1.53
2.07
8.75
2.25
2.75
2.22
1.83
3.07
5.80
5.36
2.93
6.06
9.10
4.56
0.00
4.69
4.00
5.54
10.68
2.11
3.56
5.99
0.00
12.55
0.00
4.77
0.00
2.68
3.40
0.00
3.77
3.15
6.34
4.51
2.16
4.02
2.66
6.85
5.41
0.00
6.12
0.00
2.07
2.44
8.34
3.03
4.15
5.36
2.30
4.90
5.36
5.04
2.30
6.92
2.49
3.87
9.27
4.88
6.18
5.78
5.73
6.77
9.86
10.53
9.39
12.09
10.27
9.29
0.00
8.00
3.27
4.03
0.00
3.35
5.56
1.76
2.47
4.57
3.25
5.20
7.10
5.91
5.03
6.69
3.50
3.84
6.40
5.22
3.65
5.32
5–14 15–29 30–44 45–59 60–69 70+ All ages 0–4 5–14 15–29 30–44 45–59 60–69 70+ All ages Both sexes
This table lists the age- and sex-specific prevalence of epilepsy in different study areas in China. This data is used to calculate the YLD due to epilepsy
0–4
Areas
92
. Table 92-2 The prevalence (cases per 1,000 population) due to epilepsy in six study areas in China, 2000
1596 The Disease Burden Due to Epilepsy in Rural China
The Disease Burden Due to Epilepsy in Rural China
92
. Table 92-3 Treatment gap in active epilepsy in six study areas in China, 2000 No. of patients
No. of patients receiving AEDs in previous week (%)
Treatment gap (%)
Heilongjiang
49
26(53.1)
46.9
Henan
43
16(37.2)
62.8
Jiangsu
53
12(22.6)
77.4
Ningxia
78
29(37.2)
62.8
Shanghai
41
12(29.3)
70.7
Shanxi
34
13(38.2)
61.8
298
108(36.2)
63.8
Province
Total
This table lists the percentage of AED-treated epilepsy patients and treatment gap in active epilepsy in six study areas in China
. Table 92-4 YLL, YLD and DALY lost per 1,000 population by sex due to epilepsy in six study areas in China YLL/1,000
YLD/1,000
DALY/1,000
Areas
Male
Female
Total
Male
Female
Total
Male
Female
Total
Heilongjiang
1.64
1.38
1.52
0.60
1.03
0.81
2.24
2.41
2.33
Henan
1.39
1.22
1.31
0.31
0.74
0.52
1.70
1.96
1.83
Jiangsu
1.51
1.19
1.35
0.61
0.96
0.79
2.12
2.15
2.14
Ningxia
1.53
1.42
1.47
1.00
1.01
1.01
2.53
2.43
2.48
Shanghai
1.57
1.22
1.39
0.42
0.71
0.57
1.99
1.93
1.96
Shanxi
1.49
1.29
1.39
0.49
0.44
0.46
1.98
1.73
1.85
Totala
1.52
1.30
1.41
0.55
0.81
0.67
2.07
2.11
2.08
This table lists the sex-specific YLL, YLD, and DALY lost per 1,000 population due to epilepsy in six study areas in China. Henan and Ningxia have the lowest and highest DALY lost respectively in six areas a Sum of six study areas adjusted by the sex and age structure of China population in 2000
5
Applications of the Model to Other Diseases
For the complete assessment of the epilepsy disease burden, DALY measurements should be acquired not only for neurological diseases, but also for all the diseases in different areas of China. Unfortunately, this has not been done. Compared to the disease burden due to neurologic diseases in WPRO B sub-region provided by GBD study, however, the result of 2.08 DALYs lost per 1,000 population is lower than that of Cerebrovascular disease (9.86 DALYs lost per 1,000 population), and higher than that of Alzheimer disease and other dementia (1.76 DALYs lost per 1,000 population) or Pakinson’s disease (0.21 DALYs lost per 1,000 population) (GBD, 2000).
1597
1598
92
The Disease Burden Due to Epilepsy in Rural China
. Table 92-5 YLL, YLD and DALY lost per 1,000 population by age group due to epilepsy in the entire population of the six study areas in China YLL/1,000 Age group (years)
Male
0–4
YLD/1,000
DALY/1,000
Female
Total
Male
Female
Total
Male
Female
Total
0
0.06
0.03
0.24
0.12
0.18
0.24
0.18
0.21
5–14
0.57
0.33
0.46
0.40
0.57
0.48
0.98
0.90
0.94
15–29
2.42
2.64
2.53
0.59
0.82
0.70
3.01
3.46
3.23
30–44
2.88
1.93
2.40
0.70
0.93
0.82
3.58
2.86
3.22
45–59
0.67
0.60
0.64
0.60
1.24
0.91
1.27
1.84
1.55
60–69
0.17
0.29
0.23
0.62
0.69
0.66
0.80
0.99
0.89
70+
0.12
0.34
0.24
0.10
0.21
0.16
0.21
0.55
0.41
1.52
1.30
1.41
0.55
0.81
0.67
2.07
2.11
2.08
a
Total
This table lists the adjusted YLL, YLD, and DALY lost per 1,000 population due to epilepsy. Epilepsy caused large disease burden in young population a Sum of six study areas adjusted by the sex and age structure of china population in 2000
6
Premature Mortality in People with Epilepsy
People with epilepsy have increased risk of premature death compared with the general population (Gaitatzis and Sander, 2004). However, the extent and nature of this risk has not been sufficiently examined, especially in resource-poor countries where there is often a large treatment gap with many patients not being treated (Cockerell et al., 1994). The EMPHL protocol used strict follow-up and management procedures, and so it was possible for us to calculate PMRs and SMRs to assess mortality in people with epilepsy. During the 1-year follow-up to 2,455 epilepsy patients, local primary-care physicians recorded demographic data and putative cause of death of any patient who died. Cause of death was attributed on clinical grounds and verbal autopsy. Specialists and the principal investigators in each study area gathered information about cause of death through interviews with relatives or local village physicians. Death certificates were also used for confirmation of the cause of death (Ding et al., 2006). There were 35 deaths among 2,455 people with epilepsy during follow-up. The main cause of death was accidental or as a result of injury (self-inflicted or otherwise); 13 (37%) patients died of drowning (n = 6), suicide (n = 4), poisoning (n = 2), or a road traffic accident (n = 1). In 11 (31%) patients, death was attributed to hemorrhagic or ischemic stroke, whereas in two (6%), death was attributed to pneumonia. The age-adjusted PMRs for injury (30%), stroke (30%), pneumonia (5%), and myocardial infarction (6%) were higher for the trial participants than for the 2004 Chinese population (8, 19, 2, and 3%, respectively). The adjusted PMR for neoplasm (15%) was lower in the study population than in the general population (> Figure 92-2). The SMR for people with epilepsy was 3.9 (95% CI 3.8–3.9) and was higher in women (4.1; 3.9–4.4) than in men (3.5; 3.4–3.6). > Figure 92-2 shows that in patients aged at least 15 years, and with the exception of those aged 70–74 years, the SMR was raised. SMRs were especially high in patients ages 15–19 years (23.3; 15.9–43.9), 20–24 years (40.2; 32.4–52.8), and 25–29
The Disease Burden Due to Epilepsy in Rural China
92
. Figure 92-2 Age-adjusted cause-specific PMRs in people with epilepsy compared with the general Chinese population, 2004
. Figure 92-3 Age-specific SMRs of epilepsy in the study population in rural China, adjusted for the Chinese population, 2004. Bars represents 95% CIs
years (33.3; 28.0–41.1). The SMRs for patients aged 30–49 years were in the range 2.3–8.1, and for people with epilepsy aged older than 49 years the SMRs were lower than for the 30–49 year age-group (> Figure 92-3). Cause-specific SMRs of the epilepsy population were calculated for: pneumonia (21.3; 14.5–40.0), injury (12.2; 11.4–13.0), myocardial infarction (10.7;
1599
1600
92
The Disease Burden Due to Epilepsy in Rural China
5.6–95.3), stroke (7.0; 6.5–7.6), and neoplasm (1.6; 1.5–1.8). Suicide (3.8; 3.2–4.6) and drowning (5.6; 4.7–6.5) were the major causes of death in those classified as dying of injury.
7
Conclusion
According to our measurement, there was an overall size of burden of 1,681,410 DALYs lost in rural China based on the rural population of 808,370,000 in 2000. There is an increased risk of premature mortality in people with epilepsy in rural China, especially among the young. The increase in mortality in young adults is higher than seen in developed countries. The information of disease burden can help policy planners to allocate resources and identify strategies and interventions for the reduction of the burden of epilepsy in China.
Summary Points The minimum lifetime prevalence of epilepsy was 6.8/1,000, and the prevalence of active epilepsy was 4.5/1,000. Thus, the number of people with epilepsy in PRC was estimated as almost nine million, and approximately six million people in PRC have active epilepsy. Two fifths of people had not received any treatment for epilepsy, one third had received irregular treatment, and only a quarter had received reasonable, regular doses of AEDs in the week before the survey. Of those with active epilepsy, two thirds did not receive reasonable treatment in the week before the survey. Epilepsy caused 1.41 YLLs and 0.67 YLDs per 1,000 population, thus the DALYs due to epilepsy was 2.08 per 1,000 population, representing the epilepsy disease burden in rural China. This is almost three times higher than that of the WPRO B sub-region and twice that of the worldwide burden as assessed by the GBD study. The main cause of death of epilepsy patients was accidental or as a result of injury. SMRs were especially high in patients ages 15–19 years (23.3), 20–24 years (40.2), and 25–29 years (33.3). There is an increased risk of premature mortality in young people with epilepsy in rural China.
References Cockerell OC, Johnson AL, Sander JW, Hart YM, Goodridge DM, Shorvon SD. (1994). Lancet. 344: 918–921. Ding D, Hong Z, Wang WZ, Wu JZ, de Boer HM, Prilipko L, Sander JW. (2006). Epilepsia. 47: 2032–2037. Ding D, Lu GY, Huang MS, Zhou B, Hou GQ, Zeng J, Jin MH, Zhu GX, Wang JY, Fu JH, Meng HJ, Zhao QH, Hong Z. (2004). Chin J Neurosci. 12: 122–124 (Chinese). Ding D, Wang WZ, Wu JZ, Ma GY, Dai XY, Yang B, Wang TP, Yuan CL, Hong Z, de Boer HM, Prilipko L, Sander JW. (2006). Lancet Neurol. 5: 823–827.
Gaitatzis A, Sander JW. (2004). Epileptic Disord. 6: 3–13. GBD 2000 (2000). Available from: URL: http://www3. who.int/whosis Accessed 2004. Kale R. (1997). BMJ. 315: 2–3. Kleinman A, Wang WZ, Li SC, Cheng XM, Dai XY, Li KT, Kleinman J. (1995). Soc Sci Med. 40: 1319–1330. Li SC. (1989). Chin J Nerv Ment Dis. 22: 144 (Chinese). Meinardi H, Scott RA, Reis R, Sander JW. (2001). Epilepsia. 42: 136–149. Murray CJL, Lopez AD. (1996). Global Burden of Disease and Injury Series. World Health Organization, Geneva. Scott RA, Lhatoo SD, Sander JW. (2001). Bull World Health Organ. 79: 344–351.
The Disease Burden Due to Epilepsy in Rural China Wang WZ, Wu JZ, Wang DS, Chen GS, Wang TP, Yuan CL, Yang B, Zhao DH. (2002). Natl Med J China. 82: 449–452 (Chinese). Wang WZ, Wu JZ, Wang DS, Dai XY, Yang B, Wang TP, Yuan CL, Scott RA, Prilipko L, de Boer HM, Sander JW. (2003). Neurology. 60: 1544–1545.
92
World Health Organization. (2000). The Global Campaign Against Epilepsy – Out of the Shadows. WHO, Geneva.
1601
93 Alcohol Consumption and Burden of Disease: Germany and Switzerland M. Roerecke . J. Rehm . J. Patra 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1604
2
Data Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1605
3
Exposure Measurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1606
4
Risk Relations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1606
5
Health Outcomes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1607
6
Population Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1607
7
Calculation of Attributable Fractions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1608
8
Alcohol Consumption in Germany and Switzerland . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1608
9
Alcohol-Attributable Deaths in Germany and Switzerland . . . . . . . . . . . . . . . . . . . . . . . 1609
10 Alcohol-Attributable Years of Life Lost in Germany and Switzerland . . . . . . . . . . . . 1609 11 Alcohol-Attributable Disability Adjusted Life Years in Germany and Switzerland . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1609 12 Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1613 13 Burden of Disease in Germany and Switzerland . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1613 14 Implications for Policy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1614 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1616
#
Springer Science+Business Media LLC 2010 (USA)
1604
93
Alcohol Consumption and Burden of Disease: Germany and Switzerland
Abstract: We estimated the burden of disease attributable to alcohol in Germany and Switzerland for the year 2002. We calculated disease-specific attributable fractions for diseases and conditions based on relative risk estimates from meta-analyzes, prevalence of exposure from the Global Alcohol Database and large, representative surveys, and obtained > mortality and burden of disease data from the World Health Organization. Comparatively high alcohol consumption in Switzerland caused substantial burden, in particular among Swiss women. Although alcohol consumption was high among women in Germany and Europe as well, the burden among women was much less compared to Switzerland. For men in both countries comparable burden of disease was estimated. Both detrimental and beneficial effects of alcohol consumption were considered in this analysis. Because effective and cost-effective interventions are available, the burden of disease due to alcohol should be a focus of public health policy development. Specific measures are discussed. List of Abbreviations: > DALYs, disability-adjusted life years, a health gap measure combining years of life lost and years lived with a disability, therefore taking into account mortality and morbidity; > YLL, years of life lost, a health gap measure taking into account the time of deaths relative to the normative life expectancy; Mortality, number of deaths; CRA 2000, comparative risk analysis study for the year 2000, conducted as part of the global burden of disease study by the World Health Organization; > AF, attributable fraction, the fraction of disease outcome, for example, that is caused by a specific risk factor, in this case alcohol; > WHO, World Health Organization; > GBD, global burden of disease study conducted by the World Health Organization describing the effect of 26 risk factors on burden of disease around the world; > GAD, global alcohol database, maintained by the WHO. Alcohol consumption indicators are updated annually for each member state based on data from multiple sources; Europe A, WHO subregion defined by very low child and very low adult mortality; APC, estimated adult (15 years and above) per capita alcohol consumption (in liters of pure alcohol consumed within one year); HDL, high density lipoprotein
1
Introduction
Alcohol is a cause for substantial burden of disease worldwide, ranking fifth of 26 risk factors examined by the World Health Organization (WHO) in 2000 (Ezzati et al., 2002). Four percent of the global burden of disease in 2000 was estimated to be caused by alcohol consumption; in developed countries, the contribution of alcohol is even higher with 9.2% of all disease burden attributed to alcohol, just behind tobacco and high blood pressure. However, the burden varies greatly by region and country with poorer countries and regions showing less burden attributable to alcohol than developed countries. In Europe in 2002, due to comparatively high consumption (11.9 liters per adult capita, compared to 6.2 liters worldwide) the burden is significantly higher with considerable heterogeneity across countries (Rehm et al., 2006b). For example, an East West gradient within Europe has been described recently (Popova et al., 2007; Rehm et al., 2007a,b). On a global level, most people abstain from alcohol and most recent estimates showed strong sex differences with 6.1% of all deaths worldwide attributable to alcohol among men and 1.1% among women in 2002. Among the population 60 years and older 5% of all deaths were attributable to alcohol consumption (Rehm et al., 2006b).
Alcohol Consumption and Burden of Disease: Germany and Switzerland
93
The relationship between alcohol and disease and conditions is complex and several dimensions of consumption have to be considered. We will concentrate here on direct health effects (burden of disease) and not consider social and indirect effects on families and communities. Overall, the net costs attributable to alcohol (including indirect health care costs, law enforcement and social harm job loss, marginalization, etc.) outweigh any benefits because of the strong impact of social costs. For most diseases the effect is detrimental with few exceptions, such as coronary heart disease or diabetes when average consumption is low to moderate. Most researchers to date agree on a cardioprotective effect of regular low to moderate average alcohol consumption in comparison to heavy drinkers and abstainers (Rehm et al., 2003). Rimm and colleagues investigated biological mechanisms and summarized the evidence from animal and experimental studies regarding mediating pathways. Most of the mediating beneficial effect of regular low alcohol consumption is caused by an increase in high density lipoproteins (HDL), in addition to inhibiting platelet activation, lower levels of fibrinogen, and antiinflammatory effects (Rimm et al., 1999; Rimm and Moats, 2007). It has been estimated that a risk reduction of 20–30% can be assumed for low to moderate average consumption (Corrao et al., 2000). In some developed countries the beneficial effect of alcohol on cardiovascular diseases can be substantial and, mostly among women, result in a net beneficial effect. Some diseases are caused by alcohol by definition, such as alcohol dependence or alcoholic liver cirrhosis. These conditions have by definition an attributable fraction of 100%. While measurement error in disease classification cannot be ruled out, the contribution of alcohol does not need to be estimated. For chronic diseases, such as cancer or cardiovascular diseases, alcohol is a contributory cause rather than a sufficient cause. While this presents difficulties in treatment and causal inference for an individual case, on a population level, a certain fraction of cancer cases can be attributed to alcohol by calculating attributable fractions (AF) derived from prevalence estimates of alcohol exposure and relative risk estimates from large scale epidemiological studies quantifying risk relations for diseases where alcohol has been found to be a causal factor in disease etiology. This chapter will compare alcohol consumption and attributable burden in Germany and Switzerland for the year 2002 using a standardized methodology developed for the Comparative Risk Analysis 2000 (CRA 2000) within the Global Burden of Disease study conducted by the WHO (Rehm et al., 2004). The CRA 2000 and recent updates were undertaken in an effort to compare among other risk factors alcohol consumption and resulting burden of disease across regions and countries using the same standardized methodology.
2
Data Sources
In order to calculate the burden of disease attributable to a specific risk factor, several data sources need to be evaluated: disease specific relative risk estimates usually derived from comprehensive meta-analyzes, data for prevalence of exposure from official sources in addition to survey data, and an estimate of the total number of deaths, years of life lost (YLL) and disability adjusted life years (DALYs) per disease condition in the respective population. Preferably, these estimates should be sex- and age specific. A disease specific approach is necessary because the burden of disease caused by alcohol depends on the distribution of consumption and deaths in the population (for a detailed description of the methodology, please see Rehm et al., 2004; Rehm et al., 2007a,b).
1605
1606
93 3
Alcohol Consumption and Burden of Disease: Germany and Switzerland
Exposure Measurement
Information of exposure to alcohol for a population is not readily available and several data sources have to be considered, such as official sales statistics, tax revenue statistics, production data, and survey data. Alcohol consumption of the population 15 and over is usually used to determine the overall consumption of a population because alcohol consumption typically takes place in adulthood and the proportion of people under the age of 15 varies by country, therefore using per capita consumption estimates (total population) would result in biased consumption estimates for the actual consuming population. We took adult per capita consumption for Germany and Switzerland from the WHO Global Alcohol Database (WHO, 2006). Although these data represent the most valid estimates for overall consumption in a population (Gmel and Rehm, 2004), they have to be combined with survey data in order to derive sex- and age-specific estimates of the distribution of alcohol consumption in each country. Unrecorded consumption, i.e., consumption not recorded by official sales statistics, such as home production, illegal import or production, is commonly low in Germany and Switzerland (1 L or less). Information on the distribution of alcohol consumption was taken from the Swiss Health Survey [Schweizerische Gesundheitsbefragung], a random telephone survey of the Swiss population 15 years and older living in Switzerland. Response rate for the third instalment of the Swiss survey, conducted in 2002 was 64% at the household level (Bundesamt fu¨r Statistik, 2004). Data for Germany were taken from the 2003 Epidemiological Survey of Substance Abuse [Epidemiologischer Suchtsurvey 2003], a two-stage random survey of about 8,000 residents of Germany between 18–59 years. Response rate for the selfadministered questionnaire was 55% (Augustin and Kraus, 2005). We classified drinking levels into categories based on English et al. (1995) a classification widely used in meta-analyzes examining the effect of alcohol consumption (> Table 93-1). Both countries were assigned a . Table 93-1 Definition of drinking categories based on English et al. (1995) Abstainer or very light drinker (g/day)
Drinking Drinking category Drinking category category I (g/day) II (g/day) III (g/day)
Men
0 – < 2.5
2.5 – < 40
40 – < 60
60+
Women
0 – < 2.5
2.5 – < 20
20 – < 40
40+
Tthe limits of these categories are stated in grams of pure alcohol per day. For reference, a bottle of table wine contains about 70 g of ethanol; 2.5 g/day corresponds to somewhat more than one bottle of wine per month
pattern score of one, the least detrimental score (range 1–4). The score for pattern of drinking is composed of indicators for hazardous drinking, such as drinking outside of meals, frequency of drinking in public places, and frequency of heavy drinking occasions (Rehm et al., 2004). Pattern scores were available only on the country level.
4
Risk Relations
More than 60 diseases and conditions have been found to be causally related to alcohol, including cirrhosis of the liver, several types of cancer (mouth and oropharynx, esophagus,
Alcohol Consumption and Burden of Disease: Germany and Switzerland
93
liver, breast in women, colorectal, and others), hypertension, ischaemic heart disease, ischaemic and hemorrhagic stroke, diabetes, and intentional and unintentional injuries (Boffetta and Hashibe, 2006; Cho et al., Corrao et al., 2000; Corrao et al., 2004; English et al., 1995; Gutjahr et al., 2001, 2004; Reynolds et al., 2003). These conditions have been identified in several meta-analyzes based on epidemiological criteria (Hill, 1965; Rothman and Greenland, 1998) with an emphasis on epidemiological criteria to determine causality, such as consistency across studies, temporal order, plausible biological mechanisms, strength of the effect, and dose-response relationships. Beneficial and detrimental effects were assessed using the same criteria. For acute outcomes, such as injuries, attributable fractions were taken directly from the CRA 2000 study. The most recent addition to alcohol attributable diseases is colorectal cancer for both sexes (Cho et al., 2004). There is now sufficient evidence that alcohol plays a causal role in the etiology of colorectal cancer and therefore we included this condition in our calculations.
5
Health Outcomes
Three health outcome measures summarizing the health status of a population will be used in this analysis. First, mortality (i.e., the number of deaths in a population), secondly, years of life lost (YLL) as one of two health gap measures taking into account not only the number of deaths but also the age at which the death occurs, therefore giving a higher weight to deaths at a younger age. Thirdly, as a second health gap measure, disability adjusted years of life (DALYs), a measure that combines YLL and the years lived with a given disability, hence taking into account morbidity in addition to mortality. The residual life expectancy (the difference between the normative standard and the actual time of death) for calculation of YLL and DALYs was 80 years for men and 82.5 years for women (based on life expectancy in Japan). Both DALYs and YLL were 3% discounted and age-weighted for comparison to the 2000 CRA study. Because Germany and Switzerland have not undertaken a burden of disease study yet, disease-specific estimates for DALYs, YLL, and number of deaths were obtained from WHO headquarters (Dr. Colin Mathers); diseases were categorized based on the 2000 Global Burden of Disease study.
6
Population Data
Population data were obtained from UN (United Nations, 2005) and classified into seven age groups: 0–4 years, 5–14 years, 15–29 years, 30–44 years, 45–59 years, 60–69 years, and 70 years and over. Respective population surveys on alcohol consumption from each country were used in calculating sex- and age-specific prevalence rates to derive attributable fractions for each disease and age group. Because surveys generally underestimate real consumption in a population (Midanik, 1982; Rehm, 1998), estimates form these surveys had to be adjusted to reflect total adult per capita alcohol consumption, a more valid indicator of overall consumption, as reported in the WHO Global Alcohol Database (please see Rehm et al., 2007a,b for details of this procedure). Drinking categories were based on English et al. (1995), a classification widely used in alcohol epidemiology.
1607
1608
93 7
Alcohol Consumption and Burden of Disease: Germany and Switzerland
Calculation of Attributable Fractions
Sex and, where possible, age-specific relative risk estimates from meta-analyzes are then combined with prevalence data for alcohol consumption to calculate the fraction attributable to alcohol for each disease category (Walter, 1976, 1980). The result is the proportion of a disease that would not have occurred if there were no exposure to alcohol, i.e., the whole population is abstinent. Since the effect of alcohol consumption can beneficial or detrimental, the attributable fraction can be positive or negative. AF ¼
k X
Pi ðRRi 1Þ=
k X
i¼1
Pi ðRRi 1Þ þ 1
i¼0
where i: exposure category with baseline exposure or no exposure i = 0, RR(i): relative risk at exposure level i compared to no consumption, P(i): prevalence of the ith category of exposure. Age and sex-specific AFs were then applied to mortality and burden of disease data to derive the number of deaths, YLL, and DALYs attributable to alcohol by sex and age group.
8
Alcohol Consumption in Germany and Switzerland
Germany and Switzerland both are part of the WHO subregion Europe A, with very low childhood and adult mortality. With an adult population of about 70 million, Germany represents about 20% of the total adult population in Europe A, while Switzerland has an adult population of about 6 million (> Table 93-2). Overall, the adult population (15 years and over) in Switzerland consumed 11.4 L (including 0.5 L unrecorded consumption) of pure alcohol in 2002, the corresponding figure for Germany was 13.2 L (including 1 L unrecorded consumption) according to the Global Alcohol Database. In both countries, survey data accounted for about half of the adult per capita consumption, therefore making the adjustments described above necessary. . Table 93-2 Characteristics of alcohol consumption in Switzerland, Germany and Europe A in 2002 Percent of abstainers Country or Adult region population*
M
W
Switzerland
Alcohol Unrecorded Pattern Csdconsumption{ consumption value
Recorded beverage most consumed
5,969
15
34
11.4
0.5
1
Wine (51%)
Germany
70,042
7
9
13.2
1.0
1
Beer (58%)
Europe A
347,001
11
23
12.1
1.3
1.5
Beer (41%) and wine (40%)
This table summarizes the characteristics of alcohol consumption among the adult population in Switzerland, Germany and WHO subregion Europe A * Numbers in thousands { Adult per capita (age 15+) consumption for 2002 in liters of pure alcohol, derived as average of yearly consumptions from 2001 to 2003, including unrecorded consumption
Alcohol Consumption and Burden of Disease: Germany and Switzerland
93
Among Swiss women, a total of 34.0% were abstinent in 2002, and 15.4% among men. Corresponding figures for Germany were 7.0% among men and 9.2% among women. Despite higher abstention rates for both sexes in Switzerland, a higher proportion of the population reported consuming amounts within drinking categories II and III compared to Germany. About half of German men and slightly more than 60% of women in Germany were classified in drinking category I where a cardioprotective effect can be assumed. Only slightly less than 30% of men and women in Switzerland drank within this category.
9
Alcohol-Attributable Deaths in Germany and Switzerland
In both countries and WHO subregion Europe A, proportionally many more men died because of alcohol consumption than women (> Table 93-3). Given the alcohol consumption levels for men are much higher, this is not surprising. Among men, the proportion of alcoholattributable deaths in Switzerland was much higher compared to Europe A, whereas Germany had a slightly higher proportion of deaths caused by alcohol compared to Europe A. These proportions represent net effects, i.e., detrimental and beneficial effects were considered. While the most prevalent categories (cardiovascular diseases, cancer and cirrhosis of the liver) were the same for Europe A, Switzerland and Germany, the magnitude and direction of the overall effect (i.e., beneficial or detrimental) was markedly different. Compared to Europe A and Germany, the beneficial effect of alcohol consumption on cardiovascular diseases in Switzerland was much lower and resulted in an overall detrimental effect of alcohol consumption for women in Switzerland. Contrary, in Europe A and Germany many more deaths were prevented due to alcohol than caused among women.
10
Alcohol-Attributable Years of Life Lost in Germany and Switzerland
When considering years of life lost (YLL, > Table 93-4), injury categories typically show higher influence on the overall effect of alcohol on health outcomes because intentional and unintentional injuries occur earlier in life, especially among young men, whereas chronic diseases generally occur later in life. As seen with mortality as outcome measure, YLL among women in Switzerland (4.9%) were markedly higher compared to either Germany (4.0%) or Europe A (1.2%), largely due to the cardioprotective effect of low to moderate alcohol consumption. Among men, in all countries and region displayed in > Table 93-3 a substantial detrimental overall effect was estimated. Slightly more than ten percent of all YLL were caused by alcohol in Germany and Switzerland and slightly less than ten percent in Europe A.
11
Alcohol-Attributable Disability Adjusted Life Years in Germany and Switzerland
With regard to DALYs (> Table 93-5), neuropsychiatric disorders play a more important role compared to YLL or deaths because these disorders, such as alcohol use disorders, are less fatal and more related to morbidity than mortality and usually involve multiple episodes and relapses over the lifetime. Among men in Switzerland, slightly more than 50% of all
1609
162
5.2%
30.345
1.4%
30.574
441
48
90
191
92.5
0.0
W
25.7
9
4,661
857
7,231
M
10.9
20.3
43.2
100.0 100.0
10.3
20.2
27.7
3.8%
385,388
14,755
1,465
3,694
9,477
12.4 81.8 10,925
15.4
13.2 10.7
52.0
0.0
M 7
1,288
930
5,026
W
Germany
4.3%
430,013
18,622
422
1,196
3,597
29,227
no.
6,824
100.0 100.0
48,114
1,898
7,840
13,012
94,527
3,953
3,108
22,793
25
W
Europe A
2.4%
1,970,000
no.
3.5%
1,950,000
67,772
25,594
6.4 2.3
9.9
25.0
31,850
64.2 19.3
12,912
6.9
42,962
31.6 74.0 156.9
3,290
5.0
36,809
5.8
49.0 27.0
M 34
W 0.0
0.1
M
%
0.1
W
8.2
6.5
3.9 100.0 100.0
10.1
37.8 16.3
47.0 27.0
63.4 196.5
19.1
4.9
54.3 47.4
0.0
M
%
This table displays the burden of disease by disease category in terms of number of deaths caused by alcohol consumption. Beneficial and detrimental effects have been considered in these estimates. Data taken from Rehm et al. (2006b) a The WHO (World Health Organization) subregion Europe A denotes the part of Europe with very low childhood and very low adult mortality. It comprises the following countries: Andorra, Austria, Belgium, Croatia, Czech Republic, Denmark, Finland, France, Germany, Greece, Iceland, Ireland, Israel, Italy, Luxembourg, Malta, Monaco, Netherlands, Norway, Portugal, San Marino, Slovenia, Spain, Sweden, Switzerland, and United Kingdom
Percentage of all deaths attributable to alcohol
All deaths
1.575
318
Unintentional injuries
Intentional injuries
All alcohol-attributable deaths
436
113
361
243
408
47
818
207
195
0
0
Cirrhosis of the liver
Cardiovascular diseases
Neuropsychiatric disorders
Diabetes mellitus
Cancer
Maternal and perinatal conditions (low birth weight)
W
%
93
M
no.
Switzerland
. Table 93-3 Deaths attributable to alcohol consumption in Switzerland, Germany and Europe Aa in 2002
1610 Alcohol Consumption and Burden of Disease: Germany and Switzerland
21,871
All alcoholattributable YLLs
10.5%
2,393
4.9%
144,394
7,067
911
1,212
20.8
5.5
61.3
0.0
W
33.9
12.9
17.2
100.0 100.0
14.9
27.3
24.0
8.8 40.5
14.0
7.7
36.3
0.0
M
%
10.1%
2,909,683
292,950
29,055
78,508
126,908
85,257
71,410
9,312
81,346
292
M
18,859
7,455
56,232
229
W
Germany
4.0%
2095,351
84,150
7,746
17,433
49,544
226,738
no.
1,114
406,401
0.3
27.8 66.8
528,526 144,790
9.2
100.0 100.0
60,077
25,453
258,511
829
W
Europe A
1.2%
9,830,000
121,971
37,458
108,532
173,927
735,853
no.
9.5%
14,133,000
1,335,964
427,765
43.3 58.9 26.8 20.7 9.9
334,978
199,884
24.4 22.4 29.1 269.4
37,538
8.9
3.2
0.1
M
W
M
%
0.7
W
603.3
49.3
20.9
100.0
10.8
39.6
100.0
30.7
89.0
32.0 142.6
25.1
15.0
2.8
30.4 211.9
0.1
M
%
This table displays the burden of disease by disease category in terms of years of life lost (YLL) caused by alcohol consumption. Beneficial and detrimental effects have been considered in these estimates. Data taken from Rehm et al. (2006b) a The WHO region Europe A denotes the part of Europe with very low childhood and very low adult mortality. It comprises the following countries: Andorra, Austria, Belgium, Croatia, Czech Republic, Denmark, Finland, France, Germany, Greece, Iceland, Ireland, Israel, Italy, Luxembourg, Malta, Monaco, Netherlands, Norway, Portugal, San Marino, Slovenia, Spain, Sweden, Switzerland, and United Kingdom
Percentage of all YLLs attributable to alcohol
207,539
3,270
Intentional injuries
All YLLs
5,246
5,967
2,865
1,924
Cardiovascular diseases
Unintentional injuries
1,472
3,059
Neuropsychiatric disorders
Cirrhosis of the liver
4,333
392
7,938
3
1,694
9
W
Diabetes mellitus
Cancer
Maternal and perinatal conditions (low birth weight)
M
no.
Switzerland
. Table 93-4 Years of life lost (YLL) attributable to alcohol consumption in Switzerland, Germany and Europe Aa in 2002 Alcohol Consumption and Burden of Disease: Germany and Switzerland
93 1611
6,288
7,420
3,416
54,399
Cirrhosis of the liver
Unintentional injuries
Intentional injuries
All alcoholattributable DALYs
12.9%
4.2%
378,056
15,857
976
1,791
%
56.0
3.7
30.4
0.1
W
6.2
11.3
19.1
100.0 100.0
6.3
13.6
11.6
2.6 19.3
61.2
5.4
15.3
0.0
M
115,043
11,120
63,063
324
W
0.2%
4,910,882
11,445
8,306
23,861
64,589
252,620
no.
13.4%
5,480,607
732,207
30,456
94,995
154,317
81,598
468,187
20,201
85,660
392
M
%
2.8
1005.2
97.2
551.0
W
100.0
4.2
13.0
21.1
100.0
72.6
208.5
564.3
11.1 2,207.3
63.9
2.8
11.7
0.1
M
343,956
39,953
152,646
225,561
828,914
509,204
45,281
289,368
1,419
W
1.4%
24,396,000
no.
12.2%
27,329,000
3,334,298
151,130
643,717
520,850
313,038
2,008,786
104,457
425,555
1,755
M
Europe A %
148.0
13.2
84.1
0.4
W
100.0
4.5
19.3
15.6
100.0
11.6
44.4
65.6
9.4 241.0
60.2
3.1
12.8
0.1
M
This table displays the burden of disease by disease category in terms of disability adjusted life years (DALYs) caused by alcohol consumption. Beneficial and detrimental effects have been considered in these estimates. Data taken from Rehm et al. (2006b) a The WHO region Europe A denotes the part of Europe with very low childhood and very low adult mortality. It comprises the following countries: Andorra, Austria, Belgium, Croatia, Czech Republic, Denmark, Finland, France, Germany, Greece, Iceland, Ireland, Israel, Italy, Luxembourg, Malta, Monaco, Netherlands, Norway, Portugal, San Marino, Slovenia, Spain, Sweden, Switzerland, and United Kingdom
Percentage of all DALYs attributable to alcohol
420,560
3,053
1,417
Cardiovascular diseases
All DALYs
8,886
33,271
Neuropsychiatric disorders
3,021
4,816
591
8,331
2,929
10
W
Diabetes mellitus
Cancer
19
no.
Germany
93
Maternal and perinatal conditions (low birth weight)
M
Switzerland
. Table 93-5 Disability adjusted life years (DALYs) attributable to alcohol consumption in Switzerland, Germany and Europe Aa in 2002
1612 Alcohol Consumption and Burden of Disease: Germany and Switzerland
Alcohol Consumption and Burden of Disease: Germany and Switzerland
93
alcohol-attributable DALYs were neuropsychiatric disorders, followed by cancer and cirrhosis of the liver. About 13% of all DALYs among men in both countries were attributable to alcohol, slightly less in Europe A (12.2%). Among women, 4.2% were attributable to alcohol in Switzerland, 1.4% in Europe A, and 0.2% in Germany.
12
Limitations
Although we feel that the estimates presented in this study represent the best available, we have to stress that certain assumptions were made in deriving these estimates. While the standardized procedures in the CRA methodology in combination with usage of official statistics and other data sources are certainly strengths of this research, limitations include measurement error in the aforementioned official sales statistics and we assume that this error is nondifferential across countries. Measurement error most likely is strongest for unrecorded consumption as it is not part of official statistics. However, in the case of Germany and Switzerland the proportion of unrecorded consumption is low (less than 1 L). Furthermore, relative risk estimates derived from meta-analyzes might not be the same across countries. This is most important for injuries, as there is some evidence that these estimates can be influenced considerably by cultural differences. However, because most studies reported in meta-analyzes have been conducted in English speaking countries or European countries, this should be acceptable for Switzerland and Germany because these risks are mainly the result of biochemical effects of alcohol. Certainly, especially in the case of beneficial effects on cardiovascular diseases, age-specific estimates would improve the accuracy of this analysis; however, those estimates are not yet available. In addition, pattern scores for drinking were available only at country level. Thus, we assume that the drinking pattern is homogeneous within a country. These patterns play a role not only in risk of injuries as a result of accidents, but also in the occurrence of ischaemic heart disease. There is evidence that in age groups older than 70 years relative risk estimates converge to one, i.e., the effect of alcohol might be overestimated (Rehm et al., 2006a). This overestimation pertains to beneficial and detrimental effects. In addition, limitations generally applying to all survey data, such as that the findings mirror the true distribution of exposure in a population. In order to address this issue, we adjusted survey estimates for adult per capita consumption, the most valid estimate of the true consumption in a population (Gmel and Rehm, 2004).
13
Burden of Disease in Germany and Switzerland
Despite the fact that alcohol is a commodity in many countries, resulting harm should not be taken for granted. In light of the substantial direct burden of disease in both countries examined in this analysis, implementation of interventions should be prioritized in both countries. One should keep in mind that the harm to society is much greater than just the direct health burden estimated in this analysis. Alcohol can cause much harm besides the direct effect on the person who is drinking. Almost no pattern of drinking is risk-free and risks and benefits have to be weighed against each other. Overall, the effect of alcohol on burden of disease worldwide is overwhelmingly detrimental. The only exception on a regional level is the WHO subregion Europe A, where, among women, a net beneficial effect was estimated when considering the number of deaths or YLL. For DALYs, the effect turned slightly negative even
1613
1614
93
Alcohol Consumption and Burden of Disease: Germany and Switzerland
in this subregion. Generally, the magnitude of the effect of alcohol increases when time-based gap measures, such as YLL and DALYs are considered. This is a result of the tremendous burden due to injuries, which mainly occur earlier in life compared to chronic diseases outcomes. Men drank considerably more on average than women in both countries, a trend that is observable around the world. This higher consumption in turn lead to a less pronounced cardioprotective effect and much higher disease burden due to alcohol in general compared to women in both countries. The burden among men in both Switzerland and Germany was substantial, about 13% of all DALYs among men were attributable to alcohol. Neuropsychiatric disorders were the main cause of alcohol-attributable DALYs in both countries. Although trends in both countries show slightly decreasing consumption overall, drinking among youth in both Germany and Switzerland seems to increase recently (Mohler-Kuo et al., 2004; Augustin and Kraus, 2005). Although abstention rates were much lower in Germany compared to Switzerland, the proportion of the population drinking at higher levels (drinking categories II and III) was higher in Switzerland. A cardioprotective effect is restricted to drinking category I. While the majority of the German population drank within this category, most people in Switzerland drank at higher levels, where consequently no protective effect was estimated. The difference in the distribution of drinking levels was most pronounced among women when comparing Germany and Switzerland. Whereas slightly more than 60% drank moderately in Germany, only slightly less than 30% of Swiss women drank at comparable low levels. Thus, although adult per capita alcohol consumption was higher in Germany (13.2 L vs. 11.4 L), the burden of disease attributable to alcohol in both countries was comparable among men, but much higher among women in Switzerland. Differences between Switzerland, Germany and Europe A among women were strongest when considering YLL, followed by number of deaths and DALYs. In addition to differences in alcohol consumption, the influence of alcohol on cardiovascular diseases was less pronounced in Switzerland because prevalence of cardiovascular diseases is lower compared to Germany.
14
Implications for Policy
While light to moderate alcohol consumption decreases the risk of cardiovascular events when consumed regularly without heavy drinking episodes, alcohol has at the same time many detrimental features that need to be considered when formulating public health policies. Many factors, such as family history of substance use and abuse, liver status, depression, and others, play a role in determining recommendation to individual patients. These have to be done on a case by case basis because the potential for harm is great when recommending alcohol intake to healthy or ill patients. Without randomized trials, there is some uncertainty left on risks and benefits of alcohol consumption. Only a relatively low proportion of the population in Switzerland drank at levels where a cardioprotective effect can be expected. Nevertheless, given that alcohol is not part of a necessary diet, most of this burden could theoretically be prevented if there would be no alcohol consumption in a population. Having reached the conclusion that the burden of alcohol consumption is substantial in both countries, interventions should be a public health priority in both countries. A reduction in overall consumption at a population level should be the focus of measures to reduce the burden attributable to alcohol and is possible, as several neighboring, traditionally high consumption countries have shown (> Figure 93-1). Evidence-based and cost-effective interventions include, but are not
Alcohol Consumption and Burden of Disease: Germany and Switzerland
93
. Figure 93-1 Adult per capita alcohol consumption in selected European countries. This graph shows the trend in average per capita alcohol consumption in liters of pure alcohol from 1961 to 2003 for selected countries in WHO subregion Europe A. Source: Global Alcohol Database (GAD) (www3. who.int/whosis). Reprinted from Roerecke et al. (2007), with permission from the publisher. Copyright ß2007, Neuland Verlag
limited to, increase in price by, for example, tax increases, availability of alcohol (number of outlets, hours of sales), or brief interventions by health care personnel, which have been described in detail elsewhere (Babor et al., 2003). Tax increases have been shown to be the most effective and cost-effective intervention on a population level (Chisholm et al., 2004) and have been shown to affect on mortality rates for liver cirrhosis, impaired driving, and also criminal activity. Even in high-risk drinkers an increase in price can lower the consumption level (Chaloupka et al., 2002). However, these interventions need to be supported by many stakeholders and policy makers in order to be implemented. Switzerland has made steps in the opposite direction in allowing sale of alcohol at gas stations and tax reduction for some alcoholic beverages. The result was an increase in consumption (Mohler-Kuo et al., 2004; Kuo et al., 2003). Besides price and availability, the minimum drinking age could be enforced more than it has been in the past in Switzerland (Gruppe Schweizer Alkoholpolitik, 2005); a measure which has been shown to be very effective in reducing alcohol consumption and traffic related harm among youth (Wagenaar and Toomey, 2002). There is now considerable evidence that
1615
1616
93
Alcohol Consumption and Burden of Disease: Germany and Switzerland
suggests that educational programs and public information campaigns are largely ineffective and even might have detrimental effects (Foxcroft et al., 2003; Room et al., 2005). However, public information might play a role in gaining support for evidence-based interventions on a population level. These points should be considered in formulating a comprehensive and determined public health policy, including both structural measures and raising public awareness, in formulating an effective public health response to the substantial burden of disease attributable to alcohol in Europe and among women and youth in Switzerland in particular.
Summary Points The direct health effect of alcohol consumption can be detrimental or beneficial, depending on the average volume and pattern of drinking
Alcohol consumption in Europe, including Germany and Switzerland is high compared to worldwide estimates
High overall consumption and the distribution of consumption lead to a substantial burden of disease in both countries, particular among men in both countries and women in Switzerland Effective interventions, such as price increase or restriction in availability, are available and should be implemented to reduce this burden of disease attributable to alcohol Methodological limitations apply and more research is underway to improve estimates presented
Acknowledgments This study has been supported by the Swiss Federal Office of Public Health (FOPH-Project ‘‘Alkoholbedingte Mortalitaet und Krankheitslast in der Schweiz – von der Epidemiologie zu Empfehlungen hinsichtlich alkoholpolitischer Massnahmen’’; contract 05.001178). The methodology was in part developed for the Comparative Risk Assessment for Alcohol, financially supported by the Swiss FOPH, the World Health Organization (WHO) and the Research Institute for Public Health and Addiction, a WHO collaborative centre.
References Augustin R, Kraus L. (2005). Sucht. 51: S29–S39. Babor T, Caetano R, Casswell S, Edwards G, Giesbrecht N, Graham K, Grube J, Grunewald P, ¨ sterberg E, Rehm J, Hill L, Holder H, Homel R, O Room R, Rossow I. (2003). Alcohol: No Ordinary Commodity - Research and Public Policy. Oxford University Press, Oxford. Boffetta P, Hashibe M. (2006). Lancet Oncol. 7: 149–156.
Bundesamt fu¨r Statistik. (2004). Schweizerische Gesundheitsbefragung (2002). Bundesamt fu¨r Statistik, Neuchatel. Chaloupka F, Grossman M, Saffer H. (2002). Alcohol Res Health. 26: 22–34. Chisholm D, Rehm J, Van Ommeren M, Monteiro M. (2004). J Stud Alcohol. 65: 782–793. Cho E, Smith-Warner SA, Ritz J, van den Brandt PA, Colditz GA, Folsom AR, Freudenheim JL,
Alcohol Consumption and Burden of Disease: Germany and Switzerland Giovannucci E, Goldbohm RA, Graham S, Holmberg L, Kim DH, Malila N, Miller AB, Pietinen P, Rohan TE, Sellers TA, Speizer FE, Willett WC, Wolk A, Hunter DJ. (2004). Ann Intern Med. 140: 603–613. Corrao G, Bagnardi V, Zambon A, La Vecchia C. (2004). Prev Med 38: 613–619. Corrao G, Rubbiati L, Bagnardi V, Zambon A, Poikolainen K. (2000). Addiction. 95: 1505–1523. English D, Holman C, Milne E, Winter M, Hulse G, Codde G. (1995). The Quantification of DrugCaused Morbidity and Mortality in Australia, Commonwealth Department of Human Services and Health, Canberra, Australia. Ezzati M, Lopez AD, Rodgers A, Vander Hoorn S, Murray CJ. (2002). Lancet. 360: 1347–1360. Foxcroft DR, Ireland D, Lister-Sharp DJ, Lowe G, Breen R. (2003). Addiction. 98: 397–411. Gmel G, Rehm J. (2004). Contemp Drug Probl. 31: 467–540. Gruppe Schweizer Alkoholpolitik (2005). Alkoholpolitische Massnahmen in der Schweiz 2004 – was ist realisiert und was bringt die Zukunft? In: Babor T, Caetano R, Casswell S, Edwards G, Giesbrecht N, Graham K, Grube J, Grunewald P, Hill L, Holder H, ¨ sterberg E, Rehm J, Room R, Rossow I Homel R, O (eds) Alkohol: kein gewo¨hnliches Konsumgut. Forschung und Alkoholpolitik. Hogrefe, Go¨ttingen, pp. 334–344. Gutjahr E, Gmel G, Rehm J. (2001). Eur Addict Res. 7: 117–127. Hill AB. (1965). Proc R Soc Med. 58: 295–300. Kuo M, Heeb J, Gmel G, Rehm J. (2003). Alcohol Clin Exp Res. 27: 720–725. Midanik L. (1982). Br J Addict. 77: 357–382. Mohler-Kuo M, Rehm J, Heeb JL, Gmel G. (2004). J Stud Alcohol. 65: 266–273. Popova S, Rehm J, Patra J, Zatonski W. (2007). Alcohol Alcohol. 42: 465–473. Rehm J, Room R, Monteiro M, Gmel G, Graham K, Rehn N, Sempos CT, Frick U, Jernigan D. (2004).
93
In: Ezzati M, Lopez A, Rodgers A, Murray C (eds.) Comparative Quantification of Health Risks: Global and Regional Burden of Disease Attributable to Selected Major Risk Factors. vol. 1. World Health Organization, Geneva, pp. 959–1109. Rehm J. (1998). Alcohol Clin Exp Res. 22: 4S–14S. Rehm J, Chisholm D, Room R, Lopez A (2006a). In: Disease Control Priorities in Developing Countries. Oxford University Press and World Bank, Washington, D.C, pp. 887–906. Rehm J, Klotsche J, Patra J (2007a). Int J Methods Psychiatr Res. 16: 66–76. Rehm J, Patra J, Baliunas D, Popova S, Roerecke M, Taylor B. (2006b). Alcohol consumption and the global burden of disease 2002. World Health Organization, Department of Mental Health and Substance Abuse, Management of Substance Abuse, Geneva. Rehm J, Sempos CT, Trevisan M. (2003). J Cardiovasc Risk. 10: 15–20. Rehm J, Sulkowska U, Manczuk M, Boffetta P, Powles J, Popova S, Zatonski W. (2007b). Int J Epidemiol. 36: 458–467. Reynolds K, Lewis LB, Nolen JDL, Kinney GL, Sathya B, He J. (2003). JAMA 289: 579–588. Rimm EB, Williams P, Fosher K, Criqui M, Stampfer MJ (1999). Br Med J. 319: 1523–1528D. Rimm EB, Moats C. (2007). Ann Epidemiol. 17: S3–S7. Room R, Babor T, Rehm J. (2005). Lancet. 365: 519–530. Rothman KJ, Greenland S. (1998). Modern Epidemiology. Lippincott-Raven Publishers, Philadelphia, PA. United Nations. (2005). World Population Prospects. The 2004 revision. United Nations Department of Economic and Social Affairs, Population Division. Wagenaar AC, Toomey TL. (2002). J Stud Alcohol. 14: 206–225. Walter SD. (1980). Am J Epidemiol. 112: 409–416. Walter SD. (1976). Biometrics. 32: 829–849. World Health Organization. (2006). Global Alcohol Database World Health Organization, Geneva.
1617
94 Alcoholic Beverage Preference, Morbidity and Mortality T. E. Strandberg 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1620
2
Mechanisms of Health Effects of Ethanol and Various Alcoholic Beverages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1621 Ethanol Itself . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1622 Beer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1622 Wine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1623 Liquor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1625
2.1 2.2 2.3 2.4
3 Effects of Various Beverages on Morbidity and Mortality . . . . . . . . . . . . . . . . . . . . . . . 1625 3.1 Cardiovascular Risk and Beverages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1626 3.2 Noncardiovascular Disease and Mortality Risk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1627 4
True Relationship Between Alcohol Beverages and Mortality and Confounders in Between . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1627
5
Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1628 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1629
#
Springer Science+Business Media LLC 2010 (USA)
1620
94
Alcoholic Beverage Preference, Morbidity and Mortality
Abstract: It is an interesting question whether moderate consumption of > alcohol beverages – especially red wine – could be recommended to reduce vascular disease burden in the society. A large body of studies exists showing an association between light to > moderate drinking (often of wine) and a reduced risk of vascular disease, particularly of coronary heart disease. There are also several plausible mechanisms explaining this association, for example relating to antiatherosclerotic and antithrombotic effects and improved vascular function. However, we are still dealing with observational findings, not with those from controlled studies, such as we have from statin and aspirin treatment. No randomized, controlled studies exist which would show a beneficial effect of wine drinking on meaningful vascular endpoints. On the other hand, clear health hazards are related to alcohol consumption in the society, and harms are generally related to the per caput average consumption. Therefore, increasing moderate drinking may also lead to increase in harms. Consequently, recommendations for use of particular alcohol beverages in vascular prevention are difficult to institute safely in practice. On the other hand, observational epidemiology can be used to assure that those individuals drinking moderately can safely continue their habit, and their moderation appears to be an indicator of good prognosis. But – are they better off because of their drinking or accompanying personal and lifestyle characteristics seems still, however, to be an open question. "
1
Trinke ma¨ssig aber regelma¨ssig (Old Steinha¨ger-slogan)
Introduction
Ethanol (in this review used interchangeably with ‘‘alcohol’’) is a paradoxical substance and legalized drug of modern societies. While alcohol is addictive and its excessive consumption causes well-known ill effects (Room et al., 2003), at the same time protective effects of light to moderate alcohol consumption from cardiovascular diseases and mortality have been documented in numerous studies leading to a putative ‘‘J-shaped’’ relationship between alcohol consumption and health effects (> Figure 94‐1). In this context ‘‘moderate’’ usually equals 1–2 > drinks per day for women and 2–4 drinks per day for men, and often particularly wine. These interesting associations have been studied, in fact, over 150 years (Klatsky, 1998). The protective effect of moderate alcohol consumption has especially concerned coronary heart disease (Camargo et al., 1997; Corrao et al., 2000; de Lorgeril et al., 2008; Friedman and kimball, 1986; Fuchs et al., 1995; Thun et al., 1997), but also congestive heart failure (Walsh et al., 2002), whereas for stroke results have been inconsistent. For example, in a population-based cerebral magnetic resonance imaging-study, there was no association between low to moderate alcohol consumption and cerebral infarction, and increased alcohol intake was associated with brain atrophy (Ding et al., 2004). Furthermore, moderate alcohol consumption has also been associated with less dementia and better cognitive function (Anttila et al., 2004; Letenneur et al., 2004; Lindeman et al., 2005; Mukamal et al., 2003b), which would be a clearly attractive property in ageing societies. In a recent study in women, this was especially related to wine consumption (Mehlig et al., 2008). Highly interesting issues are: Are these associations causal, and are they related to specific alcohol beverages? In this review, the clear health hazards of excessive drinking and alcoholism are handled only marginally, main focus is put on the presumed and demonstrated health effects (reducing morbidity and mortality) of ethanol itself and various > alcoholic beverages divided in three
Alcoholic Beverage Preference, Morbidity and Mortality
94
. Figure 94‐1 J-shaped association between alcohol consumption and health effects. The shape and proportion of the curve is purely schematic and is only aimed to present the proposed general relationship. Number of drinks denote the amounts usually considered to be optimal for cardiovascular benefits
major groups: beer, wine and liquor (strong spirits). Second, the somewhat mixed evidence of various alcoholic beverages, cardiovascular morbidity and mortality will be presented, with consideration of the effects of consumption type (with meals, alone, regular moderate or heavy, episodic or ‘‘binge’’ drinking). Final part is a discussion of the net effects, and whether the relationship between alcohol beverages and health effects are truly causal or explained by > confounder, > confounding factors. When both health benefits and harms of alcohol are considered, at least the factors in > Table 94‐1 must be taken into account.
2
Mechanisms of Health Effects of Ethanol and Various Alcoholic Beverages
Because alcohol is the common denominator of all alcoholic beverages, a brief overview of ethanol itself and its metabolism is presented. However, because alcoholic beverages are composites of ethanol and matrix, exact division of their effects is often difficult. Extensive reviews of the metabolism of ethanol and general properties of various alcoholic beverages can be found in current textbooks, and only a brief overview is given here.
1621
1622
94
Alcoholic Beverage Preference, Morbidity and Mortality
. Table 94‐1 Factors affecting benefits and harms of alcoholic beverages Amount of ethanol Beverage type (beer, wine, liquor, mixed) Regularity of drinking Pattern of drinking (with or outside meals, binge drinking) Gender of the consumer (physiologic differences, pattern of drinking) Age of the consumer (majority of alcohol-related violent deaths occur before 50 years of age, possible cardiovascular benefits later in life) Risk of coronary heart disease Confounding genetic, socioeconomic and lifestyle factors
2.1
Ethanol Itself
Although ethanol is not a nutrient as such, its use nevertheless affects human nutrition in several ways, for example by being an energy source and interfering with the metabolism of other nutrients. Ethanol is rapidly absorbed from the upper part of small intestine, concomitant food slows its absorption only to a small extent. Ethanol is metabolized under normal conditions in a non-tolerant individual at a relatively stable rate through oxidation to acetaldehyde and subsequently to carbon dioxide and water, partly in stomach, but mostly in liver. Metabolism is affected by alcohol amount and tolerance, gender, body composition, and possible concomitant diseases modifying gastrointestinal and hepatic function. > Alcohol has both negative and positive physiologic effects and positive effects are generally encountered at low to moderate doses, and at the epidemiological level ill effects are usually due to excessive intake. Furthermore, the harms of alcohol beverages are considered to be primarily due to ethanol and other matrix ingredients play a minor, if any, role. Beneficial effects related to human health are listed in > Table 94‐2. Although there may be a great variety of rational reasons why people consume alcohol beverages (social habit, gastronomic reasons, company to meals, reasons related to health) the general bottom line is that it is an addictive substance the effects of which many people experience as relaxing and psychologically pleasant. The type of beverage consumed in a particular geographical area is rooted in the culture and agricultural climate. For example in Europe, there has traditionally been a ‘‘wine belt’’ around the Mediterranean, a ‘‘beer belt’’ in Central Europe, and ‘‘hard liquor belt’’ in northern regions. These are also associated with type of drinking: strong liquor is associated with > binge drinking, wine drinking with meals, and beer drinking there between. But not only geographically, alcohol beverage preference (and amount) is also dictated by social class, gender, age, and reference group. Although cultural globalization is eroding the differences, all these factors are necessary confounders to be taken into account, when the associations of beverages, morbidity and mortality are considered.
2.2
Beer
Beer is a mild alcoholic beverage with alcohol content (volume percent) typically around 2–6%, but it can reach up to 10%. Beer is produced from a wide variety of natural nutrients,
Alcoholic Beverage Preference, Morbidity and Mortality
94
. Table 94-2 Beneficial effects of light to moderate ethanol consumption on human and vascular health (Guiraud et al., 2004; Kloner and Rezkalla, 2007; de Lorgeril et al., 2008) Anti-atherosclerotic effects Increase of plasma HDL cholesterol Anti-inflammatory effects Antioxidant effects Antithrombotic effects Reduction of plasma fibrinogen concentration Increase in fibrinolysis Decrease in platelet aggregation Improves cardiac and vascular function Correction at endothelial dysfunction Reduction in plasma viscosity Protection of ischemic myocardium
. Table 94‐3 Ingredients of beer (hop) with possible physiological actions Ingredient
Health effect
Alfa-acids Humulon
Antioxidant
Lupulon
Antioxidant
Etheric oils Myrsen (terpenic compound)
?
some common sources being barley and rice. In its simplest form beer includes only water, hops and barley. Beer is a source of energy, and depending on its origin it may contain several substances with biological activity (Collie and Higgings, 2002). However, most important is hop (Humulus lupulus) the cones of which are used to flavor beer and their resins give beer its bitter taste. Ingredients of hop with possible physiological effects are listed in > Table 94‐3.
2.3
Wine
The alcoholic content of wine is typically 9–13%, but both lighter and stronger versions are produced. Strong wines may contain alcohol up to 20%. Typical origin of wine are grapes produced by Vitis vinifera in the temperate regions of Europe, Americas, Africa, and Australasia. Also other sources such as currants may be used for wine-like products. Many wines are blended, i.e., include several types of grape. According to the production method, three major classes of wine are distinguished; red, white and rose (blush). Red wine is produced of crushed red grapes, which are fermented with
1623
1624
94
Alcoholic Beverage Preference, Morbidity and Mortality
their peel and seeds. White wine is produced of mostly white but also red grapes without peel and seeds. Red grapes are also used for rose wines, but their peel is removed halfway of the fermentation process. In the production of strong wines (port, madeira) the fermentation process is interrupted before all sugar is consumed by adding distilled alcohol. For the production of sherry all sugar is fermented before distilled alcohol with sugar added afterwards to the desired quantity. The main ingredients of wine are water, variable amount of sugar, phytocomplex (matrix), and alcohol which is added through fermentation of sugar. Detailed characteristics of wine related to health depend on its matrix and various phenol (probably over 100 different) and non-phenol compounds in it. Matrix is influenced by several factors: type of grape, soil, climate, conditions during harvesting, production method (red, white, rose), fermentation and ripening. Consequently, various wines may substantially differ, even between vintages. For example, within the same grape and region the concentration of resveratrol (stilbene) in red wine may vary 100-fold between vintages (Bertelli, 2007). In principle, the same applies to the dozens of phenolic (and other) substances. Some ingredients peculiar to red and white wine are presented in > Table 94‐4. . Table 94-4 Ingredients of red and white wine with possible physiological actions (Bertelli, 2007) Type
Ingredient
Health effect
Red wine
Caffeic acid
Monophenol with antioxidant properties
Catechin, quercetin, procyanid, resveratrol
Polyphenols with antithrombotic, antioxidant, antiinflammatory properties
White wine
Caffeic acid, hydroxytyrosol, Monophenols with antioxidant properties, also found in tyrosol extravirgin olive oil Shikimic acid
Non-phenol with immunostimulant properties, basic chemical compound for synthesis of antivirals
There has been controversy about the clinical significance of wine ingredients as to human health. Recently, the following conclusions of > phenolic compounds (Polyphenols) in wine and human physiology were presented (Bertelli, 2007): Phenol compounds in wine are bioavailable after chronic consumption, they have biological activity at very low levels, they may interact with other substances, and small daily doses lead to accumulation in tissues. Because especially red wine contains phenol and several other substances with favorable biological activity, wine consumption has been increasingly advocated for cardiovascular prevention in particular, among middle-aged and elderly people (de Lange and van de Wiel, 2004). Mediating mechanisms have included antioxidant, anti-inflammatory, and lipid-modifying properties beyond those of ethanol only. For example, a recent clinical study of healthy volunteers showed that while both liquor (gin) and wine had anti-inflammatory properties, only wine decreased C-reactive protein and endothelial adhesion molecules known to be associated with vascular risk (Estruch et al., 2004). One red wine ingredient, resveratrol, has also been studied as an highly interesting ‘‘anti-ageing’’ substance due to its sirtuin (produces longevity in lower organisms)- activating effects (Sinclair and Guarante, 2006), and pharmacologic products for human are being investigated.
Alcoholic Beverage Preference, Morbidity and Mortality
94
Recently, a new mechanism of action of wine was presented. Postprandial cytotoxic lipid peroxidation products may be important in the pathogenesis of > atherosclerosis, because these compounds are injurious to the vessel wall. High-fat, high-cholesterol foods, which are rich of oxidized products, can affect lipoprotein metabolism in the body and expose arteries to their cytotoxic effects. Red wine polyphenols prevented absorption of these postprandial, lipotoxin compounds in healthy volunteers (Gorelik et al., 2008). This mechanism might explain why the habit of consuming alcohol beverages regularly and with meals – such as wine often is – may be especially beneficial, although a large follow-up study failed to show benefits of drinking with meals (Mukamal et al., 2003a). In general, the pattern of regular, moderate intake – in contrast to binge drinking – has been emphasized as an important element in cardiovascular protection (Groenbaek, 2006; Ma¨kela¨ et al., 2005). Moderate alcohol consumption, and possibly wine in particular, has also beneficial effects on arterial stiffness, wave reflections and central blood pressure, which are increasingly identified as important pathophysiologic factors for cardiovascular diseases and arterial ageing (O’Rourke and Hashimoto, 2007). They are not necessarily related to the conventional brachial blood pressure. Moderate alcohol consumption was associated with reduced arterial stiffness in a large population-based study of elderly, although the association was significant only in women (Mattace-Raso et al., 2005). Concomitant red wine prevented harmful acute cardiovascular effects (wave reflections, blood pressure rise) of smoking in young healthy smokers (Papamichael et al., 2006). Recently combination of olive oil and wine had beneficial postprandial reduction on augmentation index as compared to control (Papamichael et al., 2008). It is obscure, however, whether these actions are due specifically to wine or moderate ethanol consumption in general.
2.4
Liquor
The main characteristics of liquor (strong spirits) is its high alcohol content which is due to removal of water through distillation. Because the source and consequent matrix of the beverage varies (grapes for brandy, barley for whisky, grains for vodka, various fruit for liquors etc.) a strong alcoholic beverage may resemble beer or wine in its other health effects.
3
Effects of Various Beverages on Morbidity and Mortality
As mentioned above, amount of alcohol (moderate), type (wine) and pattern (with meals) have been associated with less morbidity and mortality, especially from cardiovascular causes. Very good mechanisms, such as antiatherogenic, antithrombotic, and cardiovascular functionimproving (de Lorgeril et al., 2008; Kloner and Rezkalla, 2007; Tolstrup and Groenbaek, 2007) for these associations have been presented, but epidemiological studies are nevertheless hampered by many factors. How reliable are reports of alcohol consumption? What is the proper reference group? Non-drinkers may include those who have quitted drinking because of disease or alcoholism, and be therefore at higher risk. What is the impact of social class and other lifestyle besides consuming alcohol moderately? In fact, a long-term follow-up study in a homogenous group of at baseline middle-aged men of high social class revealed no mortality benefits between abstinence and moderate consumption (Strandberg et al., 2004). In fact, higher mortality among those with heavier consumption was explained by other risk factors
1625
1626
94
Alcoholic Beverage Preference, Morbidity and Mortality
such as smoking. These results are in accordance with a recent population-based study where no consistent health differences (not directly mortality) were observed between life-time > abstainers and moderate drinkers (Stranges et al., 2006). When discerning differences between beverages it should be borne in mind that many drinkers are mixed drinkers and being a preferrer of one beverage only (such as wine or beer-only drinkers) may again bring confounding into play. Also these caveats must be taken into account when the effects of beverage types on mortality are considered.
3.1
Cardiovascular Risk and Beverages
The following review is mainly based on three recent and carefully performed epidemiological reports of large numbers of people, both in population studies and selective cohorts. These reports have assessed the relationship between alcohol beverage type, mortality and especially vascular risk. A meta-analysis based on 26 studies (Di Castelnuovo et al., 2002) found a significant inverse association between light- to moderate consumption of wine and vascular risk. In 13 studies (over 200,000 individuals) wine intake was associated with a 32% reduced relative risk of vascular disease as compared to nondrinkers. In ten studies, a J-shape dose-response was found between wine consumption and vascular risk, the inverse relationship was observed up to an amount of 150 mL of wine daily. Moderate beer intake was also associated with a 22% reduced vascular risk in 15 studies (over 200,000 persons), however, no dose-response was observed between beer consumption and vascular risk in seven studies (over 136,000 persons). Underreporting and/or type of drinking (more binge drinking with beer) may have obscured the associations. After the meta-analysis, a large (almost 130,000 adults) and long-term cohort study from California was published (Klatsky et al., 2003). This study assessed the relationship between wine (and also wine types), beer, or liquor and mortality. A J-shaped association was observed between alcohol consumption and mortality, and much of the lower risk of light drinking was explained by wine. Although each beverage type protected against coronary risk, overall mortality risk was lower among wine drinkers as compared to beer or liquor preferrers. This study also discerned between wine types (red, white, other types or combinations), but no differences in mortality were observed. An important finding in this study was the suggestion of more benefit from wine preference among women. However, several possible confounders were identified. In contrast, the third report, a 12-year follow-up among almost 40,000 male health professionals (Mukamal et al., 2003a) did not find differences between beverage type and risk of coronary heart disease. Total mortality was not reported. However, alcohol consumption over 10 g per day (one drink) to at least 50 g per day (four drinks) protected from both nonfatal and fatal myocardial infarction. Because there were few participants consuming over 50 g per day, the possibilities to assess heavy drinking were limited. All beverage types seemed to protect and beer and liquor even more than wine, shifting again focus on the role of alcohol as such as the protective factor. Because beer and liquor were the most frequent beverages in their population, the authors explain this with the hypothesis that the most widely (and probably most regularly) consumed beverages are inversely associated with vascular risk in a particular population (Rimm, 1996). This may supported by the population level analyses in the US where significant protective effect of beer on ischemic heart disease was noted (Kerr and Ye, 2007).
Alcoholic Beverage Preference, Morbidity and Mortality
3.2
94
Noncardiovascular Disease and Mortality Risk
In contrast to vascular risk, there is less or no controversy about the relationship between alcohol beverages and noncardiovascular endpoints: the essential factor is the amount of ethanol consumed. This applies not only to violent endpoints (accidents, suicides), but to various cancers and hepatic pathology alike. For example, the development of hepatic cirrhosis is clearly related to the cumulative amount of ethanol consumed, not the type of alcoholic beverage.
4
True Relationship Between Alcohol Beverages and Mortality and Confounders in Between
As shown above, wine preference has provided protection in many epidemiological studies and plausible mechanisms have been discovered to explain these associations. However, the question of cause and effect and the role of confounders – for example social factors and cardiovascular risk factors – associated with wine preference has remained. Several studies have further considered the possibility that wine drinkers represent a special trait with overall healthy lifestyle and attitude to life from young age (Barefoot et al., 2002; Gronbaek et al., 1999; McCann et al., 2003; Mortensen et al., 2001; Rimm and Stampfer, 2002; Paschall and Lipton, 2005). For example, a recent study from Denmark showed that customers buying wine at the supermarket made more healthy choices of other food items than people who bought beer (Johansen et al., 2006). Moderate users and wine preferrers may thus simply be protected by other beneficial effects of their life course and lifestyle (‘‘the healthy user bias’’). Therefore, controlling for confounders, especially social class, would be very important in observational studies. Moreover, alcohol consumption is usually a life-long habit, and the health effects should also be considered in the long-term, not only for 5–10-year follow-up times. Possible confounders associated with beverage preference and vascular risk are presented in > Table 94-5. Looking at the long list of important factors related to both wine preference and lower vascular and mortality risk, it is no wonder that in observational, epidemiological studies a significant correlation would emerge between them without true causal connection. On the other hand, the equally long list of beneficial mechanisms related to wine drinking in careful clinical studies implicates that a causal relation could exist. In a recent long-term study from middle-age to old age showed that wine preference was associated with lower mortality in a socioeconomically homogenous group of men (Strandberg et al., 2007). Moreover, the result remained after adjustments for various cardiovascular risk factors. Maybe only a full life-course study (or a randomized trial) accounting for also pre-midlife characteristics could finally decide whether it is truly the characteristics of the alcoholic beverages – and not the characteristics of the people who make the choices – that account for these differences. The most complete evidence would come from a > randomized clinical trial (Kloner and Rezkalla, 2007), where participants are treated with red wine or a control substance (although it is difficult to imagine a placebo wine). This kind of trial would meet complex ethical and organizational problems, and as a primary prevention trial in healthy persons would be close to impossible to carry out. A trial in secondary prevention, for example after myocardial infarction, could be more feasible (for example one drink a day or every other day, as suggested by Kloner and Rezkalla (2007), but even without such study, some
1627
1628
94
Alcoholic Beverage Preference, Morbidity and Mortality
. Table 94-5 Associations of type of alcohol consumption or beverage preference with some factors related to vascular risk Factor
Beverage
Female gender
More women among wine drinkers
References Paschall and Lipton (2005)
Some of the better health indicators related to wine preference seen especially among women Subjective health
Better among wine drinkers
Paschall and Lipton (2005)
Cigarette consumption
Greater in heavier drinkers Lower among wine drinkers
McCann et al. (2003), Strandberg et al. (2004)
Depressive symptoms
Less among female wine drinkers
Paschall and Lipton (2005)
Least among abstainers
Paschall and Lipton (2005)
Physical activity
More among wine drinkers Better dietary habits
More among wine drinkers
McCann et al. (2003), Paschall and Lipton (2005), Johansen et al. (2006)
Larger waist circumference
Moderate to high consumption of alcohol and of beer and liquor
Vadstrup et al. (2003)
Moderate drinking
More among wine consumers
College education, More among wine drinkers or higher socioeconomic status
Paschall and Lipton (2005) Mortensen et al. (2001), McCann et al. (2003), Paschall and Lipton (2005), Strandberg et al. (2007)
authorities already are enthusiastic about the prospects of moderate amounts of wine in secondary prevention (de Lorgeril et al., 2008). However, in the present day coronary patients the benefits should appear atop evidence-based treatments such as statins, aspirin, beta-blockers and ACE-inhibitors. In > Table 94-6 there is a comparison of proof from cholesterol-lowering and current proof from wine drinking in the prevention of vascular diseases.
5
Conclusions
Could moderate consumption of alcohol beverages – and red wine in particular – be generally recommended as a way to reduce vascular disease burden in the society? The answer at present is for several reasons: No (perhaps only exception being a heavy drinker with a heavy family history of coronary heart disease). Despite a large body of studies showing an association between light to moderate and wine drinking and despite plausible beneficial mechanisms, we are still dealing with observational findings not with those from randomized controlled studies, such as we have from statin and aspirin treatments. Recommendations for use could be given more lightly if alcohol would not have harms. Truly moderate consumption of wine hardly involves them, but ‘‘moderation’’ is a relative subject and ‘‘two to three glasses per day’’
Alcoholic Beverage Preference, Morbidity and Mortality
94
. Table 94-6 Comparison of evidence from cholesterol lowering and wine drinking in vascular prevention Type of evidence
Cholesterol lowering
Wine drinking
Epidemiologic studies
Abundant and consistent evidence between lower (or lowered) cholesterol levels and lower vascular risk. Relationship not affected by covariates
Abundant, but not totally consistent evidence between wine preference and lower vascular risk. Relationship often abolished after adjustments. Association with moderate drinking in general is more robust
Mechanistic studies
Well-documented mechanisms explaining the link between cholesterol and vascular risk
Well-documented, but not totally consistent, mechanisms explaining the link between wine consumption (especially red) and vascular risk
Animal studies
Well documented animal models
Not consistent (Bentzon et al., 2001)
Randomized, Robust evidence showing reduction of vascular vascular events after cholesterol lowering endpoint studies (especially with statins)
Not available
may be taken very differently by individuals and lead to heavier consumption with time in some. After all, the majority of harms at the population level occur paradoxically among moderate consumers of alcohol (Poikolainen et al., 2007). Due to the wide and positive media coverage on health effects of alcohol, earlier abstainers, for example older women, may start drinking their daily portion of red wine without robust evidence that starting drinking in old age would promote health (Wannamethee and Shaper, 2002), and may on the contrary lead to misuse (Loukissa, 2007). On the other hand, observational epidemiology may testify that those individuals really drinking moderately (any beverage) – and keeping it that way – can safely continue their habit. Their moderation appears to be an indicator of good prognosis, also among coronary patients. But – are they better off because of their drinking or accompanying personal and lifestyle characteristics seems still, however, to be an open question.
Summary Points Alcoholic beverages are divided in three major classes, beer, wine and spirits, which differ by their average alcohol (ethanol) content and the composition of matrix (phytocomplex).
Both ethanol itself and the beverage matrix have effects on health. Alcohol is an addictive substance and alcohol consumption has well-known harmful effects, which have to be carefully weighed against possible benefits.
There is a large body of studies showing an association between light to moderate and especially red wine drinking with reduced risk of vascular disease.
Regular (but light to moderate) > drinking pattern is better that binge-type.
1629
1630
94
Alcoholic Beverage Preference, Morbidity and Mortality
Plausible beneficial mechanisms (antiatherosclerotic, antihrombotic, improved vascular function) exist to explain this association.
But beverage preference is also associated with lifestyle factors (diet, smoking, socioeconomic status etc) which may confound the relationship with vascular risk.
Taking into account the risks of excessive consumption, general recommendations cannot be given to use alcohol beverages in vascular prevention.
On the other hand, there is no need to change the habit of drinking – IF it is truly light to moderate.
Acknowledgments Dr. Kari Poikolainen is gratefully acknowledged for expert advice during the preparation of the manuscript. Sources of support: The Academy of Finland, the Pa¨ivikki and Sakari Sohlberg Foundation, the Helsinki University Central Hospital, and the Finnish Foundation for Cardiovascular Research are gratefully acknowledged for financial support.
References Anttila T, Helkala EL, Viitanen M, Kareholt I, Fratiglioni L, Winblad B, Soininen H, Tuomilehto J, Nissinen A, Kivipelto M. (2004). Br Med J. 329: 539. Barefoot JC, Gronbaek M, Feaganes JR, McPherson RS, Williams RB, Siegler IC. (2002). Am J Clin Nutr. 76: 466–472. Bentzon JF, Skovenborg E, Hansen C, et al. (2001). Circulation. 103: 1681–1687. Bertelli AAE. (2007). Atherosclerosis. 195: 242–247. Camargo CA Jr, Hennekens CH, Gaziano JM, Glynn RJ, Manson JE, Stampfer MJ. (1997). Arch Intern Med. 157: 79–85. Collie ME, Higgins JC. (2002). Arch Intern Med. 162: 364–365. Corrao G, Rubbiati L, Bagnardi V, Zambon A, Poikolainen K. (2000). Addiction. 95: 1505–1523. de Lange DW, van de Wiel A. (2004). Semin Vasc Med. 4: 173–186. de Lorgeril M, Salen P, Corcos T, Defaye P, Juneau M, Pavy B, Saoudi N. (2008). Eur Heart J. 29: 4–6. Di Castelnuovo A, Rotondo S, Iacoviello L, Donati MB, de Gaetano G. (2002). Circulation. 105: 2836–2844. Ding J, Eigenbrodt ML, Mosley TH Jr, Hutchinson RG, Folsom AR, Harris TB, Nieto J. (2004). Stroke. 35: 16–21. Estruch R, Sacanella E, Badia E, Antunez E, Nicolas JM, Fernandez-Sola J, Rotilio D, de Gaetano G, Rubin E, Urbano-Marquez A. (2004). Atherosclerosis. 175: 117–123.
Friedman LA, Kimball AW. (1986). Am J Epidemiol. 124: 481–489. Fuchs CS, Stampfer MJ, Colditz GA, Giovannucci EL, Manson JE, Kwachi I, et al. (1995). N Engl J Med. 332: 1245–1250. Gorelik S, Ligumsky M, Kohen R, Kanner J. (2008). FASEB J. 22: 41–46. Groenbaek M. (2006). Curr Opin Lipidol. 17: 17–21. Groenbaek M, Mortensen EL, Mygind K, Andersen AT, Becker U, Gluud C, Sorensen TI. (1999). J Epidemiol Community Health. 53: 721–4. Guiraud A, de Lorgeril M, Boucher F, de Leiris J. (2004). J Mol Cell Cardiol. 36: 561–566. Johansen D, Friis K, Skovenborg E, Groenbaek M. (2006). Br Med J. 332: 519–522. Kerr WC, Ye Y. (2007). Alcohol Clin Exp Res. 31: 1913–1919. Klatsky AL. (1998). Novartis Found Symp. 216: 2–12. Klatsky AL, Friedman GD, Armstrong MA, KIpp H. (2003). Am J Epidemiol. 158: 585–595. Kloner RA, Rezkalla SH. (2007). Circulation. 116: 1306–1317. Letenneur L, Larrieu S, Barberger-Gateau P. (2004). Biomed Pharmacother. 58: 95–99. Lindeman RD, Wayne SJ, Baumgartner RN, Garry PJ. (2005). J Gerontol. 60A: 1065–1070. Loukissa D. (2007). Br J Nurs. 16: 1254–1258. Mattace-Raso FUS, van der Cammen TJM, van den Elzen APM, Schalekamp MADH, Asmar R, Reneman RS,
Alcoholic Beverage Preference, Morbidity and Mortality Hoeks APG, Hofman A, Witteman JCM. (2005). J Gerontol Med Sci. 60A: 1479–1483. Ma¨kela¨ P, Palja¨rvi T, Poikolainen K. (2005). J Stud Alcohol. 66: 722–728. McCann SE, Sempos C, Freudenheim JL, Muti P, Russell M, Nochajski TH, Ram M, Hovey K, Trevisan M. (2003). Nutr Metab Cardiovasc Dis. 13: 2–11. Mehlig K, Skoog I, Guo X, Schu¨tze M, Gustafson D, Waern M, Ostling S, Bjo¨rkelund C, Lissner L. Am J Epidemiol. (2008 Jan 24). [Epub ahead of print]. Mortensen EL, Jensen HH, Sanders SA, Reinisch JM. (2001). Arch Intern Med. 161: 1844–1848. Mukamal KJ, Conigrave KM, Mittleman MA, et al. (2003a). N Engl J Med. 348: 109–118. Mukamal KJ, Kuller LH, Fitzpatrick AL, Longstreth WT Jr, Mittleman MA. (2003b). Siscovick DS JAMA. 289: 1405–13. O’Rourke MF, Hashimoto J. (2007). J Am Coll Cardiol. 50: 1–13. Papamichael C, Karatzi K, Karatzis E, Papaionnou TG, Katsichi P, Zampelas A, Lekakis J. (2006). J Hypertens. 24: 1287–92. Papamichael C, Karatzi K, Papaionnou TG, Karatzis E Katsichi P, Sideris V, Zakopoulos N, Zampelas A, Lekakis J. (2008). J Hypertens. 26: 223–229. Paschall M, Lipton RI. (2005). Drug and Alcohol Dependence. 78: 339–344. Poikolainen K, Palja¨rvi T, Ma¨kela¨ P. (2007). Addiction. 102: 571–578.
94
Rimm EB. (1996). Am J Epidemiol. 143: 1094–1098. Rimm EB, Stampfer MJ. (2002). Circulation. 105: 2806–2807. Room R, Graham K, Rehm J, Jernigan D, Monteiro M. (2003). Eur Addict Res. 9: 165–175. Sinclair DA, Guarante L. (2006). Scientific American. 16: 68–75. Strandberg AY, Strandberg TE, Salomaa VV, Pitkala K, Miettinen TA. (2004). Am J Clin Nutr. 80: 1366–1371. Strandberg TE, Strandberg AY, Salomaa VV, Pitkala K, Tilvis RS, Miettinen TA. (2007). J Gerontol Med Sci. 62A: 213–218. Stranges S, Notaro J, Freudenheim JL, et al. (2006). Addiction. 101: 1265–1276. Thun MJ, Peto R, Lopez AD, Monaco JH, Henley SJ, Heath CW Jr, et al. (1997). N Engl J Med. 337: 1705–1714. Tolstrup J, Groenbaek M. (2007). Curr Atheroscler Rep. 9: 116–124. Vadstrup ES, Petersen L, Soerensen TIA, Groenbaek M. (2003). Int J Besit. 27: 238–246. Walsh CR, Larson MG, Evans JC, Djousse L, Ellison RC, Vasan R, Levy D. (2002). Ann Intern Med. 136: 181–191. Wannamethee SG, Shaper AG. (2002). Heart. 87: 32–36.
1631
95 Burden of Disease Due to Alcohol and Alcohol Related Research R. Rajendram . G. Lewison . V. R. Preedy 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1634
2
Effects of Alcohol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1635
3
Assessment of the Burden of Disease Due to Alcohol . . . . . . . . . . . . . . . . . . . . . . . . . . . 1635
4
Global Burden of Disease Due to Alcohol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1636
5
Alcohol-Related Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1636
6
Assessment of WARR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1639
7
Assessment of Regional Contribution and Relative Commitment to WARR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1642
8
Regional Contribution and Relative Commitment to WARR . . . . . . . . . . . . . . . . . . . 1643
9
Global Expenditure on Alcohol Related Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1643
10 Implications: Is the Glass Half Full or Half Empty? . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1644 10.1 The Glass is Half Empty . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1644 10.2 The Glass is Half Full . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1645 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1646
#
Springer Science+Business Media LLC 2010 (USA)
1634
95
Burden of Disease Due to Alcohol and Alcohol Related Research
Abstract: Ethanol is one of the most commonly used recreational drugs worldwide. Whilst alcohol may have some beneficial effects, it is responsible for 4% of the global burden of disease. > Alcohol-related research (ARR) is a multidisciplinary specialty with the fundamental aim of improving the lives and health of those who consume alcohol. When the worldwide literature published during 1992–2003 was reviewed we found that there were relatively few ARR publications. Biomedical research and the global disease burden due to alcohol both increased whilst the number of papers from ARR remained static. Nearly 58% of all ARR papers were from centers in Canada and the United States. A further 30% were from Western Europe and 10% originated from Australia, New Zealand or Japan. The rest of the world contributed to only 3% of all ARR publications. There is thus a need for greater interest in ARR in the developing world. The estimated annual expenditure on ARR published in 2001 was $730 million (between $12 and $13 per DALY). This represents 0.7% of the expenditure on global biomedical research. Despite the decline in ARR, from 1993 to 2003 there has been substantial progress in treatment and strategies to reduce the GBD due to alcohol. Policies to reduce overall levels of drinking and the rates of some alcohol-related problems have been identified through research and implemented effectively. Examples include increasing the price of alcohol through taxation and campaigns against driving under the influence of alcohol. However, research does not offer much guidance on successful ways of changing patterns of drinking, other than with unpalatable measures such as prohibition or alcohol rationing. Alcohol-related research is necessary, cost-effective and could significantly reduce the GBD due to alcohol. Increasing ARR is therefore likely to benefit society as a whole. List of Abbreviations: ARR, alcohol-related research; GBD, global burden of disease; ISO, International Organization for Standardization; NIH, National Institutes of Health; MRCC, Medical Research Council of Canada; UK, United Kingdom; RBD, Regional Burden of Disease; RBDA, Regional Burden of Disease Due to Alcohol; RC, > relative commitment; RL, research level; SCI, > science citation index; SSCI, > social sciences citation index; WARR, worldwide alcohol-related research; WHO, World Health Organization
1
Introduction
Ethanol is one of the most commonly used recreational drugs and is freely available throughout most of the world. Approximately 2 billion people worldwide (half the total adult population) drink alcohol (WHO, 2004). The proportion of drinkers varies between countries; from 18 to 90% of adult men and 1–81% of adult women (WHO, 2004). > Per caput consumption also varies widely. However, the amounts consumed vary significantly between individuals and some communities prohibit the consumption of alcohol on religious grounds. Many enjoy the pleasing effects of alcohol, usually without affecting themselves or others. Moderate alcohol consumption is even considered to be beneficial in reducing the risk of ischemic heart disease (Booyse and Parks, 2001; Preedy and Watson, 2004; Rehm et al., 2003c). However, in England alone approximately 1.5 million drink harmful amounts of alcohol (men drinking over 50 units/week; women drinking over 35 units/week) whilst a further 6.3 million drink hazardous amounts (men drinking 22–50 units/week; women drinking 15–35 units/ week; Deacon et al., 2007). Unfortunately, alcohol may induce any of at least 60 different
Burden of Disease Due to Alcohol and Alcohol Related Research
95
alcohol-related pathologies (Peters and Preedy, 1999; Preedy and Watson, 2004). Alcoholinduced disease may be acute (e.g., intentional and unintentional injuries) or chronic. Alcohol also has important effects on people other than the drinker, such as newborns whose mothers drank while pregnant or victims of drunk drivers.
2
Effects of Alcohol
In 2002, the most common cause of deaths attributed to alcohol consumption worldwide was drinking before or whilst doing something else such as driving (Rehm et al., 2003b). In 2001, 38% of deaths attributed to alcohol use in Canada were related to driving; alcohol-attributable cancer ranked second (23%; Rehm et al., 2006a); and liver cirrhosis third (15%). Life lost from mortality from alcohol use in England is increasing. Alcohol-specific mortality in men under 75 years old increased from 10.8/100,000 in 2001 to 12.5/100,000 in 2005 and in women increased from 5.1 to 5.7/100,000 (Deacon et al., 2007). On average, each man dying from an alcohol-attributable cause loses 20.2 years and each woman loses 15.1 years (Deacon et al., 2007). As a result the average loss life of life expectancy due to alcohol use in England is 9.9 months for men and 4.4 months for women (Deacon et al., 2007). The severity of the complications of alcohol use generally depends on the amount consumed over time, the pattern of drinking and the quality of the alcoholic beverage (Rehm et al., 2004). Morbidity and mortality rise as intake increases (Her and Rehm, 1998). High levels of alcohol intake, especially in binges (>60 g pure ethanol/day) increase the risk of liver cirrhosis, injuries, some forms of malignant neoplasms, alcohol use disorders and cardiovascular disease (Rehm et al., 2003b). It is a commonly held belief that alcohol-related diseases are entirely due to misuse as defined by regular consumption of more than the recommended maximum daily (binge drinking) or weekly allowances for alcohol. This is simply not true. Many deaths from cancer associated with ingestion of alcohol occur in people who do not misuse alcohol. For example, the risk of developing breast cancer increases with consumption levels as low as one alcoholic beverage per day (Fenoglio et al., 2003). These detrimental effects of alcohol consumption are far more significant than its cardioprotective or other potential beneficial effects (Rehm et al., 2006b).
3
Assessment of the Burden of Disease Due to Alcohol
Widely accepted methods of quantifying the adverse effects of alcohol include mortality and disability-adjusted life years (DALYs). The DALY is a well-established summary measure of population health that takes into account years lost due to premature death and time spent living with disabling conditions (years ‘‘lost’’ due to disability; Murray et al., 2000). The number of years lost due to disability is calculated by multiplying the time living with a disease condition by the disability weight of the condition (Murray and Lopez, 1996). Disability weights measure the degree of impact of a disease condition on disability, i.e., activity limitations and impairments. Disability weighting varies from no disability (zero) to death (one). Every disease state has been assigned one or more disability weight, dependent on the severity and progression of the disease, using a complicated procedure which estimated the relative impact of different disease conditions (Murray and Lopez, 1996).
1635
1636
95 4
Burden of Disease Due to Alcohol and Alcohol Related Research
Global Burden of Disease Due to Alcohol
In established market economies, 10.3% of the disease burden as quantified by disabilityadjusted life years (DALYs) is attributable to alcohol (WHO, 2002). In 2000 the global burden of disease due to alcohol was estimated at 58 million DALYs (WHO, 2002). This is significantly more than the disease burden due to illicit drug use (2.3% of DALYs) but comparable to that caused by tobacco (11.7% of DALYs). These figures significantly underestimate the true extent of the effects of ethanol as they focus on health and exclude the social problems caused by alcohol misuse. Alcohol misuse is associated with criminal and anti-social behavior, inefficiency in the workforce and violence within the families of alcoholics. The social cost of alcohol misuse is significant and far greater than that from tobacco or drug abuse (Catalyst Health Consultants, 2001; Fenoglio et al., 2003; Preedy and Watson, 2004; The Prime Minister’s Strategy Unit, 2003; Varney and Guest, 2002). For example in England, the cost of alcohol misuse to the health service has been estimated to be £1.7 billion per annum (The Prime Minister’s Strategy Unit, 2003). In comparison, the estimated annual cost of crime and anti-social behavior linked to alcohol use is thought to be £7.3 billion whilst lost productivity costs £6.4 billion per annum (The Prime Minister’s Strategy Unit, 2003). The effects of alcohol in England are thought to cost up to £20 billion per year in total (The Prime Minister’s Strategy Unit, 2003). However, the alcohol industry of the UK is associated with one million jobs and has an annual turnover over £30 billion. Alcohol consumption and the associated burden of disease are rife throughout the modern world (> Table 95-1, > Figures 95-1 and > 95-2). The global burden of disease (GBD) 1990 study revealed that 1.5% of global mortality was attributable to alcohol (Murray and Lopez, 1997) whilst in the GBD 2000 study, 3.2% was attributable to alcohol (Rehm et al., 2003b; WHO, 2002). The number of DALYs attributed to use of alcohol also increased, but not to the same degree. In 2000, alcohol accounted for 4.0% of the total number of DALYs worldwide (Rehm et al., 2003b; WHO, 2002). The GBD due to alcohol increased in terms of both disability and mortality and is probably continuing to rise. However, there is international variation in the alcohol-related burden of disease, mainly in relation to the average per caput volume of consumption and drinking patterns (Rehm et al., 2003a–c; WHO, 2002).
5
Alcohol-Related Research
Alcohol-related research (ARR) is a multidisciplinary specialty that covers a wide spectrum of research ranging from epidemiological studies of international patterns of drinking to studies at the cellular level, and it includes clinical trials of treatment for patients (Rajendram et al., 2006). One of the most fundamental aims of research is to improve the wellbeing of individuals and of society in general. The amount of alcohol-related research conducted worldwide should therefore reflect the proportion of the GBD due to alcohol. Some studies have suggested that the research grants from the United States (US) National Institutes of Health (NIH) and the Medical Research Council of Canada (MRCC) correlate relatively well with the burden of disease in the USA and Canada (Gross et al., 1999; LamarreCliche´ et al., 2001). This suggests that the commitment to ARR in the USA and Canada may also correlate with disease burden. We investigated the worldwide commitment to ARR during 1992–2003 in relation to the GBD due to alcohol (Rajendram et al., 2006).
Afghanistan, Djibouti, Egypt, Iraq, Morocco, Pakistan, Somalia, Sudan, Yemen
EMR-D
Bolivia, Ecuador, Guatemala, Haiti, Nicaragua, Peru
AMR-D
Bahrain, Cyprus, Iran, Jordan, Kuwait, Lebanon, Libyan Arab Jamahiriya, Oman, Qatar, Saudi Arabia, Syrian Arab Republic, Tunisia, United Arab Emirates
Canada, Cuba, United States of America
Antigua and Barbuda, Argentina, Bahamas, Barbados, Belize, Brazil, Chile, Colombia, Costa Rica, Dominica, Dominican Republic, El Salvador, Grenada, Guyana, Honduras, Jamaica, Mexico, Panama, Paraguay, Saint Kitts and Nevis, Saint Lucia, Saint Vincent and the Grenadines, Suriname, Trinidad and Tobago, Uruguay, Venezuela
AMR-A
AMR-B
Eastern EMR-B Mediterranean
America
Botswana, Burundi, Central African Republic, Congo, Ivory Coast, Democratic Republic of the Congo, Eritrea, Ethiopia, Kenya, Lesotho, Malawi, Mozambique, Namibia, Rwanda, South Africa, Swaziland, Uganda, United Republic of Tanzania, Zambia, Zimbabwe
343
139
71
431
325
346
294
AFR-E
6045
Algeria, Angola, Benin, Burkina Faso, Cameroon, Cape Verde, Chad, Comoros, Equatorial Guinea, Gabon, Gambia, Ghana, Guinea, GuineaBissau, Liberia, Madagascar, Mali, Mauritania, Mauritius, Niger, Nigeria, Sao Tome and Principe, Senegal, Seychelles, Sierra Leone, Togo
AFR-D
Africa
Countries
Alcohold
113
23
17
81
47
210
148
1467
0.4
0.2
1.0
9.3
3.6
4.4
1.8
58.3
0.4
0.3
1.6
16.1
6.3
7.5
3.3
100.0
0.3
0.8
5.7
11.7
7.9
2.1
1.3
4.0
1.2
1.4
14.1
21.6
11.1
12.7
6.1
9.6
DALYs % per 1,000 Pop’nb DALYsc Million % GBD pop’n millions millions DALYsd alcohold RBDd
World
WHO regiona
. Table 95-1 Burden of disease due to alcohol in the 14 WHO regions
Burden of Disease Due to Alcohol and Alcohol Related Research
95 1637
Australia, Brunei Darussalam, Japan, New Zealand, Singapore
Cambodia, China, Cook Islands, Fiji, Kiribati, Lao, Malaysia, Marshall Islands, Micronesia, Mongolia, Nauru, Niue, Palau, Papua New Guinea, Philippines, Republic of Korea, Samoa, Solomon Islands, Tonga, Tuvalu, Vanuatu, Vietnam
WPR-B
Bangladesh, Bhutan, Korea, India, Maldives, Burma, Nepal
SEA-D
WPR-A
Indonesia, Sri Lanka, Thailand
Belarus, Estonia, Hungary, Kazakhstan, Latvia, Lithuania, Republic of Moldova, Russian Federation, Ukraine
EUR-C
SEA-B
Albania, Armenia, Azerbaijan, Bosnia and Herzegovina, Bulgaria, Georgia, Kyrgyzstan, Poland, Romania, Slovakia, The Former Yugoslav Republic of Macedonia, Tajikistan, Turkmenistan, Turkey, Uzbekistan, Yugoslavia
EUR-B
1533
154
1242
294
243
218
412
241
16
358
61
59
39
53
14.0
0.8
5.6
2.1
9.1
2.6
3.5
22.9
1.2
9.1
3.5
15.2
4.3
6.0
5.6
4.6
1.5
3.4
15.2
6.5
6.7
9.1
5.2
4.5
7.1
37.4
11.9
8.5
The alcohol related disease burden in the 14 WHO Mortality Stratified Regions (WHO Geographically Defined, Mortality Stratified Regions (WHO, 2000)). Adapted from Rajendram et al. (2006) with the kind permission of Oxford University Press and the Medical Council on Alcoholism a The World Health Organisation (WHO) regions are classified by geography and mortality stratum (WHO, 2002). Mortality strata: A very low child and adult; B low child and adult; C low child, high adult; D high child and adult; E high child, very high adult (WHO, 2000) b Regional population in millions (pop’n) c Total regional disease burden in millions of disability-adjusted life years (DALYs; WHO, 2002) d Estimated burden of disease due to alcohol is shown as in millions of DALY, percentage of total regional burden of disease (% RBD) and percentage of global burden of disease (% GBD) due to alcohol (WHO, 2002)
Western Pacific
South East Asia
Andorra, Austria, Belgium, Croatia, Czech Republic, Denmark, Finland, France, Germany, Greece, Iceland, Ireland, Israel, Italy, Luxembourg, Malta, Monaco, Netherlands, Norway, Portugal, San Marino, Slovenia, Spain, Sweden, Switzerland, United Kingdom
EUR-A
Countries
DALYs % per 1,000 Pop’nb DALYsc Million % GBD pop’n millions millions DALYsd alcohold RBDd
Alcohold
95
Europe
WHO regiona
. Table 95-1 (continued)
1638 Burden of Disease Due to Alcohol and Alcohol Related Research
Burden of Disease Due to Alcohol and Alcohol Related Research
95
. Figure 95-1 Alcohol-related disease burden (DALYs). The alcohol related disease burden in various regions presented as disability adjusted life years. Data derived from WHO (2002)
. Figure 95-2 Percent of the global disease burden due to alcohol. The alcohol related disease burden in various regions presented as a percentage of the global burden of disease due to alcohol. Data derived from WHO (2002)
6
Assessment of WARR
We identified all ARR papers published during the 12 year period from 1992 to 2003 from the Science Citation Index (SCI) and Social Sciences Citation Index (SSCI) databases using a ‘‘filter’’ based on keywords (often in combination) and on specialist and semi-specialist journals (Rajendram et al., 2006). The filter was calibrated by inspection of the lists generated. The specificity (> precision, p) was 0.92 and the > sensitivity (> recall, r) was 0.88 (Rajendram et al., 2006). The calibration factor of the filter (i.e., p/r; the ratio of the true estimated total to the number of papers found) was 1.045 (Rajendram et al., 2006).
1639
1640
95
Burden of Disease Due to Alcohol and Alcohol Related Research
Duplicate papers listed in both databases and papers that were not relevant to the effects of ethanol on humans or animals (mainly chemical and industrial) were removed. Of the remaining 22,370 papers, 19% were only in the SSCI, 55% were only in the SCI and 25% were in both databases. To put ARR in context, we also determined the total number of biomedical research publications worldwide from 1992 to 2003. The data for the whole 12-year period are presented in > Table 95-2. To demonstrate the trends in biomedical and ARR the data are divided into 4-year periods in > Table 95-3. During the period 2000–2003 worldwide ARR (WARR) represented less than 0.7% of global biomedical research output (> Table 95-3; Rajendram et al., 2006). The majority (82%) was published in journals classified as clinical medicine; the rest was in biomedical research journals (8%) or ones covering psychology or sociology. The total numbers of ARR papers and biomedical papers increased during 1992–2003. However, > Figure 95-3 and > Table 95-3 illustrate the global decline in relative commitment to ARR over the same period (1992–1995, 0.77%; 2000–2003, 0.69%).
. Table 95-2 Publication of WARR due to alcohol regions from 1992 to 2003 1992–2003 Region
Papersa
% WARRb
% Biomedicalc
World
22,593
100
0.73
AFR-D
34
AFR-E
87
AMR-A
13,243
AMR-B
524
0.15 0.39 58.6 2.32
0.44 0.38 0.98 0.79
AMR-D
3
0.01
0.15
EMR-B
32
0.14
0.23
EMR-D
22
EUR-A
6,727
0.10
EUR-B
336
EUR-C
352
1.56
0.60
SEA-B
27
0.12
0.36
SEA-D
184
0.81
0.52
WPR-A
2,164
9.58
0.53
WPR-B
227
1.00
0.33
29.8 1.49
0.24 0.51 0.62
Publication of WARR due to alcohol regions from 1992 to 2003. The data are presented for the whole 12 year period. Data derived from Rajendram et al. (2006) with the kind permission of Oxford University Press and the Medical Council on Alcoholism a Number of publications from alcohol-related research b The alcohol-related publications from each WHO region as a percentage of WARR output c Alcohol-related research publications as a percentage of biomedical research output from that region
7
2,372
134
119
11
73
792
121
EMR-D
EUR-A
EUR-B
EUR-C
SEA-B
SEA-D
WPR-A
WPR-B
1.58
10.32
0.95
0.14
1.55
1.75
30.91
0.09
0.17
0.00
2.53
57.81
0.42
0.12
100.0
%
WARRb
2000–2003
%
0.31
0.52
0.52
0.31
0.58
0.53
0.50
0.19
0.22
0.00
0.62
0.94
0.40
0.31
0.69
Biomedicalc
63
707
57
10
89
118
2,293
6
11
2
173
4,519
36
15
7,664
Papersa
0.82
9.22
0.74
0.13
1.16
1.54
29.92
0.08
0.14
0.03
2.26
58.96
0.47
0.20
100.0
%
WARRb
1996–1999 %
0.30
0.51
0.50
0.44
0.48
0.66
0.51
0.19
0.24
0.28
0.76
1.00
0.48
0.59
0.74
Biomedicalc
43
665
54
6
144
84
2,062
9
8
1
157
4,288
19
10
7,255
Papersa
0.59
9.17
0.74
0.08
1.98
1.16
28.42
0.12
0.11
0.01
2.16
59.10
0.26
0.14
100.0
% WARRb
1992–1995
0.38
0.57
0.53
0.34
0.73
0.67
0.53
0.33
0.23
0.17
1.00
0.99
0.26
0.43
0.77
% Biomedicalc
Publication of WARR due to alcohol in the 14 WHO regions during 1992–2003. Adapted from Rajendram et al. (2006) with the kind permission of Oxford University Press and the Medical Council on Alcoholism. The data are presented in three four-year periods a Number of publications from alcohol-related research b The alcohol-related publications from each WHO region as a percentage of WARR output c Alcohol-related research publications as a percentage of biomedical research output
0
13
EMR-B
194
AMR-B
AMR-D
32
4,436
9
AFR-D
AMR-A
7,674
World
AFR-E
Papersa
Years Region
. Table 95-3 Publication of WARR (1992–2003)
Burden of Disease Due to Alcohol and Alcohol Related Research
95 1641
1642
95
Burden of Disease Due to Alcohol and Alcohol Related Research
. Figure 95-3 A comparison of worldwide alcohol-related research (WARR) and biomedical research (divided by 150; biomed/150) published during 1993–2002. Papers from WARR listed in the Science Citation Index (SCI) and Social Sciences Citation Index (SSCI) databases compared to biomedical research in the SCI, three-year running means during 1993–2002. From Rajendram et al. (2006), with the kind permission of Oxford University Press and the Medical Council on Alcoholism
. Figure 95-4 The numbers of ARR papers from various regions in published between 2000 and 2003
7
Assessment of Regional Contribution and Relative Commitment to WARR
We identified the countries of origin of each paper from the addresses of the authors listed in the SCI and the SSCI. The contributions of the 14 WHO regions stratified on the basis of geography and mortality were determined (See > Table 95-1 for the countries in each region; WHO, 2000). We used > integer counting for these analyses; so a paper from several centers in both Australia and the UK was counted as one each for both the EUR-A and WPR-A regions. > Figure 95-4 shows the numbers of ARR papers from various regions published during 2000–2003.
Burden of Disease Due to Alcohol and Alcohol Related Research
95
The relative percentage presence of each region in WARR was then compared with its presence in global biomedical research to show its relative commitment (RC) to ARR. For example, the UK published about 6% of WARR but just over 10% of all biomedical research, so the RC of the UK to ARR is 0.6 (Rajendram et al., 2006). This method of > bibliometric analysis which has been described previously (Lewison and Wilcox-Jay, 2003) aims to compensate for the tendency of researchers in some countries to publish papers in local non-English language journals not indexed in the SCI and SSCI databases. However, some bias against non-English language clinical and social science research remains.
8
Regional Contribution and Relative Commitment to WARR
The regional commitment to ARR as a percentage of their biomedical output (relative commitment) varied from almost nothing (AMR-D) to just under 1% (0.94% AMR-A; > Figure 95-6). The relative commitment to ARR declined in all of the WHO regions although in most regions the actual number of publications increased. Nearly 60% of all ARR papers originated from AMR-A (mainly Canada and the United States). A further 30% were from Western Europe (EUR-A) and 10% originated from the WPR-A region (mainly Australia and Japan). The rest of the world contributed to only 3% of all ARR. > Figure 95-5 and > Table 95-4 compare the ARR outputs from each of the WHO regions as a percentage of WARR with the corresponding disease burden due to alcohol as a percentage of the GBD due to alcohol. For all but three (AMR-A, EUR-A and WPR-A) the contribution to GBD due to alcohol is far greater than their contribution to WARR. > Figure 95-5 and > Table 95-4 compare the regional relative commitment to ARR (as a percentage of regional biomedical research published during 2000–2003) to the percentage of the total regional burden of disease attributable to alcohol in 2001. The world ratio of ARR (as a percentage of global biomedical research published during 2000–2003) to the percentage of GBD due to alcohol in 2001 was 0.17 (> Table 95-4 Ratio RBDA). This suggests that WARR activity is only one sixth of that appropriate to the GBD due to alcohol. Although the ratio was less than one in all of the 14 WHO regions, the discrepancy between ARR and burden of disease is of particular concern in Europe (EUR-A, ratio = 0.07; EUR-C, 0.04) and Latin America (AMR-B, 0.05; AMR-D, 0). The ratio was highest in the Middle East (EMR-D; 0.63) where the disease burden due to alcohol is particularly low (0.3% DALYs; WHO, 2002).
9
Global Expenditure on Alcohol Related Research
We estimated global expenditure on ARR on the basis of a mean cost per published paper, assuming that ARR is typical of biomedical research in general. The estimated worldwide biomedical expenditure ($106 billion; Global Forum, 2004) and the number of papers indexed in the SCI per year in 2001 (277,000) suggested an average cost per biomedical paper of $372,000 in 2001 (Rajendram et al., 2006). Given that in 2001 the total number ARR papers were published worldwide was 1924, this suggests that the annual worldwide expenditure on ARR was approximately $730 million in 2001 (Rajendram et al., 2006). This was approximately 0.7% of the annual global biomedical expenditure of $106 billion in 2001. The WHO (2002)
1643
1644
95
Burden of Disease Due to Alcohol and Alcohol Related Research
. Figure 95-5 Alcohol-related research, 2000–2003, compared to the global and regional burdens of disease due to alcohol in 2001 (see > Table 95-1 for codes). The line shows equal percentages of both. From Rajendram et al. (2006) with the kind permission of Oxford University Press and the Medical Council on Alcoholism
estimated that alcohol caused 59 million DALYs in 2001. Expenditure on WARR was therefore approximately $12.4 per DALY due to alcohol (Rajendram et al., 2006). See > Figure 95-6 for a comparison of the expenditure on ARR with that spent research on to malaria, tuberculosis and diabetes.
10
Implications: Is the Glass Half Full or Half Empty?
10.1
The Glass is Half Empty
These data suggest that alcohol-related research is receiving insufficient attention from both researchers and policy-makers, who appear to be having difficulty in stemming the rising tide of alcohol-related harm in many countries. The situation is particularly dire in Eastern Europe and Russia, where consumption of ‘‘surrogates’’ has caused great harm because of their widespread availability and very low prices (McKee et al., 2005; Leon et al., 2007; Parna et al., 2007). The problem is not so much an absence of knowledge of the biomolecular and clinical effects of excess ethanol on the body, which are likely to be the same in most countries (with some notable exceptions because of genetic differences) as the lack of evidence-based policy options for the control of excessive drinking in individual states. Whereas tobacco use is declining in many countries and is very much a minority activity in most industrialized ones, so that sharply rising prices and restrictions on where smoking is permitted do not incur serious opposition, such blanket controls on alcoholic beverages and
Burden of Disease Due to Alcohol and Alcohol Related Research
95
. Table 95-4 Publication of WARR 2000–2003 relative to the burden of disease due to alcohol worldwide in 2001 Years
2001
2000–2003 a
Region
% RBD
Ratio RBDAb
World
4.0
0.17
AFR-D
1.3
0.25
AFR-E
2.1
0.19
AMR-A
7.9
0.12
AMR-B
11.7
0.05
AMR-D
5.7
0.00
EMR-B
0.8
0.27
EMR-D
0.3
0.63
EUR-A
6.7
0.07
EUR-B
6.5
0.08
EUR-C
15.2
0.04
SEA-B
3.4
0.09
SEA-D
1.5
0.34
WPR-A
4.6
0.11
WPR-B
5.6
0.05
Publication of WARR during 2000–2003 relative to the burden of disease due to alcohol in the 14 WHO regions in 2001. Adapted from Rajendram et al. (2006) with the kind permission of Oxford University Press and the Medical Council on Alcoholism a Burden of disease due to alcohol as percentage of regional DALYs attributable to alcohol in 2001 (WHO, 2002) b RBDA represents the ratio of the regional alcohol-related research in 2000–2003 (% biomed) to the regional burden of disease due to alcohol in 2001 as a percentage of total regional burden of disease (Table 95-1)
their consumption are unlikely to be popular where large percentages of the electorate drink, most of them with little apparent ill-effect. What is needed is a good sociological understanding of the factors leading to the abuse of alcohol, which are likely to be highly culturally dependent, and in need of appropriate studies in each locale. Indeed, the UK Prime Minister’s Strategy Unit’s alcohol harm reduction project interim analysis (2003) contained 71 references to papers in peer-reviewed journals, and 50 of these were concerned with the social determinants and effects of alcoholism – mainly domestic violence, and its possible treatment. This strongly suggests that what is needed to influence governments is not so much more research in general as more work on the social and economic levers that can be pulled so as to allow policy-makers to respond appropriately to the problems caused by excessive drinking in their countries.
10.2
The Glass is Half Full
Despite the decline in WARR, the past 12 years have seen substantial progress in treatment and strategies to reduce the GBD due to alcohol (Mann et al., 2000). For example brief interventions by family practitioners can reduce alcohol consumption and increase participation in
1645
1646
95
Burden of Disease Due to Alcohol and Alcohol Related Research
. Figure 95-6 Expenditure on research based on bibliometric analysis. This is comparable to that for tuberculosis ($10) and malaria ($7), both widely regarded as under-researched diseases (Global Forum, 2004). However it falls far below expenditure per DALY on diabetes research ($107; Lewison et al., 2004)
treatment programs (Bien et al., 1993). The effectiveness of cognitive behavior therapy, the Alcoholics Anonymous programed (12-step facilitation) and motivational enhancement therapy in inducing abstinence in alcoholics has been confirmed (Project Match Research Group, 1998). In addition, naltrexone and acamprosate have been found to reduce the relapse risk during early abstinence (O’Malley et al., 1996; Sass et al., 1996). Some of the lessons from such research have been brought together in a clinical guideline from the Scottish Intercollegiate Guidelines Network (SIGN, 2003). However many questions remain unanswered and even with an accompanying medical treatment, most alcoholics relapse (Mann et al., 2000). Additional research work on the treatment of people who consume alcohol to excess is therefore needed. A more prominent role for ARR would also have other positive effects. For example, a greater public awareness and understanding could reduce the confusion caused by mixed messages on the health benefits or risks of drinking different amounts of alcohol. Increasing the prominence of researchers of alcohol-related disease could facilitate the development of advisory committees and task forces to influence governmental policy makers. This will help to ensure that alcohol control policies are based on good evidence, and that popular support is forthcoming for the public health campaigns and legislative changes that will be needed in future.
Summary Points Ethanol is one of the most commonly used recreational drugs worldwide. Alcohol is responsible for 4% of the global burden of disease.
Burden of Disease Due to Alcohol and Alcohol Related Research
95
Not all complications related to alcohol result from misuse. Biomedical research and the GBD due to alcohol are increasing whilst the number of papers from ARR remains static.
Nearly 58% of all ARR papers were from the United States or Canada. A further 30% were from Western Europe and 10% originated from Australia, New Zealand or Japan.
The rest of the world contributed to only 3% of all ARR publications. Despite the decline in ARR there has been substantial progress in treatment and strategies to reduce the GBD due to alcohol.
Policies to reduce overall levels of drinking and the rates of some alcohol-related problems have been identified through research and implemented effectively.
Alcohol-related research is necessary, cost-effective and could significantly reduce the GBD due to alcohol. Increasing ARR is therefore likely to benefit society as a whole.
References Bien TH, Miller WR, Tonigan SJ. (1993). Addiction. 88: 315–336. Booyse FM, Parks DA. (2001). Thromb Haemost. 86: 517–528. Catalyst Health Consultants. (2001). Alcohol Misuse in Scotland: Trends and Costs – Final Report. Northwood, Middlesex, Catalyst Health Consultants available at URL http://www.scotland.gov.uk/ health/ alcoholproblems/docs/trco.pdf Deacon L, Hughes S, Tocque K, Bellis MA. (eds.) (2007). Indications of Public Health in the English Regions 8: Alcohol. Association of Public Health Observatories, York UK. Fenoglio P, Parel V, Kopp P. (2003). Eur Addict Res. 9: 18–28. Global Forum for Health Research. (2004). Monitoring Financial Flows for Health Research. Global Forum for Health Research, Geneva. Gross CP, Gerard FA, Powe NR. (1999). N Engl J Med. 340: 1881–1887. Her M, Rehm J. (1998). Addiction. 93: 1335–1340. Lamarre-Cliche´ M, Castilloux A-M, LeLorier J. (2001). Clin Invest Med. 24: 83–89. Leon DA, Saburova L, Tomkins S, Andreev E, Kiryanov N, McKee M, Shkolnikov VM. (2007). Lancet. 369: 2001–2009. Lewison G, Rippon I, de Francisco A, Lipworth S. (2004). Res Eval 13: 181–188. Lewison G, Wilcox-Jay K. (2003). Getting biomedical research into practice: the citations from UK clinical guidelines. Proceedings of the 9th International Conference on Scientometrics and Informetrics, Beijing, China, August; pp. 152–160. Mann K, Hermann D, Heinz A. (2000). Alcohol Alcohol. 35: 10–15.
McKee M, Suzcs S, Sarvary A, Adany R, Kiryanov N, Suburova L, Tomkins S, Andreev E, Leon DA. (2005). Alcohol: Clin Exp Res. 29: 1884–1888. Murray CJL, Lopez AD (eds.). (1996). The Global Burden of Disease: A Comprehensive Assessment of Mortality and Disability from Diseases, Injuries and Risk Factors in 1990 and Projected to 2020. Harvard School of Public Health on behalf of the World Health Organisation and the World Bank, Boston. Murray CJL, Lopez A. (1997). Lancet. 349: 1436–1442. Murray CJL, Salomon JA, Mathers C. (2000). Bull World Health Org. 8: 981–994. O’Malley SS, Jaffe AJ, Chang G, Rode S, Schottenfeld R, Meyer RE, Rounsaville. (1996). Arch Gen Psychol. 53: 217–224. Parna K, Lang K, Raju K, Vaeli M, McKee M. (2007). Int J Public Health. 52: 402–410. Peters TJ, Preedy VR. (1999). Medicine. 27: 11–15. Preedy VR, Watson RR, editors. (2004). Handbook of Alcohol-Related Pathology, vol 1–3. Academic Press, London. Project MATCH Research Group. (1998). Alcohol: Clin Exp Res. 22: 1300–1311. Rajendram R, Lewison G, Preedy VR. (2006). Alcohol Alcohol. 41: 99–106. Rehm J, Baliunas D, Brochu S, et al. (2006a). The Costs of Substance Abuse in Canada 2002. Canadian Centre on Substance Abuse, Ottawa (ON). Rehm J, Patra J, Popova S. (2006b). Addiction. 101: 373–384. Rehm J, Room R, Monteiro M. et al. (2004) Alcohol use. In Comparative Quantification of Health Risks. In: Ezzati M, Lopez AD, Rodgers A, Murray CJL (eds.) Global and Regional Burden of Disease Attributable
1647
1648
95
Burden of Disease Due to Alcohol and Alcohol Related Research
to Selected Major Risk Factors, vol. 1. World Health Organization, Geneva, pp. 959–1109. Rehm JT, Rehn N, Room R, Monteiro M, Gmel G, Jernigan D, Frick U. (2003a). Eur Addict Res. 9: 147–156. Rehm JT, Room R, Monteiro M, Gmel G, Graham K, Rehn N, Sempos CT, Frick U, Jernigan D. (2003b). Eur Addict Res. 9: 157–164. Rehm JT, Sempos CT, Trevisan M. (2003c). J Cardiovasc Risk. 10: 15–20. Sass H, Soyka M, Mann K, Zieglga¨nsberger W. (1996). Arch Gen Psych. 53: 673–680. SIGN (Scottish Intercollegiate Guidelines Network) (2003). The management of harmful drinking and alcohol dependence in primary care – a national clinical guideline. Available at: http://www.sign.ac. uk/pdf/sign74.pdf
The Prime Minister’s Strategy Unit. (2003). Alcohol misuse: How much does it cost? Annexe to the Strategy Alcohol Harm Reduction project Interim Analytical Report. The Cabinet Office. Varney SJ, Guest JF. (2002). Pharmacoeconomics. 20: 891–907. World Health Organization. (2000). The World Health Report 2000 – Health Systems: Improving Performance. WHO, Geneva. World Health Organization. (2002). World Health Report 2002. Reducing Risks, Promoting Healthy Life. WHO, Geneva. World Health Organisation. (2004). Global Status Report on Alcohol 2004. World Health Organisation, Geneva.
96 Years Life Lost Due to Smoking: A Korean Focus S. Yoon 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1650
2
Measuring Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1651
3
Defining Standard Expected Years of Life According to Age . . . . . . . . . . . . . . . . . . . . . 1651
4
Selection of Smoking-Related Diseases or Disease Groups . . . . . . . . . . . . . . . . . . . . . . . 1651
5 Calculation of YLL Due to Premature Death . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1652 5.1 Calculation of Years of Life Lost Due to Premature Death (YLLs) . . . . . . . . . . . . . . . . 1653 6
Estimation of SAF (Smoking-Attributable Fraction) of YLLs Due to Smoking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1653
7
Years of Life Lost Due to Smoking in Korea . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1654
8
Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1654 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1659
#
Springer Science+Business Media LLC 2010 (USA)
1650
96
Years Life Lost Due to Smoking: A Korean Focus
Abstract: This chapter estimates the burden of premature death due to smoking in Korea during the period between 2001 and 2005 using the > years of life lost due to premature death (YLLs) method. To implement this study, we calculated the years of life lost due to premature death by age group using the > standard expected years of life lost (SEYLL) method. YLLs due to smoking were calculated based on the assumptions and methods developed by the > Global Burden of Disease Study Group. The burden of premature death due to smoking was estimated by multiplying the > Smoking-Attributable Fraction (SAF) by the YLLs due to smoking-related diseases. In 2005, the burden of premature death due to smoking using YLLs was 19.2% of YLLs due to all causes in males and 3.3% of YLLs due to all causes in females in Korea. The incidence of premature death due to smoking decreased from 350,800 in 2001 to 341,141 in 2005 in the male population and increased from 36,764 in 2001 to 38,058 in 2005 in the female population in Korea. Our results suggest that the method employed in this study allowed for the burden of premature death due to smoking to be assessed quantitatively, and it was comparable with those used by other international studies in this field and thus can provide a rational basis for national health policy planning regarding premature death from smoking and smokingrelated risk factors in Korea. List of Abbreviations: CEYLL, > cohort expected years of life lost; GBD, global burden of disease; KMOHW, > Korea Ministry of Health and Welfare; KNSO, National Statistical Office of Korea; PAR, > population attributable risk; PEYLL, > period expected years of life lost; PYLL, > potential years of life lost; SAF, smoking attributable fraction; SEYLL, standard expected years of life lost; YLL, Years of life lost
1
Introduction
Cigarette smoke contains some 4,000 chemicals, many of which are toxins that are known to cause malignant tumors (Kaufman et al., 1989; Lee et al., 2006; Nagata et al., 2006; Sturgis and Cinciripini, 2007; Sung et al., 2007), respiratory diseases (Gan et al., 2005) and circulatory diseases (Bazzano et al., 2003; Rijken and Britton, 1998; Saucer et al., 2002). In addition, not only is smoking harmful to the smoker, it is also harmful to others who are chronically exposed to cigarette smoke, especially children (Hackshaw et al., 1997; Park and Kim, 1986). Smoking is now widely regarded as one of the greatest risk factors for mortality. The Korean Association of Smoking and Health estimates that there are approximately 35,000 deaths due to smoking-related diseases each year, and the national economic burden of premature death due to smoking-related diseases exceeded 3 trillion won (approximately 2.5 billion dollars) in Korea in the year 2000 (Ha et al., 2003). The smoking rate among Koreans is one of the highest in the world, with a dramatic increase being seen among teenagers and women. The smoking rate among Korean women was 5.8% in 2005, and the smoking rate among Korean men was a staggering 52.3%, making Korea the nation with the highest male smoking rate in the world (Ministry of Health and Welfare of Korea, 2005). Although the unwanted impact of smoking on health has been the most important public health issue in Korea for quite some time, scientific studies in this field in Korea have been less than satisfactory in qualitative and quantitative terms. The majority of previous studies were focused on estimating the smoking rate or the number of deaths related to smoking.
Years Life Lost Due to Smoking: A Korean Focus
96
The recent development of the concept of and methods used to calculate the number of years of life lost (YLLs) due to premature death has enabled us to calculate the burden of disease due to premature death in more realistic and comparative terms. Furthermore, it is important to mention that significant efforts have been made internationally to improve the quality of the production and collection of vital data in order to enhance the comparability of disease burden measurements between countries. Considering the fact that smoking is the most widespread unhealthy behavior among the general population, the estimation of the burden of disease due to smoking in a scientific and internationally comparable manner should have a profound impact on the formation of public health policy. In this context, the present chapter, which is based on the concepts and methods recently developed by the Global Burden of Disease (GBD) Study Group (Murray and Lopez, 1996) for the calculation of YLLs due to smoking, can provide a means of estimating the unwanted smoking-related impact on the health of the Korean population.
2
Measuring Methods
The study procedure consisted of the following five stages. First, nationwide data on the causes of death and the life table from the > Korea National Statistical Office (KNSO) for the period between 2001 and 2005 were used to analyze the standard expected years of life, the age and gender of the deceased and the cause of death in each age group. Second, diseases or disease groups related to smoking were defined after reviewing all of the available information. Third, the index of standard expected years of life lost (SEYLL) was used to calculate the years lost due to premature death by year and disease group. Fourth, YLLs due to smoking-related diseases were calculated using a functional formula developed by the Global Burden of Disease (GBD) Study Group (Murray and Lopez, 1996) using the above estimates. Finally, the smokingattributable fractions (SAFs) of smoking-related diseases were estimated by age and gender, and the burden of premature death due to smoking was calculated by multiplying the YLLs by the SAFs of smoking-related diseases.
3
Defining Standard Expected Years of Life According to Age
The standard expected years of life by age and gender in the Korean population were estimated from the life tables of the Korean population published by the National Statistical Office in 2001 and 2005. To estimate the standard expected years of life in the over-85 age group, a regression equation was developed using age and life expectancy according to the standard life table of the Korean population (KNSO, 2005a; KNSO, 2005b; KNSO, 2005c; KNSO, 2006a; KNSO, 2006b).
4
Selection of Smoking-Related Diseases or Disease Groups
The smoking-related diseases or disease groups to be included for the Korean population were selected based on the following criteria. The diseases included in this study had to be acknowledged as a smoking-related disease or disease group in a previous study. The previous
1651
1652
96
Years Life Lost Due to Smoking: A Korean Focus
studies referred to in this study were primarily performed among the Korean population. Studies proving the relationship between smoking and specific diseases in other populations were also included in this study. Smoking-related malignant neoplasms, cardiovascular diseases and respiratory diseases were identified and considered. The 20 diseases included in this study and listed by disease group are as follows. The smoking-related malignant neoplasms (11 disease groups) included oral and pharyngeal cancer, esophageal cancer, stomach cancer, liver cancer, pancreatic cancer, laryngeal cancer, cancers of the trachea, bronchi and lungs, cervical cancer, renal and other urinary cancers, urinary bladder cancer, and acute myeloid leukemia. The smoking-related cardiovascular diseases (6 disease groups) included ischemic heart disease, other heart diseases, cerebrovascular disease, atherosclerosis, aortic aneurysm, and other arterial diseases. The respiratory diseases (3 disease groups) included pneumonia, influenza, bronchitis, emphysema, and chronic obstructive airway disease. The 35 years and older age group was chosen as the study population since the lag time of smoking effects is believed to be 20 years or more after exposure (> Figure 96-1).
5
Calculation of YLL Due to Premature Death
Previous studies have proposed various indices that can be used to calculate the number of years of life lost due to premature death (Murray and Lopez, 1996). In general, these studies attempted to use these indices to calculate the difference between the standard expected years of life and that associated with premature death due to a disease or disease group at a certain age as the index of disease burden. Four typical indices were introduced in previous studies: potential years of life lost (PYLL), period expected years of life lost (PEYLL), cohort expected years of life lost (CEYLL), and SEYLL.
. Figure 96-1 Prevalence of current smoking between 1989 and 2005 in a population aged 20–59 years (unit: %). Source: Korean National Health and Nutrition Examination Survey (KNHANES III, 2005) - Health Behaviors of Adults- pp 55
Years Life Lost Due to Smoking: A Korean Focus
96
In the present study, SEYLL was used to measure the burden of disease due to premature death. SEYLL is expressed in the following equation: X L SEYLL ¼ dx ex x¼0
where dx is the number of deaths and e*x is the standard expected years of life at the corresponding age.
5.1
Calculation of Years of Life Lost Due to Premature Death (YLLs)
In order to calculate YLL from SEYLL, a number of decisions had to be made, such as whether to apply different weights to different ages, whether to apply a discounting rate on time lag differences, and how to apply the above assumptions to the calculation of SEYLL. The parameters developed by the Global Burden of Disease (GBD) Study Group (Murray and Lopez, 1996) were adopted in this study in order to improve the comparability of the study result between different countries. In other words, an age differential weight of 0.04 was used as a variable, with a correcting constant of 0.1658 being applied, and the discount rate was 3.0% per year of age. The functional equation for determining the YLLs is as follows: YLL ¼
KCera ðg þ bÞðLþaÞ ½e ½ðg þ bÞðL þ aÞ 1 eðg þ bÞa ½ðg þ bÞa 1 gb2 1K ð1 egL Þ þ g
where r is the discount rate, b is the age weight parameter, K is the modulation factor (which is 1 when using age weight and 0 when not using it), C is a correcting constant applied to changes in the overall burden of disease, a is the age at the time of death, and L is the standard life expectancy at the corresponding age.
6
Estimation of SAF (Smoking-Attributable Fraction) of YLLs Due to Smoking
An equation based on an epidemiological model developed by the Centers for Disease Control and Prevention, USA (CDC, 2003) was used to measure the smoking-attributable fraction of mortality. It categorizes cigarette smoking status as current, former and never. The equation is given below. SAF ¼ ½ðp0 þ p1ðRR1Þ þ p2ðRR2ÞÞ 1=½ p0 þ p1ðRR1Þ þ p2ðRR2Þ
P0
Percentage of adult never smokers in study group
P1
Percentage of adult current smokers in study group
P2
Percentage of adult former smokers in study group
RR1
Relative risk of death for adult current smokers relative to adult never smokers
RR2
Relative risk of death for adult former smokers relative to adult never smokers
1653
1654
96
Years Life Lost Due to Smoking: A Korean Focus
where, SAF is the fraction of the burden that is attributable to smoking, P is the proportion of the population exposed, and RR is the relative risk of death in the exposed population compared with the unexposed population. In calculating the SAF due to smoking, three parameters should be considered: the number of deaths, the relative risk of smoking among different smoking-related diseases, and the smoking rate. The number of deaths according to age, gender and cause of death are published in the annual report of the Korea National Statistical Office. The relative risks of death from cancer were mostly obtained from a study conducted in Korea (Jee et al., 2004a; Jee et al., 2004b). However, the previous study only reported the risks associated with very limited types of cancer in women. The relative risks of death from other cancers and diseases were obtained from the CDC in the USA and Germany (CDC, 2003; Neubauer et al., 2006). The smoking rates of the Korean population by age and gender were extracted from the national health survey conducted by the Korea Ministry of Health and Welfare (KMOHW) in 2005 (Ministry of Health and Welfare of Korea, 2005). An average 20 years of lag time was applied when calculating the SAF values of smoking-related diseases or disease groups with a chronic degenerative nature. The SAF values according to age, sex and disease group are shown in > Table 96-1. Finally, the SAF values of the YLLs due to smoking were calculated by multiplying the SAF values by the YLLs due to smoking.
7
Years of Life Lost Due to Smoking in Korea
The YLLs due to premature death resulting from smoking were calculated from the SEYLL based on the life table prepared by the National Statistical Office during the study period. The YLLs due to premature death resulting from smoking in Korean males increased from 224,537 in 2001 to 357,679 in 2004. The YLLs due to smoking in the male population increased for all smoking-related diseases except for esophageal cancer, stomach cancer, liver cancer, laryngeal cancer, cerebrovascular diseases, bronchitis, and emphysema (> Table 96-2). The YLLs due to smoking in females increased from 36,764 in 2001 to 38,058 in 2005, with an average annual rate of increase of 323 YLLs due to all types of smoking-related diseases except for esophageal cancer, laryngeal cancer, cerebrovascular disease, bronchitis and emphysema (> Table 96-3). The population attributable rate of YLLs due to smoking (i.e., the YLLs due to preventable premature death due to smoking in men.) was 350,800 (19.5% of YLLs due to all causes) in 2002 and decreased to 341,141 (19.2% of YLLs due to all causes) in 2005 (> Tables 96-2 and > 96-4), which means only by smoking, amount of 341,141years’ premature life lost occurs in the year of 2005 in male and 38,058 in female. The YLLs in women increased from 36,764 (3.5% of YLLs due to all causes) in 2001 to 38,058 (3.3% of YLLs due to all causes) in 2005 (> Tables 96-3 and > 96-5).
8
Discussion
The results of this study showed that the YLLs due to smoking increased between 2001 and 2004 in both males and females. The rate of smoking in females was especially high during the 1990s, and considering the time lag of 20–30 years between the initiation of smoking and the onset of related diseases, it is anticipated that the YLLs due to smoking in females will
40–49
60–69
70–79
80þ
47.57
50.51
6.70
6.76
28.16
4.63
24.28
52.73
3.24
24.69
9.28
87.30
93.07
33.41
34.32
76.85
45.20
20.69
47.20
31.63
22.49
49.47
23.02
40.00
52.22
NA
68.81
78.56
23.61
21.30
30.77
61.63
85.11
50.95
53.30
7.80
7.86
31.41
5.44
17.02
40.41
3.74
21.26
47.20
2.05
10.36
2.37
7.40
11.38
49.45
12.16
NA
3.68
34.65
24.68
86.25
92.71
31.01
29.45
74.56
41.26
15.28
36.94
28.51
20.34
45.62
22.69
38.04
49.17
NA
65.79
76.62
22.34
19.57
29.04
58.39
83.00
34.51
36.74
4.10
4.13
18.80
2.82
2.88
15.67
1.93
4.03
11.72
1.06
5.55
1.21
3.92
6.11
33.13
6.57
NA
1.90
21.16
14.24
86.03
92.97
29.93
23.92
73.08
37.83
3.38
17.92
26.13
4.63
13.44
24.28
37.92
47.37
NA
63.51
75.66
22.54
18.86
28.94
56.08
80.98
60.26
65.69
9.31
9.47
36.93
6.19
1.74
9.93
4.98
2.40
7.28
4.44
15.44
3.24
12.46
15.73
57.55
17.28
NA
5.72
41.48
30.96
85.59
92.85
28.93
21.30
71.93
35.86
4.02
21.09
24.71
6.27
16.40
24.26
37.18
45.98
NA
61.96
74.75
22.09
18.17
28.29
54.48
79.69
74.69
79.23
16.12
16.41
52.59
10.87
7.24
33.41
9.13
11.41
27.04
8.77
26.50
6.05
22.25
26.63
72.32
29.02
NA
10.70
57.55
46.29
96
NA not applicable 1 Proportion of the death in the Korean population that would be saved in the absence of smoking
87.86
37.80
Other arterial disease
Chronic obstructive pulmonary disease
78.24
Aortic aneurysm
93.20
47.85
Atherosclerosis
Bronchitis, emphysema
26.93
Cerebrovascular disease age over 65þ
34.94
56.44
Cerebrovascular disease age between 35 and 64
Pneumonia, influenza
26.16 33.78
Cardio vascular diseases Ischemic heart disease age between 35 and 64 54.93
Other heart disease
52.79
22.80
Acute myeloid leukemia
Ischemic heart disease age over 65
1.94
41.09
Urinary bladder
2.06
6.74
10.08
NA
70.69
Trachea, lung, bronchus
45.87
10.84
NA
3.27
31.36
22.08
54.15
24.27
22.42
Liver 79.70
31.74
Stomach
Larynx
63.67
Esophagus
Pancreas
86.38
Kidney, other urinary
Respiratory diseases
50–59
Male Female Male Female Male Female Male Female Male Female
Lip, oral cavity, pharynx
Female
Cervix uteri
Malignant neoplasm
Male
. Table 96-1 Population attributable risk (PAR) of smoking1 by age group in Korea (unit: %)
Years Life Lost Due to Smoking: A Korean Focus 1655
1656
96
Years Life Lost Due to Smoking: A Korean Focus
. Table 96-2 Years of life lost due to premature death from smoking in males2 2001 Cancer
Lip, oral cavity, pharynx
2003
2004
2005
6,940
8,778
8,626
9,411
8,117
Esophagus
11,325
12,188
11,645
11,677
10,909
Stomach
32,247
32,615
32,223
31,010
30,385
Liver
26,264
26,607
26,205
26,269
25,446
5,471
5,586
5,382
5,705
6,139
Pancreas Larynx Trachea, lung, bronchus Cervix uteri
6,906
6,197
6,035
5,527
5,713
83,861
85,887
86,079
90,658
90,796
NA
NA
NA
NA
0
Kidney, other urinary
2,849
3,205
3,568
3,583
3,490
Urinary bladder
2,749
3,347
3,211
3,005
3,166
477
933
788
1,099
1,121
Acute myeloid leukemia Subtotal Cardiovascular Ischemic heart disease
179,088 185,344 183,762 187,944 185,281 34,073
38,247
36,824
37,427
36,252
Other heart disease
13,606
12,678
11,970
12,645
13,291
Cerebrovascular disease
85,955
86,005
83,052
76,190
62,984
Atherosclerosis
1,257
1,578
1,684
1,427
1,720
Artic aneurysm
1,602
1,818
1,735
1,737
2,070
725
698
631
634
498
Other arterial disease Subtotal Respiratory
2002
137,217 141,023 135,896 130,060 116,815
Pneumonia, influenza
6,914
5,578
5,303
6,982
7,844
Bronchitis, emphysema
4,078
3,485
2,120
1,994
1,919
Chronic obstructive pulmonary disease
23,503
32,032
30,598
32,994
29,283
Subtotal
34,495
41,095
38,021
41,971
39,046
Average annual total
324,537 350,800 367,462 357,679 341,141
NA not applicable 2 Among adults aged 35 years and older
significantly increase in the future. According to the results of this study, the burden of premature death due to smoking using YLLs was 19.2% of YLLs due to all causes in males and 3.3% of YLLs due to all causes in females in Korea in 2005, suggesting that the YLLs of 341,141 in males and 38,058 in females could have been prevented if the population did not smoke. Although SAF, PYLL and smoking rate were used to monitor the burden of disease, the use of YLLs is known to have many advantages when it comes to determining data in quantitative terms and can be used as a tool for comprehensive analysis (Yoon et al., 2000). The burden of premature death due to smoking in Korea according to YLL will make it possible to measure the impact of smoking on the state of public health in Korea, particularly when
Years Life Lost Due to Smoking: A Korean Focus
96
. Table 96-3 Years of life lost due to premature death from smoking in females3 2001 Cancer
2003
2004
2005
Lip, oral cavity, pharynx
457
505
475
513
583
Esophagus
491
456
461
449
411
1,376
1,394
1,393
1,339
1,766
Stomach Liver Pancreas Larynx Trachea, lung, bronchus Cervix uteri Kidney, other urinary Urinary bladder Acute myeloid leukemia Subtotal Cardiovascular Ischemic heart disease Other heart disease Cerebrovascular disease
NA
NA
NA
NA
0
1,407
1,349
1,540
1,491
2,180
432
458
363
312
376
3,379
3,561
3,606
3,687
4,824
623
753
804
777
1,157
34
47
47
44
61
184
160
191
175
231
22
34
39
41
97
8,405
8,718
8,918
8,829 11,687
4,082
4,749
4,620
4,945
4,610
948
948
901
829
1,080
14,831 15,551 15,091 13,747 11,056
Atherosclerosis
76
102
109
118
117
Artic aneurysm
376
282
333
359
488
43
55
46
30
34
Other arterial disease Subtotal Respiratory
2002
Pneumonia, influenza
20,356 21,686 21,100 20,027 17,385 751
758
756
894
Bronchitis, emphysema
1,057
1,076
692
412
455
Chronic obstructive pulmonary disease
6,194
8,464
7,500
6,670
7,317
8,003 10,297
8,948
7,977
8,985
Subtotal Average annual total
1,213
36,764 40,702 38,966 36,833 38,058
NA not applicable 3 Among adults aged 35 years and older
this information is compared with data from other countries, thus allowing for the evaluation of current public health policies and the revision of health policy priorities in Korea. This study has the following limitations. First, there are some anomalies associated with the data used in the study, i.e. the computerized data on the causes of death are not entirely accurate because death certificates are only issued by physicians in approximately 50–60% of cases (Korea National Statistical Office, 1995), and even in cases in which the diagnosis is made by a physician, there is sometimes a discrepancy between the recorded and the actual cause of death (Nam et al., 1996). Additionally, the fact that the classification system of the NSO accepts decrepitude as a cause of death distorts the information on the causes of death in the elderly (Nam et al., 1996). Second, since the relative risk of death from disease among smokers in Korea has not yet been investigated thoroughly in all disease categories, some of the relative
1657
1658
96
Years Life Lost Due to Smoking: A Korean Focus
. Table 96-4 Percentage of total YLLs preventable by non-smoking in males4 (unit: %) 2001 2002 2003 2004 2005 Cancer
Lip, oral cavity, pharynx
86.5
86.5
86.5
86.5
85.4
Esophagus
63.5
63.5
63.5
63.5
62.1
Stomach
30.7
30.7
30.7
30.7
30.9
Liver
22.0
22.0
22.0
22.0
21.5
Pancreas
23.3
23.3
23.3
23.3
23.7
Larynx
79.3
79.3
79.3
79.3
78.8
Trachea, lung, bronchus
70.6
70.6
70.6
70.6
69.3
Cervix uteri
NA
NA
NA
NA
NA
Kidney, other urinary
53.8
53.8
53.8
53.8
52.7
Urinary bladder
40.0
40.0
40.0
40.0
40.2
Acute myeloid leukemia
21.0
21.0
21.0
21.0
22.8
Cardiovascular Ischemic heart disease age between 35 and 64
Respiratory
55.5
55.5
55.5
55.5
54.0
Ischemic heart disease age over 65
23.0
23.0
23.0
23.0
22.0
Other heart disease
33.9
33.9
33.9
33.9
32.2
Cerebrovascular disease age between 35 and 64
58.9
58.9
58.9
58.9
55.0
Cerebrovascular disease age over 65
21.9
21.9
21.9
21.9
18.7
Atherosclerosis
48.2
48.2
48.2
48.2
77.2
Aortic aneurysm
78.0
78.0
78.0
78.0
45.9
Other arterial disease
39.0
39.0
39.0
39.0
35.3
Pneumonia, influenza
34.4
34.4
34.4
34.4
33.7
Bronchitis, emphysema
92.7
92.7
92.7
92.7
93.1
Chronic obstructive pulmonary disease
87.4
87.4
87.4
87.4
87.4
18.1
19.5
20.4
19.8
19.2
Average annual total5 NA not applicable 4 Among adults aged 35 years and older 5 Percentage For Overall total YLLs due to all causes
risk data used in this study were taken from previous studies performed in the United States and Germany, and therefore may not fully represent the actual situation in Korea. In studies related to public health policies, it is becoming more and more important to determine the magnitude of the health hazards and to assess the burden of disease in quantifiable terms. However, tools for measuring the burden of disease with reasonable accuracy have not been readily available in Korea. Thus, it is essential to improve the accuracy of the estimates of the related epidemiological indices in order to improve the utility of studies on the burden of premature death. Based on the premise that the above limitations can be overcome, the scope of research should be broadened in order to evaluate the burden of disease in general and not simply diseases related to smoking, as was the case in this study.
96
Years Life Lost Due to Smoking: A Korean Focus
. Table 96-5 Percentages of total YLLs preventable by non-smoking in females6 (unit: %) 2001 2002 2003 2004 2005 Cancer
Lip, oral cavity, pharynx
20.1
20.1
20.1
20.1
23.2
Esophagus
29.1
29.1
29.1
29.1
32.6
Stomach
2.6
2.6
2.6
2.6
3.6
Liver
NA
NA
NA
NA
NA
Pancreas
9.2
9.2
9.2
9.2
11.6
Larynx
42.7
42.7
42.7
42.7
47.5
Trachea, lung, bronchus
8.7
8.7
8.7
8.7
10.8
Cervix uteri
4.9
4.9
4.9
4.9
7.5
Kidney, other urinary
1.8
1.8
1.8
1.8
2.2
Urinary bladder
7.6
7.6
7.6
7.6
10.1
Acute myeloid leukemia
1.1
1.1
1.1
1.1
2.3 11.2
Cardiovascular Ischemic heart disease age between 35 and 64
7.6
7.6
7.6
7.6
Ischemic heart disease age over 65
8.1
8.1
8.1
8.1
6.1
Other heart disease
2.9
2.9
2.9
2.9
3.4
10.5
10.5
10.5
10.5
15.0
Cerebrovascular disease age over 65
6.3
6.3
6.3
6.3
3.9
Atherosclerosis
4.5
4.5
4.5
4.5
29.2
Aortic aneurysm
4.8
Cerebrovascular disease age between 35 and 64
Respiratory
26.6
26.6
26.6
26.6
Other arterial disease
6.4
6.4
6.4
6.4
7.0
Pneumonia, influenza
6.4
6.4
6.4
6.4
7.0
Bronchitis, emphysema
43.7
43.7
43.7
43.7
53.1
Chronic obstructive pulmonary disease
43.5
43.5
43.5
43.5
49.4
3.2
3.5
3.3
3.2
3.3
Average annual total7 NA not applicable 6 Among adults aged 35years and older 7 Percentage For overall total YLLs due to all causes
Summary Points The smoking rate among Korean men is one of the highest in the world. The unwanted impact of smoking on health has been the most important public health issue in Korea.
This chapter estimated the burden of premature death due to smoking in Korea during the period between 2001 and 2005 by calculating the years of life lost due to premature death.
The burden of premature death due to smoking using YLLs was 19.2% in males and 3.3% in females in Korea in 2005.
The study results can provide a rational basis for international comparison studies regarding premature death from smoking.
1659
1660
96
Years Life Lost Due to Smoking: A Korean Focus
Acknowledgments The author would like to thank Seo, Hyun-Ju and Kyung, Min-Ho for their assistance with the preparation of this manuscript.
References Bazzano LA, He J, Muntner P, Vupputuri S, Whelton PK. (2003). Ann Intern Med. 138: 891–897. Centers for Disease Control and Prevention (CDC). Smoking-attributable mortality, morbidity, and economic costs (SAMMEC): Adult and maternal and child health software. US Department of Health and Human Services, Atlanta, GA, 2003, www.cdc.gov. Gan WQ, Man SFP, Sin DD. (2005). Chest. 127: 558–564. Ha BM, Yoon SJ, Lee HY, Ahn HS, Kim CY, Shin YS. (2003). Public Health. 117(5): 358–365. Hackshaw AK, Law MR, Wald NJ. (1997). Br Med J. 315: 980–988. Jee SH, Ohrr H, Sull JW, Samet JM. (2004a). J Natl Cancer Inst. 24: 1851–1856. Jee SH, Samet JM, Ohrr H, Kim JH, Kim IS. (2004b). Cancer Causes Control. 15: 341–348. Kaufman DW, Palmer JR, Rosenberg L, Stolley P, Warshauer E, Shapiro S. (1989). Am J Epidemiol. 129: 703–711. Korea National Statistical Office. (1995) Korean Standard Classification of Disease, vol. 3. National Statistical Office Republic of Korea. Republic of Korea, pp. 38. Korea National Statistical Office. (2005a) 2001 Life Tables for Korea. National Statistical Office Republic of Korea, http://kosis.nso.go.kr Korea National Statistical Office. (2005b) 2002 Life Tables for Korea. National Statistical Office Republic of Korea, http://kosis.nso.go.kr Korea National Statistical Office. (2005c) 2003 Life Tables for Korea. National Statistical Office Republic of Korea, http://kosis.nso.go.kr Korea National Statistical Office. (2006a) 2004 Life Tables for Korea. National Statistical Office Republic of Korea, http://kosis.nso.go.kr
Korea National Statistical Office. (2006b) 2005 Life Tables for Korea. National Statistical Office Republic of Korea, http://kosis.nso.go.kr Korean National Health and Nutrition Examination Survey (KNHANES III, 2005) - Health Behaviors of Adults- pp 55. Lee H, Yoon SJ, Ahn HS. (2006). Cancer Sci. 97: 530–534. Ministry of Health and Welfare of Korea. (2005). Korean National Health and Nutrition Examination Survey (KNHANES III), pp 55. Ministry of Health and Welfare of Korea. (2005). Korean National Health and Nutrition Examination Survey (KNHANES III), pp 108–112. Murray CJL, Lopez AD. (1996). Rethinking DALY. Harvard University Press, Cambridge MA, pp. 1–98. Nagata C, Mizoue T, Tanaka K, Tsuji I, Wakai K, Inoue M, Tsugane S. (2006). Jpn J Clin Oncol. 36: 387–394. Nam HS, Park KS, Sun BH, Shin JH, Sohn SJ, Choi JS. (1996). Kor J Prev Med. 29(2): 227–238 (Korean Language). Neubauer S, Welte R, Beiche A, Koenig H, Buesch K, Leidl R. (2006). Tob Control. 15: 464–471. Park JK, Kim IS. (1986). Yonsei Med J. 27: 261–270. Rijken B, Britton J. (1998). Eur Respir J. 7: 41–73. Saucer WH, Berlin JA, Strom BL, Miles C, Carson JL, Kimmel SE. (2002). Arch Intern Med. 162: 300–306. Sturgis EM, Cinciripini PM. (2007). Cancer. 110: 1429–1435. Sung NY, Choi KS, Park EC, Park K, Lee SY, Lee AK, Choi IJ, Jung KW, Won YJ, Shin HR. (2007). Br J Cancer. 97: 700–704. Yoon SJ, Kim YI, Kim CY, Chang HJ. (2000). Kor J Prev Med. 33: 231–238 (Korean Language).
97 The Health and Economic Consequences of Smoking and Smoking Cessation Interventions: The Dutch Perspective M. P. M. H. Rutten-van Mo¨lken . T. Feenstra 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1662
2 2.1 2.2 2.3 2.4
Smoking Epidemiology in The Netherlands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1663 Smoking Prevalence Over Time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1663 Smoking Prevalence Among Young People . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1664 Comparison of Smoking Prevalence in EU Countries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1665 Relapse Rate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1665
3 The Health and Economic Impact of Smoking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1665 3.1 The Burden of Smoking in DALYs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1665 3.2 The Burden of Smoking in Monetary Terms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1669 4 Tobacco Control Policy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1671 4.1 Tobacco Control Policy in The Netherlands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1671 4.2 Comparison of Dutch Tobacco Control Policy with other Countries . . . . . . . . . . . . . 1671 5 Cost-Effectiveness of Smoking Cessation Interventions . . . . . . . . . . . . . . . . . . . . . . . . . . 1672 5.1 Cost-Effectiveness of Individual Smoking Cessation Support, Tax Increase and Mass Media Campaigns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1672 5.2 Cost-Effectiveness When Including Costs in Life Years Gained . . . . . . . . . . . . . . . . . . . . 1673 5.3 Health Gains of Smoking Cessation Over Time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1674 5.4 Cost-Effectiveness of Pharmacotherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1674 5.5 Comparison with other Cost-Effectiveness Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1675 5.6 Limited Reach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1676 6
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1676 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1677
#
Springer Science+Business Media LLC 2010 (USA)
1662
97
The Health and Economic Consequences of Smoking and Smoking Cessation Interventions
Abstract: Cigarette smoking causes a high health burden worldwide. This paper reviews trends in cigarette use, the health and economic consequences of smoking, and the costeffectiveness of smoking cessation interventions (> cost-effectiveness analysis) from a Dutch perspective. The Netherlands are in the third phase of the smoking epidemic, but > smoking prevalence is still around 28% in adults. The country ranks in the middle of European countries with respect to the implementation of tobacco control policies. It has been estimated that about 20% of all life years lost, 7% of all disease-year equivalents, 13% of all Disability Adjusted Life Years (DALYs), and 4% of total health care costs in the Dutch population is attributable to smoking. This is far more than the burden attributable to other single risk factors. The overall > life expectancy of a male smoker is 7.7 years less than the life expectancy of a male non-smoker. Hence, there is a lot to be gained from the prevention of smoking. Due to the severely reduced life expectancy, lifetime health care costs are lower in smokers than in never smokers. Nevertheless, strong evidence exists that individual smoking cessation interventions consisting of counseling with or without pharmacotherapy are cost-effective. Collective policy measures like tobacco tax increases and mass media campaigns are probably even more efficient. In The Netherlands, structured stop advice by the general practitioner was found to be cost saving, looking at intervention costs net of savings from a reduced incidence of smoking related diseases. When health care costs for diseases not related to smoking that occur during the life years gained by smoking cessation are included, the cost-effectiveness worsens, but the ratios remain below €20,000 per > Quality Adjusted Life Year (QALY). Despite this evidence, few smokers receive cessation aid and most try to stop without any professional support. Reimbursement of smoking cessation interventions might stimulate their use and indications were found that reimbursement might be cost-effective. To conclude, smoking prevention will result in large health gains, but not in cost savings. Smoking cessation can realize a substantial health gain, even if quit rates are at most 20% and 75% of quitters relapses. Even when accounting for additional costs in life years gained, many smoking prevention and smoking cessation policy interventions are cost-effective with ratios ranging from a few thousand up to €20,000 per QALY. List of Abbreviations: BU, bupropion; COPD, chronic obstructive pulmonary disease; CVD, cardiovascular disease; DALYs, disability adjusted life years; EU, European Union; FCTC, framework convention on tobacco control; GP, general practitioner; HALE, > health adjusted life expectancy; H-MIS, minimal intervention strategy by the general practitioner or his assistant (Huisartsen-Minimale Interventie Strategy); IC, intensive counseling; LE, life expectancy; MMC, mass media campaign; NRT, nicotine replacement therapy; OR, odds ratio; PAR, > population attributable risk; QALYs, quality adjusted life years; RIVM, National Institute for Public Health and the Environment (RijksInstituut voor Volksgezondheid en Milieuhygiene); STIVORO, Dutch Foundation for a Smokeless Future; TC, telephone counseling; WHO, World Health Organization
1
Introduction
Tobacco consumption is one of the major preventable life style risk factors that threaten human health. Current estimates are that worldwide 5.4 million tobacco deaths occur each year. Unless urgent action is taken this is estimated to increase to over 8 million deaths per year by 2030, most of which will occur in low and middle income countries (WHO, 2008).
The Health and Economic Consequences of Smoking and Smoking Cessation Interventions
97
Cigarette smoking, the major form of tobacco consumption, harms nearly every organ of the human body, causing a wide range of diseases and a massive burden of chronic illness (Surgeon General, 2004). The leading causes of death in high income countries, such as chronic obstructive pulmonary disease (COPD), lung cancer, and cardiovascular diseases (CVD) are strongly related to tobacco smoking (Doll et al., 2004). About one-third to half of the smokers who continue to smoke will eventually die from a smoking-related disease, (Peto et al., 1996) on average 10 years prematurely (Doll et al., 2004). It is now well-established in the scientific literature that not only active smoking, but also inhalation of environmental tobacco smoke increases the > relative risk to develop COPD, lung cancer, coronary heart disease and stroke. It was estimated that 79,500 people died in 2002 in the 25 countries of the EU as a result of passive smoking (Smoke-Free-Partnership, 2006). This is a relatively small number compared to the number of deaths from active smoking. Prevalence data are sometimes hard to compare, since countries use different definitions and methods to estimate smoking prevalence. Taking that into account, in 2005 the average prevalence of smoking among European males was 40%, ranging from about 14% in Sweden to over 60% in the Russian Federation. For European women, the average smoking prevalence was lower, 18.2%, ranging from less than 2% in Azerbaijan to about 40% in Austria (WHO, 2007a). Current mortality patterns are indicative of smoking trends two to three decades ago. In most Western European countries male smoking is declining but the rate of female smoking is either static or increasing. However, the most worrying trend is the continuing high proportion of teenage smokers (18–20%). The Global Youth Tobacco Survey showed that the rate of current smoking among 13–15 years old students was 17.9% in the, mainly Eastern European, countries that participated in the survey (Warren et al., 2006). This rate is increasing, especially among young girls (Warren et al., 2008). Many interventions, including collective policy measures – like smoking bans, tax increase, advertising and promotion bans, public health campaigns – and individual interventions – like behavioral counseling and pharmacotherapy – have been shown to be effective in reducing cigarette smoking. Increasing the price of tobacco products is one of the most effective means of reducing cigarette smoking (Jha and Chaloupka, 2000). Individual interventions, such as pharmacological and behavioral therapies, significantly increase the probability of long term smoking cessation. They are cost-effective, and when combined with collective interventions may be even more effective. Nevertheless, there is significant room for improvement because the 12-months continuous smoking cessation rates of the best available interventions usually do not exceed 25%. This chapter aims to present the impact of smoking on the health of the Dutch citizens. We first present the current smoking epidemiology and the trends therein, comparing the Netherlands to other European countries. We than discuss the health and health economic impact of smoking in terms of disability adjusted life years (DALYs) and healthcare costs. In the last section we present the evidence on the effectiveness and cost-effectiveness of preventive smoking interventions and smoking cessation interventions in The Netherlands, including policy measures and individual interventions.
2
Smoking Epidemiology in The Netherlands
2.1
Smoking Prevalence Over Time
The Netherlands is a country in the 3rd phase of the smoking epidemic. Cross sectional data on smoking are gathered yearly on initiation of STIVORO, an independent Dutch foundation
1663
1664
97
The Health and Economic Consequences of Smoking and Smoking Cessation Interventions
that concentrates on smoking and health.(STIVORO, 2008) Their data show that prevalence rates have been consistently declining since the 60s, to a rate of around 28%. In recent years however, the decrease seems to slow down somewhat. Dutch men smoke more than women, but this difference has diminished (See > Figure 97-1).
. Figure 97-1 Percentage of smokers in the population aged 15 years and over between 1970 and 2006. Based on cross sectional data gathered by STIVORO (Source: www.STIVORO.nl). Black line for men, gray line for women
2.2
Smoking Prevalence Among Young People
Taking a closer look at young people, the percentages of smoking adolescents are unstable and it is harder to identify trends. Few adolescents at younger ages are frequent smokers; most of them experiment with smoking and it is therefore hard to estimate prevalence. If any, trends in The Netherlands have been downward over the past decades. Among the 10 to12-year-old less than 10% is experimenting with smoking, while 89% of the adult (ex-)smokers report that they tried their first cigarette before the age of 18. Hence, most people start smoking between the ages of 13 and 18 years. In the age group 15 to 19 years, about 30% is a daily smoker. An additional 10–15% smoke infrequently, but at least once in the last 4 week period. > Figure 97-2 shows the development over age in experimental and daily smoking. The prevalence rates of daily smoking are almost identical for boys and girls. Among the 17 to 19-year-old, about 70% reports that they ‘‘ever’’ smoked a cigarette. This shows that experimenting with cigarettes in youth does not necessarily lead to continued daily smoking in later adult life. Among the 20 to 24-year old, the prevalence rates of daily smoking are 32% in men and 23% in women. Hence, for men these are similar to the 19-year-old, but for women the prevalence rates are lower (STIVORO, 2008; Van Baal et al., 2005).
The Health and Economic Consequences of Smoking and Smoking Cessation Interventions
97
. Figure 97-2 Percentage of daily and experimental smokers per age category in 2004. Based on data gathered by STIVORO (Source: www.STIVORO.nl). Black bar for experimental smoker and gray bar for daily smoker
2.3
Comparison of Smoking Prevalence in EU Countries
Compared to other Western European countries, the percentage of smokers in the Netherlands is relatively high. > Figure 97-3 is based on a rank ordering of the 27 countries in the EU (EU-27) according to smoking prevalence among men and women. The upper part of the figure, which was based on WHO data, shows the countries with the highest prevalence and the lower part the countries with the lowest prevalence. It is important to note that the definition of smokers varied. The Dutch data included all current smokers, while for instance the Northern countries only included daily smokers (RIVM, 2007).
2.4
Relapse Rate
The National Institute of Public Health and the Environment (RIVM) has used the above mentioned STIVORO data to estimate > smoking relapse rate (Hoogenveen et al., 2008). > Figure 97-4 displays the estimated relapse curve for former smokers as a function of time since smoking cessation. On the x-axis time since smoking cessation is displayed in months. The y-axis presents the percentage of quitters that has not yet relapsed. Within 6 months more than 50% of all quit attempts failed. Smokers that have been abstinent for more than 3 years will almost never start again. It was estimated that, in the long run, 25–30% of all quit attempts were successful.
3
The Health and Economic Impact of Smoking
3.1
The Burden of Smoking in DALYs
The health burden of smoking can be measured in several ways, e.g. in terms of morbidity or in terms of mortality. A measure that combines morbidity and mortality is the DALY, the ‘‘disability adjusted life year.’’ DALYs are estimated as the sum of the life years lost due to
1665
1666
97
The Health and Economic Consequences of Smoking and Smoking Cessation Interventions
. Figure 97-3 Smoking prevalence in EU countries in 2003–2005, unless another year is indicated (Source: (WHO, 2007b)). Black lines for men, gray lines for women. The upper part of Figure 97-3 shows countries with the highest prevalence and the lower part shows countries with the lowest prevalence. Figure taken from (RIVM, 2007), based on data from the WHO Health for All database (HFA-DB), Copenhagen, WHO Regional Office for Europe, (http://www.euro.who.int/hfadb) (WHO, 2007b)
. Figure 97-4 Percentage of quitters that does not smoke related to time since successful cessation. Computed by the RIVM, from data gathered by STIVORO (see (Hoogenveen et al., 2008)). Black line for men, gray line for women
The Health and Economic Consequences of Smoking and Smoking Cessation Interventions
97
premature mortality and the number of years lived with a disease, weighted by the average loss in quality of life due to the disease (Murray, 1994). When DALYs are added over all diseases the total burden of disease can be estimated, provided corrections for co-morbidity are made. In The Netherlands, the total disease burden of the 71 most important diseases and health conditions was attributed to various risk factors using population attributable risks and including both current and former smokers (Steenland and Armstrong, 2006). > Table 97-1 below presents RIVM estimates of smoking attributable DALYs in comparison to DALYs attributable to other risk factors (De Hollander et al., 2007).
. Table 97-1 Percentage of the total number of life years lost, disease year equivalents, and DALYs in the Dutch population over 20 years of age that is attributable to various risk factors in 2003 (Source: (De Hollander et al., 2007)) % of total life years lost
% of total disease year equivalents
% of total DALYs
20.9
7.1
13.0
Too much saturated fat
0.9
0.6
0.8
Low fruit intake
3.9
1.4
2.4
Low vegetables intake
2.0
0.8
1.4
Low fish intake
3.3
1.7
2.3
Physical inactivity
4.9
3.5
4.1
Alcohol use
2.7
5.4
4.5
Overweight
5.8
12.7
9.7
High cholesterol
3.3
2.2
2.7
10.8
5.6
7.8
Risk factors Smoking
High blood pressure DALYs disability adjusted life years
This table shows that cigarette smoking is the most important, single cause of morbidity and mortality with 21% of all life years lost, 7% of all disease-year equivalents and 13% of all DALYs in the Dutch population over 20 years of age being attributable to smoking. The WHO has also attributed the burden of disease in terms of DALYs to several risk factors using population attributable risks (WHO, 2007b). Within the EU-25, The Netherlands scored better than the EU-average for all risk factors except smoking. That is, a relatively large burden was attributable to smoking in the Netherlands (> Figure 97-5). However, the use of population attributable risks (PAR) to estimate the burden of disease attributable to risk factors has limitations. The PAR method is essentially static and relates current DALYs to current risk factor prevalence (> cross-sectional study). This is problematic, since current diseases were caused by past smoking behavior rather than current smoking behavior. The outcomes of a PAR-based estimation roughly reflect the risk behavior of several decades ago (Perez-Rios and Montes, 2008; Steenland and Armstrong, 2006). Hence, the relatively high DALY burden of smoking in the Netherlands can be explained from the high smoking prevalence in the Netherlands in the seventies. Another problem with the
1667
1668
97
The Health and Economic Consequences of Smoking and Smoking Cessation Interventions
. Figure 97-5 Burden of disease in DALY’s attributable to different risk factors. The Netherlands and other European countries (EU-15 en EU-25). Figure from (De Hollander et al., 2007), based on data from the WHO (Source: (WHO, 2005)). Black bars for The Netherlands, gray bars for EU-15 and bars with diagonal lines for EU-25; EU European Union; NL Netherlands
PAR estimate is that it ignores competing disease risks, and clustering of risk factors in individuals. Therefore it overestimates the burden of disease that is attributed to a certain risk factor. A non-smoker has a longer and healthier life expectancy than a smoker, but during those additional years diseases will occur and these should be accounted for (Barendregt et al., 1998). Another way to estimate the burden of disease related to smoking avoids some of these problems and links the future disease burden to current smokers using dynamic modeling. This has been done for the Netherlands by Barendregt and Bonneux, (Barendregt et al., 1997) Hoogenveen et al. (2008) and Van Baal et al. (2006, 2008). The latter compared the lifelong disease histories of cohorts of 20 year old smokers, obese people, and ‘‘healthy living’’ people, which is defined as non-smokers with a normal weight. Their results are given in > Table 97-2 below. The figures given in > Table 97-2 as well as our discussion of the economic burden of smoking in the next section mainly focus on 14 of the most important chronic diseases for which indisputable evidence of an increased risk from active smoking is present, i.e., coronary heart disease (myocardial infarction and other coronary heart disease), chronic heart failure, stroke, COPD, diabetes, lung cancer, stomach cancer, larynx cancer, oral cavity cancer, esophagus cancer, pancreas cancer, bladder cancer and kidney cancer (Surgeon General, 2004). The figures provide a conservative estimate of the burden of smoking, since damage from acute illnesses, like flu and pneumonia, was left out, as well as damage from diseases for which the evidence of the relation with smoking is disputed, for instance dementia (Surgeon General, 2004). Finally, damage from passive smoking was not included. The 14 diseases however cover the major part of smoking related health damage (Hoogenveen et al., 2008).
The Health and Economic Consequences of Smoking and Smoking Cessation Interventions
97
. Table 97-2 Life expectancy (LE) and Health adjusted life expectancy (HALE) for men and women, for a cohort of Dutch people aged 20 years at baseline (Source: (van Baal et al., 2006)) ‘‘Healthy living’’ cohort Outcome
Smoking cohort
Diff.
‘‘Healthy living’’ cohort
Men
Smoking cohort
Diff.
Women
LE
63.1
55.4
7.7
65.7
59.4
-6.3
HALE
54.8
46.9
7.8
55.4
49.4
-6.0
LE life expectancy; HALE health adjusted life expectancy; Diff. difference
3.2
The Burden of Smoking in Monetary Terms
The same approaches may be used to attribute healthcare costs to smokers and non-smokers and estimate the costs of smoking. Estimates of the costs of smoking have a long and troublesome history with the outcomes often ostensibly influenced by the sponsor of the study. Good overviews of the foot angles involved in estimating the costs of smoking are for instance found in the article by Barendregt and Bonneux, (Barendregt et al., 1997) with estimates for the Dutch situation, and from an international perspective in the book by Sloan et al. (2004). Like for mortality and morbidity, also for costs, it is important to distinguish between the static approach, attributing current resource use and costs to current smoking prevalence using PARs, and the dynamic approach, that tries to estimate the lifetime costs of smokers, using dynamic models of the relation between smoking and disease. > Table 97-3 below shows the percentage of total healthcare costs attributable to smoking and other risk factors in the Netherlands, using the static approach (De Hollander et al., 2007).
. Table 97-3 Percentage of the total costs of illness in the Dutch population in 2003 that is attributable to various risk factors (Source: (De Hollander et al., 2007)) Risk factors
% of total cost of illness
Smoking
3.7
Physical inactivity
1.4
Alcohol use
0.4
Overweight
2.0
High cholesterol
0.7
High blood pressure
3.3
> Table 97-4 shows the difference in lifetime healthcare costs between a cohort of smokers and a healthy cohort (van Baal et al., 2008). Interestingly smokers are less expensive than non-smokers taking a lifetime perspective. The reason for this is that smokers, on average, have a shorter life expectancy. Meanwhile smokers, if alive, are more expensive than nonsmokers at any given age (Barendregt et al., 1997).
1669
1670
97
The Health and Economic Consequences of Smoking and Smoking Cessation Interventions
. Table 97-4 Expected lifetime healthcare costs for a smoking cohort at age 20 compared to a healthy living cohort (Source: (van Baal et al., 2008)) Outcome measure Remaining life expectancy Expected remaining lifetime healthcare costs (€1,000)
‘‘Healthy living’’ cohort
Smoking cohort
64.4
57.4
281
220
Coronary heart disease
12
14
Stroke
Expected remaining lifetime healthcare costs (€1,000) specified by disease group: 13
12
Chronic obstructive pulmonary disease
1
5
Diabetes
2
2
Musculoskeletal diseases
12
8
Lung cancer
productivity costs, and the costs that smokers infer on their family members and others exposed to passive smoking need to be included. Other cost items are the costs of buying cigarettes and the net effects of tobacco taxes and other payments. For the Netherlands such a complete estimate has not yet been made. For the United States, a comprehensive study found that the societal costs were $220,000 for a smoking man, and $106,000 for a smoking women (price level 2000, discount rate 3%) (Sloan et al., 2004). These amounts included a monetary valuation of the loss of healthy life expectancy. They consisted of: Private costs at $183,000 for men and $86,000 for women; Costs imposed on family members of $29,000 for men and $16,000 for women; External costs outside the family of $8,000 for men and $4,000 for women, net of taxes. Expressed per pack of cigarettes costs were almost $40. The outcomes of the lifetime approach indicate that smoking cessation and prevention will result in huge gains in terms of morbidity and mortality avoided. At the same time, health care costs in the long term may increase. To fully assess the impact of smoking cessation, it needs to be taken into account that there is relapse and there is only a gradual decline of health risks of smokers to that of never smokers. Doing so, the health gains of a Dutch smoker that quits were estimated to be on average about 1 life year and 0.75 QALY. In contrast, preventing an adolescent to become an adult smoker realizes the full lifetime gains as presented in > Table 97-2 (Feenstra et al., 2005a). In section five, the health effects of smoking cessation and smoking prevention are related to their costs, to find the cost-effectiveness of tobacco control policies. But first we briefly describe the tobacco control policy in The Netherlands.
The Health and Economic Consequences of Smoking and Smoking Cessation Interventions
4
Tobacco Control Policy
4.1
Tobacco Control Policy in The Netherlands
97
In June 2003, The Netherlands signed the WHO Framework Convention on Tobacco Control (FCTC). The treaty was ratified early 2005. The Dutch Tobacco policy has three objectives, i.e., stimulating smoking cessation, preventing youth from taking up smoking and protecting nonsmokers against passive smoking. The national government further aims to reduce smoking prevalence from 28% in 2006 to 20% in 2010. To achieve these goals a comprehensive package of instruments is implemented, among which is legislation through the Tobacco Act, which was amended in 2002. This act mandates smoke-free educational and healthcare facilities, public buildings, public transport and indoor workplaces. The Tobacco Act includes a ban on direct advertising on billboards, national TV, radio, magazines and newspapers, but not on direct advertising at the point of sale. Promotion through sponsored events, free distribution by mail or other means or promotional discounts are not allowed, but tobacco products can still appear in films. The cigarette packages need to contain large, clearly visible health warnings, but no pictures. Packages are not allowed to contain deceitful terms like ‘‘low tar,’’ ‘‘(ultra) light,’’ ‘‘mild’’ etc. Cigarettes are not allowed to be sold to people under the age of 16. Among the comprehensive package of instruments to achieve the goals of tobacco policy are various national non-smoking campaigns for different target groups, offered through different channels including television, internet, schools and sports clubs. Smokers have access to toll-free quit lines that offer telephone and internet counseling. Health care providers offer face-to-face smoking cessation support. Nicotine replacement therapy (NRT) is available through pharmacies without a prescription. For bupropion and varenicline a prescription is required. However, the vast majority of smokers undertaking a quit attempt do not receive any professional support and the costs of smoking cessation counseling and pharmacotherapy are not covered by the mandatory public health insurance. In 2004, the Partnership Stop Smoking, a partnership of a large number of professional associations and scientific organizations has issued a clinical guideline for the treatment of tobacco addicts (Partnership Stop Smoking, 2004). This has been the basis for a guideline that was specifically developed for use in primary care (Chavannes et al., 2007). These guidelines aim to stimulate optimal use of currently available means to support smoking cessation.
4.2
Comparison of Dutch Tobacco Control Policy with other Countries
International comparisons have shown that many European countries have more stringent anti-smoking policies than the Netherlands (WHO, 2008). On the Tobacco Control Scale of 2007, which quantifies the implementation of tobacco control policies in 30 European countries, The Netherlands rank 14th (Joossens and Raw, 2007). In addition to the lack of reimbursement for smoking cessation support, the Dutch are especially behind with respect to price increases of cigarettes and a smoking ban in restaurants and bars, which has not become effective until 1 July 2008, together with a tax increase to raise the price of cigarettes. After July 1st designated smoking rooms in restaurants and bars are still allowed with no restrictions on their size. The price of a 20-cigarette pack of the most widely consumed brand is about €4.00 in The Netherlands, which is still relatively low compared to the highest price of €8.17 in
1671
1672
97
The Health and Economic Consequences of Smoking and Smoking Cessation Interventions
Norway. There are a number of countries outside Europe, such as Australia, Canada and the US (particularly California) which have comprehensive tobacco control policies. The percentage of smokers in these countries has decreased to 20% or less. This suggests that such intensive and comprehensive policies are effective (Van der Wilk et al., 2007).
5
Cost-Effectiveness of Smoking Cessation Interventions
5.1
Cost-Effectiveness of Individual Smoking Cessation Support, Tax Increase and Mass Media Campaigns
We have previously published a model-based cost-effectiveness study of five ‘‘face-to-face’’ smoking cessation interventions that are available in The Netherlands, i.e., telephone counseling, minimal counseling by the general practitioner, minimal counseling in combination with NRT, intensive counseling in combination with NRT and intensive counseling in combination with bupropion (Feenstra et al., 2005b). We adopted a healthcare perspective. It was found that, compared to current practice in the year 2000, minimal GP counseling was a dominant intervention, generating both gains in life-years and QALYs and savings that were higher than intervention costs. For the other interventions, incremental costs per QALY ranged from about €1,100 for telephone counseling to €4,900 for intensive counseling with NRT. This study has been updated and extended in 2005 to eight interventions among which there were two interventions at the population level, i.e., mass media campaigns and increased tobacco taxes (Feenstra et al., 2005a). In addition to the costs of the smoking cessation interventions and the health care costs of smoking-related diseases, the updated study also included the healthcare costs during the life years that are gained due to smoking cessation. The eight interventions (see > Table 97-5) were selected because of the proven effectiveness. As in the first study, a dynamic population model, the RIVM Chronic Disease Model, was used to project the impact of smoking cessation with each of the eight interventions on future health and health care costs. More specifically, the model simulates the effects of increased smoking cessation on smoking prevalence and incidence, prevalence, mortality and costs of 14 smoking-related diseases (coronary heart disease, chronic heart failure, stroke, COPD, diabetes, lung cancer, stomach cancer, larynx cancer, oral cavity cancer, esophagus cancer, pancreas cancer, bladder cancer and kidney cancer) as well as total mortality, morbidity and healthcare costs. It estimates the effects of the repetitive application of increased smoking cessations rates for a period of 5 years to a population with a changing demographic and risk factor mix. The individual interventions were assumed to reach 25% of smokers each year. The model accounts for risks of relapse. The incidence of smoking related diseases depend on age, gender, smoking status and time since smoking cessation (Hoogenveen et al., 2008). The results are shown in the table below. In the third column of > Table 97-5, the health outcomes of smoking cessation in terms of QALYs were related to the intervention costs only and the effects on costs of healthcare were ignored. It is obvious that this leads to worse > cost-effectiveness ratios than when savings resulting from a reduction in the incidence of smoking-related diseases are deducted from the interventions costs, as was done in column 4.
The Health and Economic Consequences of Smoking and Smoking Cessation Interventions
97
. Table 97-5 Long-term cost-effectiveness of smoking cessation interventions compared to current practice (100 year projection, price level 2004, 4% discounting costs and effects, effect of the current practice scenario is a 5.1% cessation rate) (Source: (Feenstra et al., 2005a)) Costs per QALY in 2004 Euros
Intervention
12-months continuous abstinence rate
When including intervention costs minus savings due to When including the When the reduced previous costs/ including incidence of savings a´nd intervention smoking-related healthcare costs in cost only diseases life years gained
Individual MC
4.4%
3,900
200
H-MIS
7.9%
3,600
cost savings
9,100 8,800
TC
9%
16,000
12,600
21,500
H-MIS + NRT 13.5%
8,200
4,500
13,400
IC + NRT
22%
11,000
7,700
16,600
IC + BU
17%
11,000
7,700
16,600
Population
Effect
MMC
Decrease in smoking 100–1,000 prevalence from 0.2 to 2.1% in year 1; no effect thereafter
cost savings
5,200–6,100
Tax increase
Decrease in smoking 0 prevalence from 3 to 10% in year 1; thereafter effect is washing out
cost savings
5,100
Individual interventions: MC minimal counseling (short counseling by GP or assistant in a single consultation); H-MIS minimal intervention strategy by the general practitioner or his assistant in 1–2 consultations following a 5-step protocol; TC telephone counseling (intake call of 20 min and 2–8 follow-up calls of 15 min each); H-MIS + NRT minimal counseling plus nicotine replacement therapy for 8 weeks; IC + NRT intensive counseling by a trained counselor (e.g., respiratory nurse) plus nicotine replacement therapy for 12 weeks; IC + BU intensive counseling plus bupropion for 8 weeks. Population interventions: MMC mass media campaign; tax increase a tax increase on tobacco products that translates into a price increase
5.2
Cost-Effectiveness When Including Costs in Life Years Gained
In column 5 of > Table 97-5 the healthcare costs of diseases not related to smoking that occur in the life years gained were included in the cost calculation. This was done to account for the fact that smoking reduction results in a substitution of healthcare costs for ‘‘inexpensive’’ lethal smoking-related diseases towards ‘‘expensive,’’ less lethal non-smoking related diseases in life years gained, as was reported, among others, by Barendregt et al. (1997). This leads to a
1673
1674
97
The Health and Economic Consequences of Smoking and Smoking Cessation Interventions
worsening of all cost-effectiveness ratios. Despite this worsening, cost-effectiveness ratios remained low, with the highest ratio being around €21,500 per QALY for telephone counseling. This relatively limited effect of additional costs in life years gained is related to the fact that costs are presented in > net present values. Most costs in added life years occur far in the future and are therefore heavily discounted. Ultimately, it was concluded from > Table 97-5 that a tax increase was the most efficient intervention with zero intervention costs from the healthcare perspective. Additional tax revenues resulting from a 20% tax increase were about €5 billion. Because the price elasticity of demand for tobacco is below 1, tax revenues will increase as a result of a tobacco tax raise. These revenues outweighed the healthcare costs in life years gained (van Baal et al., 2007). However, these revenues are not part of the healthcare budget. Costs per smoker of a mass media campaigns were relatively low (€3,-) and costs per QALY were below €10,000, but the uncertainty around the effectiveness was large. Therefore, the results were computed for a broad range of reductions in the percentage of current smoking, i.e., between 0.2 and 2.1 percentage points (see > Table 97-5). Costs per smoker of the individual interventions varied from €5- to €400. Compared to current practice, cost-effectiveness ratios varied between about €8,800 for structured GP stop-advice (H-MIS) to €21,500 for telephone counseling. These cost-effectiveness ratios were robust to changes in effectiveness and costs, but not to changes in discount rate and time horizon. The cumulative health gains and the cumulative costs and savings are of course not robust, they depend heavily on the effectiveness of the interventions and the number of smokers using the interventions.
5.3
Health Gains of Smoking Cessation Over Time
> Figure 97-6 displays the annual gain in life years and QALYs over time for a 5-year implementation of either intensive counseling combined with nicotine replacement therapy (IC + NRT) or structured GP counseling (H-MIS). The largest health effects occur about 30–35 years after the intervention when the smokers that received the intervention have become on average middle aged. The health gains approach zero as the cohorts that received the intervention die. Over the first 25 years the gain in QALYs is larger than the gain in life years. This is caused by the reduced incidence of smoking related diseases. However, in the long run the gain in life years is larger than the gain in QALYs, because substitute diseases decrease quality of life in life years gained.
5.4
Cost-Effectiveness of Pharmacotherapy
The most recent pharmacological intervention to support smoking cessation, Varenicline, had not yet been included in the above mentioned studies. The 12-months abstinence rate in people using varenicline was found to be significantly higher than the rate in people using bupropion (OR 1.58, 95% CI 1.22–2.05), (Wu et al., 2006) suggesting that it is the most effective pharmacotherapy currently available. The cost-effectiveness of a single quit attempt (in contrast to repetitive quit attempts) with Varenicline compared to unaided cessation, NRT, bupropion and nortriptyline was assessed with the BENESCO model (Benefits of Smoking Cessation on Outcomes model) (Hoogendoorn et al., 2008), which is a > probabilistic decision analytic model that follows a cohort consisting of 25% of Dutch smokers over a lifetime.
The Health and Economic Consequences of Smoking and Smoking Cessation Interventions
97
. Figure 97-6 Annual number of life years gained (LYG) and quality adjusted life years (QALYs) gained over time for a 5-year implementation of IC + NRT or H-MIS compared to current practice (Source: (Feenstra et al., 2005a)). Straight line for gain in QALYs, dotted line for gain in life years; LYG life years gained; QALY quality adjusted life years gained; IC intensive counseling; NRT nicotine replacement therapy; H-MIS minimal intervention strategy by the general practitioner or his assistant
This model has a more limited scope than the RIVM Chronic Disease Model in that it only includes five smoking-related diseases (COPD, lung cancer, coronary heart disease, stroke and severe asthma exacerbations) and does not take into account the additional costs in life years gained. It does however take account of relapse using rates identical to the rates in the RIVM Chronic Disease Model. It also uses higher RR for the incidence of smoking-related diseases for recent quitters than for long-term quitters (abstinent for >5 years). Varenicline was found to be cost saving when compared to bupropion and NRT. Compared with nortriptyline and unaided cessation, varenicline had cost-effectiveness ratios €1650 per QALY and €320 per QALY, respectively. The cumulative number of incident cases of smoking-related diseases prevented by these interventions is given in the figure below (> Figure 97-7).
5.5
Comparison with other Cost-Effectiveness Studies
The results of the above mentioned studies are in line with the results of other cost-effectiveness studies of individual (Cornuz et al., 2003; Javitz et al., 2004; Song et al., 2002; Woolacott et al., 2002) and population directed smoking cessation interventions (Ratcliffe et al., 1997; Ronckers et al., 2005; Tengs et al., 2001; Vijgen et al., 2008). Although difficult to compare due to methodological differences like the types of costs included, all these studies show that these kind of interventions have consistently low cost-effectiveness ratios, often below €10,000 per QALY. A study by Cornuz et al., comparing the cost-effectiveness of NRT in primary care across countries showed that despite considerable differences in cost-effectiveness between countries, the results were favorable for NRT in all of them (Cornuz et al., 2006). Thus, it can be concluded that these interventions are highly cost-effective.
1675
1676
97
The Health and Economic Consequences of Smoking and Smoking Cessation Interventions
. Figure 97-7 Cumulative number of incident cases of smoking-related diseases (COPD, lung cancer, coronary heart disease, stroke) prevented expressed per 1,000 smokers making a quit attempt in the first year, when each pharmacotherapy is compared to unaided cessation (Source: (Hoogendoorn et al., 2008)). NRT nicotine replacement therapy
5.6
Limited Reach
In practice however, individual smoking cessation support reaches far fewer smokers than the 25% often assumed in these studies (Kotz et al., 2007). It has been estimated that, in The Netherlands, only 2–7% of smokers above the age of 20 receive a short stop-smoking advice from their GP and only 4% gets intensive counseling upon referral by their respiratory physician or cardiologist in combination with NRTor bupropion (Vijgen et al., 2006). So the absolute gain in quality adjusted life years and the amount of savings due to the reduction in the incidence of smoking-related diseases is much smaller than anticipated in the modeling studies. One of the reasons why few smokers receive any (pharmacological) support may be the lack of reimbursement of smoking cessation interventions. Pharmacotherapies are paid out of pocket by the patient. In a randomized controlled trial in 1,266 Dutch smokers the effect of reimbursement of smoking cessation interventions for a period of 6 months was compared to no reimbursement (Kaper et al., 2005, 2006). The 12-month continuous abstinence rate was 5.5% in the intervention group compared to 2.8% in the control group, a risk difference of 2.7 (95% CI 0.5–4.9). The costeffectiveness of reimbursement was estimated to be €1,802 per QALY gained. The increase in costs was about €30 per participant and €1,120 per additional quitter. Although results in terms of quit rates are not spectacular, the study seems to suggest that reimbursement is efficient.
6
Conclusion
In conclusion, the health burden attributable to smoking is far more than that of any other single risk factor. In The Netherlands successful smoking prevention would increases the life
The Health and Economic Consequences of Smoking and Smoking Cessation Interventions
97
expectancy of men with 7.7 years and the life expectancy of women with 6.3 years. Health care costs would increase, since people also need healthcare in these additional life years. Increasing the price of cigarettes is the most cost-effective smoking cessation intervention. Other smoking cessation interventions are also cost-effective, even when the extra costs of care during the years of life gained are included. Nevertheless, these interventions do not reach most smokers. Reimbursement of smoking cessation interventions by health care insurers might increase the cessation rate. However, most health gains would be obtained by preventing young people to become addicted.
Summary Points Although declined since the sixties, smoking prevalence in The Netherlands is high (28% in adults) compared to other Western European countries.
Compared to other risk factors such as high blood pressure, high cholesterol,
and alcohol, cigarette smoking is the most important, single cause of morbidity and mortality. 20.9% of all life years lost, 7.1% of all disease-year equivalents and 13% of DALYs in the Dutch population over 20 years of age is attributable to smoking. In the Netherlands, the overall life expectancy of a male smoker is 7.7 years less than the life expectancy of a non-smoker. The healthy life expectancy is 7.8 years less. Because of a reduced life expectance, lifetime health care costs are lower in a smoking cohort compared to a non-smoking cohort. The gains from smoking cessation are substantially less than the gains from the prevention of smoking. With respect to the implementation of tobacco control policies The Netherlands rank somewhere in the middle of European countries. An increase in tobacco taxes is one of the most cost-effective means to increase smoking cessation. Counseling with or without pharmacotherapy is a cost-effective smoking cessation intervention. In the Netherlands, structured GP-stop advice was the most cost-effective, but most of the more intensive interventions that include pharmacotherapy also have cost-effectiveness ratios below €10,000 per QALY. Even when costs of diseases not related to smoking that occur in the added life years are included, the cost-effectiveness ratios for all types of effective tobacco policy remain favorable, i.e., below €20,000 per QALY. Like in many other countries, individual smoking cessation support with or without pharmacotherapy is not covered by compulsory social healthcare insurance. There is data suggestion that reimbursement would be a cost-effective policy measure. Over the first 25 years after successful smoking cessation, the annual gain in quality of life is higher than the annual gain in length of life, because of the prevention of smokingrelated diseases. More than 25 years after successful smoking cessation, the annual gain in life years is larger than the annual gain in QALYs, because of substitute disease incidence in added years of life.
1677
1678
97
The Health and Economic Consequences of Smoking and Smoking Cessation Interventions
References Barendregt JJ, Bonneux L, van der Maas PJ. (1997). N Engl J Med. 337: 1052–1057. Barendregt JJ, Van Oortmarssen GJ, Van Hout BA, Van Den Bosch JM, Bonneux L. (1998). Math Popul Stud. 7: 29–49, 109. Chavannes NH, Kaper J, Frijling BD, Van der Laan IR, Jansen PWM, Guerrouj S, Drenthen AJM, Bax W, Wind LA. (2007). Huisarts Wetensch. 50: 306–314. Cornuz J, Gilbert A, Pinget C, McDonald P, Slama K, Salto E, Paccaud F. (2006). Tob Control. 15: 152–159. Cornuz J, Pinget C, Gilbert A, Paccaud F. (2003). Eur J Clin Pharmacol. 59: 201–206. De Hollander AEM, Hoeymans N, Melse J, van Oers JAM, Polder JJ. (2007). Care for Health. The 2006 Dutch Public Health Status and Forecasts Report. National Institute for Public Health and the Environment. Bohn Stafleu Van Loghum, Houten. Doll R, Peto R, Boreham J, Sutherland I. (2004). BMJ. 328: 1519–1528. Feenstra T, Van Baal P, Hoogenveen R, Vijgen S, Stolk E, Bemelmans W. (2005a). Cost-Effectiveness of Interventions to Reduce Tobacco Smoking in The Netherlands. An Application of the RIVM Chronic Disease Model. Report 260601003. National Institute for Public Health and the Environment (RIVM), Bilthoven. Feenstra TL, Hamberg-van Reenen HH, Hoogenveen RT, Rutten-van Molken. MPMH (2005b). Value Health. 8: 178–190. Hoogendoorn M, Welsing P, Rutten-van Mo¨lken MPMH. (2008). Curr Med Res Opin. 24: 51–61. Hoogenveen R, Van Baal P, Boshuizen H, Feenstra T. (2008). Cost Eff Resour Alloc. 6: 1–15. Javitz HS, Swan GE, Zbikowski SM, Curry SJ, McAfee TA, Decker DL, Patterson R, Jack LM. (2004). Am J Manag Care. 10: 217–226. Jha P, Chaloupka FJ. (2000). BMJ. 321: 358–361. Joossens L, Raw M. (2007). Progress in Tobacco Control in 30 European Countries 2005 to 2007. Swiss Cancer League, Berne. Kaper J, Wagena EJ, van Schayck CP, Severens JL. (2006). Pharmacoeconomics. 24: 453–464. Kaper J, Wagena EJ, Willemsen MC, van Schayck CP. (2005). Addiction. 100: 1012–1020. Kotz D, Wagena EJ, Wesseling G. (2007). Respir Med. 101: 568–573. Murray CJ. (1994). Bull World Health Organ. 72: 429–445. Partnership Stop Smoking. (2004). Guideline for the Treatment of Tobacco Addicts. CBO, Institute for Quality Assurance in Health Care, Alphen aan de Rijn.
Perez-Rios M, Montes A. (2008). BMC Public Health. 8: 22. Peto R, Lopez AD, Boreham J, Thun M, Heath C, Doll R. (1996). Br Med Bull. 52: 12–21. Ratcliffe J, Cairns J, Platt S. (1997). Tob Control. 6: 104–110. RIVM. (2007). Volksgezondheid Toekomst Verkenning, Nationaal Kompas Volksgezondheid (National Public Health Compass) version 3.12. National Institute for Public Health and the Environment, Bilthoven. Ronckers ET, Groot W, Ament AJ. (2005). Med Decis Making. 25: 437–448. Sloan FA, Ostermann J, Picone G, Conover C, Taylor DH. (2004). The Price of Smoking. The MIT press, Cambridge, MA. Smoke-Free-Partnership. (2006). Lifting the Smokescreen. ERSJ Ltd. Reproduction, Brussels. Song F, Raftery J, Aveyard P, Hyde C, Barton P, Woolacott N. (2002). Med Decis Making. 22: S26–37. Steenland K, Armstrong B. (2006). Epidemiology. 17: 512–519. STIVORO. (2008). Website of STIVORO for a Smoke Free Future. Available at: www.stivoro.nl. Surgeon General. (2004). The Health Consequences of Smoking: A Report of the Surgeon General, 2004. Tengs TO, Osgood ND, Chen LL. (2001). Prev Med. 33: 558–570. Van Baal P, Vijgen S, Bemelmans W, Hoogenveen R, Feenstra T. (2005). Potential Health Benefits and Cost Effectiveness of Tobacco Tax Increases and School Intervention Programs Targeted at Adolescents in the Netherlands. National Institute for Public Health and the Environment (RIVM), Bilthoven. Van Baal PH, Brouwer WB, Hoogenveen RT, Feenstra TL. (2007). Health Policy. 82: 142–152. Van Baal PH, Hoogenveen RT, de Wit GA, Boshuizen HC. (2006). Popul Health Metr. 4: 14. Van Baal PH, Polder JJ, de Wit GA, Hoogenveen RT, Feenstra TL, Boshuizen HC, Engelfriet PM, Brouwer WB. (2008). PLoS Med. 5: e29. Van der Wilk E, Melse J, Den Broeder J, Achterberg P. (2007). Leren van de buren. Beleid publieke gezondheid internationaal bezien: roken, alcohol, overgewicht, depressie, gezondheidsachterstanden, jeugd, screening. Report 70051010. National Institute for Public Health and the Environment (RIVM), Bilthoven. Vijgen S, Van Gelder BM, Van Baal PHM, Zutphen M, Hoogenveen RT, Feenstra TL. (2006). Kosten en effecten van een pakket maatregelen voor tabaksontmoediging. National Institute for Public Health and the Environment (RIVM), Bilthoven.
The Health and Economic Consequences of Smoking and Smoking Cessation Interventions Vijgen SMC, van Baal PHM, Hoogenveen RT, de Wit GA, Feenstra TL. (2008). Health Educ Res. 23(2): 310–318. Warren CW, Jones NR, Eriksen MP, Asma S. (2006). Lancet. 367: 749–753. Warren CW, Jones NR, Peruga A, Chauvin J, Baptiste JP, Costa de Silva V, El Awa F, Tsouros A, Rahman K, Fishburn B, Bettcher DW, Asma S. (2008). MMWR Surveill Summ. 57: 1–28. WHO. (2005). The European Health Report 2005. Public Health Action for Healthier Children and Populations. World Health Organization, Regional Office for Europe, Copenhagen. WHO. (2007a). World Health Organization, Regional Office for Europe. The European Tobacco Control
97
Report 2007. World Health Organization, Regional Office for Europe, Copenhagen. WHO. (2007b). World Health Organization. Health for All database (HFA-DB). World Health Organization, Regional Office for Europe, Copenhagen. WHO. (2008). WHO Report on the Global Tobacco Epidemic, 2008. World Health Organization, Geneva. Woolacott NF, Jones L, Forbes CA, Mather LC, Sowden AJ, Song FJ, Raftery JP, Aveyard PN, Hyde CJ, Barton PM. (2002). Health Technol Assess. 6: 1–245. Wu P, Wilson K, Dimoulas P, Mills EJ. (2006). BMC Public Health. 6: 300.
1679
98 Life Years Saved, QualityAdjusted Life Years Saved and Cost-Effectiveness of a School-Based Tobacco Prevention Program L. Y. Wang . C. Linda . R. Lowry . G. Tao 1 1.1 1.2 1.3
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1682 Objective of the Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1683 Project TNT and its Efficacy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1684 Focus of the Cost-Effectiveness Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1684
2 2.1 2.2 2.3 2.4 2.5
Analytical Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1685 Estimating Intervention Costs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1685 Projecting Number of Established Smokers Prevented . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1686 Estimating LYs and QALYs Saved . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1688 Estimating Medical Costs Averted . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1692 Estimating Cost-Effectiveness of Project TNT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1692
3
Sensitivity Analyses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1692
4
Results of Base-Case Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1693
5
Results of Sensitivity Analyses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1694
6
Summary and Implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1694 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1697
#
Springer Science+Business Media LLC 2010 (USA)
1682
98
Life Years Saved, Quality-Adjusted Life Years Saved, and Cost-Effectiveness
Abstract: Project Toward No Tobacco Use (TNT) is a school-based education curriculum designed to prevent tobacco use among junior and senior high school students. An efficacy evaluation of Project TNT showed that a combined-strategy curriculum was effective in reducing both trial and weekly cigarette use. During the 2 years, trial cigarette use among intervention students increased from 37 to 53%, and weekly use increased from 6 to 10%. Among control students, trial cigarette use increased from 35 to 58%, and weekly cigarette use increased from 4 to 13%. On the basis of these findings, we conducted an economic evaluation to determine the cost-effectiveness of Project TNT. The benefits measured were life years (LYs) saved, > quality-adjusted life years (QALYs) saved, and medical care costs saved under a 3% discount rate. The costs measured were the costs of Project TNT program. We quantified TNT’s cost-effectiveness as cost per LY saved and cost per QALY saved. Under base case assumptions, at an intervention cost of $16,403, Project TNT prevented an estimated 34.9 students from becoming > established smokers. As a result, we estimated a cost savings of $10,733 accompanying every LY saved and a cost savings of $7,667 accompanying every QALY saved. Results showed TNT to be cost saving over a reasonable range of model parameter estimates. Project TNT is highly cost-effective compared with other widely accepted prevention interventions. School-based prevention programs of this type warrant careful consideration by policy makers and program planners. List of Abbreviations: CDC, Centers for Disease Control and Prevention; HRQL, healthrelated quality of life; IOM, Institute of Medicine; LYs, life years; Project TNT, Project Toward no Tobacco Use; QALYs, quality-adjusted life years; SAMHSA, Substance Abuse and Mental Health Services Administration
1
Introduction
This chapter is based on our previously published study, ‘‘Cost-effectiveness of a school-based > tobacco-use prevention program’’ (Wang et al., 2001), to determine the cost-effectiveness of the Project Toward No Tobacco Use (TNT) which was designed to prevent tobacco use among junior and senior high school students. The chapter makes new contributions by (1) providing more details on how we calculated life years (LYs) saved and (2) deriving more accurate estimates on quality-adjusted life years (QALYs) for established smokers that has further impact on cost-effectiveness estimates. In the original study, we used a published estimate of QALYs per quitter to estimate intervention benefits. However, because established smokers (ever smoked 100 cigarettes) include current smokers and former smokers, the quality of life and > life expectancy of an established smoker are different from those of a quitter or former smoker. In this chapter, we first derived health-related quality of life (HRQL) estimates for smokers and nonsmokers using the Healthy People 2000 > years of healthy life (YHL) measure in conjunction with the 1990 National Health Interview Survey (NHIS) data. Then we combined the estimated HRQL scores with published life expectancy estimates to calculate QALYs gained per established smoker prevented. Using our newly derived estimates, we recalculated the cost-effectiveness of the Project TNT. In addition, we updated the burden of disease from smoking with most recent mortality and morbidity information. TOBACCO USE is widely acknowledged to be the leading cause of preventable death in the United States (CDC, 1989). Approximately 438,000 Americans die each year as a result of smoking, including approximately 38,000 deaths from secondhand smoke exposure; these
Life Years Saved, Quality-Adjusted Life Years Saved, and Cost-Effectiveness
98
deaths have been associated with more than 5.5 million years of potential life lost (CDC, 2005). Direct > medical costs attributable to smoking total at least $75 billion per year; these expenditures plus the productivity losses ($ 92billion) exceeded $167 billion per year. Each day in the United States, approximately 4,000 young people between the ages of 12–17 years initiate cigarette smoking (see > Table 98-1), and an estimated 1,140 young people become . Table 98-1 Key facts on tobacco use by young people In the United States, each day, approximately 4,000 young people between the ages of 12 and 17 years initiate cigarette smoking and 1,140 young people become daily cigarette smokers If current patterns of smoking behaviors continue, an estimated 6.4 million of today’s children can be expected to die prematurely from a smoking-related disease Among high school students, 23% report current cigarette smoking (smoked one or more cigarettes in the previous 30 days) Among high school students, 54% have ever tried cigarette smoking (even one or two puffs) and 16% have smoked a whole cigarette before age 13 years Cigarette smoking by young people leads to immediate and serious health problems, including respiratory and nonrespiratory effects, addiction to nicotine, and the associated risk of other drug use Cigarette smoking causes heart disease, stroke, chronic lung disease, and cancers of the lung, mouth, pharynx, esophagus, and bladder Smoking at an early age increases the risk of lung cancer, and for most smoking-related cancers, the risk rises as the individual continues to smoke The younger people begin smoking cigarettes, the more likely they are to become strongly addicted to nicotine Studies have found nicotine to be addictive in ways similar to heroin, cocaine and alcohol. Of all the addictive behaviors, cigarette smoking is the one most likely to become established during adolescence This table lists the key facts of tobacco use by young people, the risks of initiating smoking at an early age, and the immediate and long-term health effects of tobacco use
daily cigarette smokers (SAMHSA, 2005). Because most daily smokers (82%) begin smoking before age 18 years (CDC, 1994a), school-based policies and interventions have been identified as part of a comprehensive effort to prevent tobacco use among youth, especially when implemented in combination with mass media campaigns and additional community programs to create tobacco-free social norms (CDC, 1994a, b, CDC, 2007). In the past decade, numerous school-based primary prevention programs to reduce tobacco use among youth have been developed and implemented across the United States. These programs can be effective means of preventing tobacco use among youth, especially those programs that focus on counteracting the social influences that may facilitate adolescent tobacco use (Botvin et al., 1995; Dent et al., 1995; Dobbins et al., 2008; IOM, 2007; Sussman, 2001).
1.1
Objective of the Analysis
Because resources to fund school-based tobacco-use prevention programs are limited, recognizing that a program is effective may not be enough to justify its implementation.
1683
1684
98
Life Years Saved, Quality-Adjusted Life Years Saved, and Cost-Effectiveness
Issues of practical concern to policy makers and program planners are cost (i.e., whether they can afford a particular prevention program) and cost-effectiveness (i.e., whether the effects of a program justify the cost of its implementation). Therefore, we conducted an economic evaluation to assess the cost-effectiveness of the Project TNT, a school-based education program designed to prevent tobacco use among junior and senior high school students.
1.2
Project TNT and its Efficacy
Project TNT is a classroom–based curriculum that aims to prevent and reduce tobacco use by students in grades 5–9 (10–14 years old). Project TNT is a comprehensive social skills program comprising activities that counteract normative and informational social influences on tobacco use and misconceptions about the physical consequences of tobacco use. The program teaches refusal skills, awareness of social misperceptions about tobacco use, and misconceptions about physical consequences. A detailed description of the efficacy study design, including the randomization process, is provided elsewhere (Sussman et al., 1995). In general, the efficacy evaluation was designed to test the effectiveness of three separate social influence curricula (a physical consequences curriculum, an informational social influences curriculum, and a normative social influences curriculum) and a fourth combined-strategy curriculum. Forty-eight junior high schools in southern California were assigned randomly to 1 of the 4 curricula or to a ‘‘usual care’’ curriculum. The 10-lesson curricula were first delivered to a cohort of seventh-grade students (most were 12 years old) in 1989, and a 2-lesson booster session was given to the eighth-grade cohort the following year. The baseline data were collected from 6,716 seventhgrade students; 50% of the students were boys; 60%, white; 27%, Hispanic; 7%, black; and 6%, Asian or ‘‘other.’’ Two-year follow-up data were collected from 7,219 ninth-grade students, 65% of whom reported attending a junior high school at which TNT curricula were offered 2 years earlier. The outcome variables tested were changes in trial and weekly cigarette and smokeless tobacco use 2 years after the intervention. Results of the 2-year follow-up study showed that each single-strategy curriculum was effective only on trial tobacco use but that the combined-strategy curriculum was effective on both trial and weekly tobacco use. On the basis of these findings, the combined intervention program was subsequently disseminated. More specifically, of a cohort of 1,234 seventh-grade students who participated in the combined intervention, 770 participated in the 2-year followup survey as ninth-graders. Of 1,956 students recruited as a control group, 1,565 were surveyed at the 2-year follow-up. During the 2 years, trial cigarette use among students participating in the combined intervention increased from 37 to 53%, and weekly use increased from 6 to 10%. Among students in the control group, trial cigarette use increased from 35 to 58%, and weekly cigarette use increased from 4 to 13%. There was no difference in effectiveness by gender.
1.3
Focus of the Cost-Effectiveness Analysis
We used the combined intervention for our economic analyses. The efficacy study showed that Project TNT was effective in preventing both cigarette use and smokeless tobacco use. We focused our study on the prevention of cigarette smoking because there were more detailed
Life Years Saved, Quality-Adjusted Life Years Saved, and Cost-Effectiveness
98
descriptions of mortality directly associated with cigarette smoking than with smokeless tobacco use. Although this efficacy study was somewhat dated, we chose to use Project TNT in our cost-effectiveness study for three main reasons: (1) economic evaluations of behavioral interventions usually are conducted on the basis of results from efficacy or effectiveness studies, and most of the rigorously evaluated school-based tobacco-use prevention programs, including Project TNT, were implemented during the late 1980s and early 1990s (CDC, 2000); (2) Project TNT has been chosen as a model program by the Center for Substance Abuse Prevention and has been recognized by the U.S. Department of Education as an exemplary program; and (3) Project TNT has been implemented as part of a statewide evaluation in Texas with successful results (www.modelprograms.samhsa.gov).
2
Analytical Approach
Because program selection decisions often are made in the interest of society as a whole, we conducted this study from a societal perspective, which considers everyone affected by the intervention and counts the most significant health outcomes and costs that are attributable to the intervention. We used standard methods of > cost-effectiveness analysis and measured benefits in terms of LYs saved, QALYs saved, and lifetime medical costs saved, discounted at a 3% annual rate as recommended by the Panel on Cost-effectiveness in Health and Medicine (Gold et al., 1996). Program costs incurred during the 2-year implementation were included as > intervention costs. All costs were in 1990 dollars to correspond with the timing of the intervention. The cost-effectiveness of Project TNT was compared with the control scenario and was assessed in terms of cost per LY saved and cost per QALY saved. Although nonsmokers generally have longer life expectancies than smokers, no study to our knowledge has examined the impact of primary smoking prevention on life expectancy. To overcome the gaps in existing research, we used an intermediate outcome measure – number of established smokers prevented. Established smokers were defined as current and former smokers who have ever smoked more than 100 cigarettes. We first translated the relative reduction in trial cigarette use and weekly cigarette use into the number of established smokers prevented and then translated the number of established smokers prevented into LYs saved and QALYs saved. To our knowledge, our study is the first one that used such translations. The base-case analysis was conducted in five steps: (1) a retrospective estimation of the intervention costs, (2) an estimation of the number of students prevented from becoming established smokers by age 26 years, (3) an estimation of the number of LYs saved and QALYs saved by the intervention, (4) an estimation of the lifetime medical care costs saved by the intervention, and (5) a calculation of the cost-effectiveness of the intervention. We conducted multivariate and univariate sensitivity analyses to determine the robustness of the base-case analysis and identify the parameters that had the most influence on the results.
2.1
Estimating Intervention Costs
We estimated the direct costs of program delivery (> Table 98-2) incurred in the combined intervention, including the cost of training of health educators, the cost of teaching students, and the cost of materials used. In the trial study, eight schools were assigned to each of the
1685
1686
98
Life Years Saved, Quality-Adjusted Life Years Saved, and Cost-Effectiveness
. Table 98-2 Intervention costsa Intervention
Cost, $
Training of health educators Two health educators received $10/h for 15 days (120 h) of training
2 $10/h 120 h = 2,400
Two health educators received the training at a fee of $56/day for 15 days 2 $56/day (120 h) of training 15days = 1,680 Subtotal
4,080
Teaching Two health educators taught at four schools each for 10 days (80 h) for $10/h
2 4 80 h $10/h = 6,400
Two health educators taught 2-day (16-h) booster sessions at four schools 2 4 16 h each for $10/h $10/h = 1,280 Subtotal
7,680
Materials Two teachers manuals @ $45/manual
90
1,234 students manuals @ $3.69/manual
4,553
Subtotal ($)
4,643
Total
16,403
a
Values were provided by the evaluation study group of the Project TNT This table summarizes the direct costs of program delivery incurred in the combined intervention, including the cost of training of health educators, the cost of teaching students, and the cost of materials used
4 curricula. Nine health educators received 3 weeks of training (120 h) at an hourly rate of $10 before delivering the curriculum. A master trainer charged $500 per day for conducting the training, or $56 per health educator trained. On the basis of the total number of students who received 1 of the 4 interventions (5,263) and the number of students who received the combined intervention (1,234), we estimated that two health educators would actually be needed for the combined intervention only. During the first year of implementation, the 10-lesson combined curriculum was delivered to 45 classes of seventh-grade students with an average of 5.6 classes per school. Each health educator taught in four schools during an 8-week period, 2 weeks for each school. During the second year, 2-lesson booster sessions were delivered to the eighth-grade students at each school. The health educators worked 8 h a day (5–6 h of teaching and 2–3 h of preparation) at an hourly rate of $10. Each health educator received a copy of the teacher manual, which cost $45, and each student received a copy of a student guide book, which cost $3.69.
2.2
Projecting Number of Established Smokers Prevented
As shown in > Figure 98-1, we developed a > smoking progression model to estimate the number of students (of the 770 total participants) who would become established smokers by
Life Years Saved, Quality-Adjusted Life Years Saved, and Cost-Effectiveness
98
. Figure 98-1 Smoking progression model. This figure shows the smoking progression model we developed to project the number of students who would become established smokers by age 26 years in each of two scenarios – intervention and control. Abbreviations used in the table: Exp experimenters; Est established
age 26 years in the intervention scenario and in the control scenario. At the 2-year follow-up, the 770 students were divided into nonsmokers (ever smoked Table 98-4, the life expectancy of a nonsmoker is 2 years longer than that of a former smoker, 7.3 years longer than that of a current smoker. When we discounted those LYs at an annual rate of 3%, we estimated an average gain of 0.27 discounted LYs for a former smoker prevented and 1.09 discounted LYs for a current smoker prevented. Among all smokers, 68% were current smokers and 32% were former smokers (Appendix of Rogers et al.). Thus, for each established smoker prevented, the weighted average of discounted LYs saved is 0.8 (68% 1.09 + 32% 0.27). We calculated the total number of discounted LYs saved by the intervention by multiplying 0.83 and the number of established smokers prevented. In the present study, using the YHL measure (CDC, 1995) in conjunction with the 1990 NHIS (CDC, 1990), we calculated mean HRQL scores by smoking status (current, former smoker, and non-smoker) for men and women aged 25 years and older adjusted for age, race, and body mass index (BMI) status. Then we combined the estimated HRQL scores with the life expectancy estimates from Rogers and Powell-Griner (1991) to calculate QALYs gained per established smoker prevented.
Life Years Saved, Quality-Adjusted Life Years Saved, and Cost-Effectiveness
98
. Table 98-3 Data used to estimate the number of established smokers prevented Parameter definition
Base Symbol case
Intervention students, No.
N
Sources
na
Dent et al. (1995)
Percentage of students who had initiated Pi smoking by age 14 years, intervention group, %
53.0 na
Dent et al. (1995)
Percentage of students who had initiated smoking by age 14 years, control group, %
60.0 57.5–62.5a Dent et al. (1995)
Pc
770
Rangea
Percentage of students who had become Qi weekly smokers by age 14 years, intervention group, %
10.0 na
Dent et al. (1995)
Percentage of students who had become weekly smokers by age 14 years, control group, %
Qc
15.0 13.3–16.7a Dent et al. (1995)
Percentage of nonsmokers at age 14 years who initiate smoking by age 18 years, %
A1
41.4 37.9–45.1a Pierce et al. (1996)
Percentage of nonsmokers at age 18 years who initiate smoking by age 22 years, %
A2
35.5 30.7–40.4a Pierce et al. (1996)
Percentage of nonsmokers at age 14 years B1 who become established smokers by age 18 years, %
8.1
6.2–10.3a Pierce et al. (1996)
Percentage of nonsmokers at age 18 years B2 who become established smokers by age 22 years, %
3.0
1.6–5.2a
Pierce et al. (1996)
Percentage of nonsmokers at age 22 years B3 who become established smokers by age 26 years, %
1.0
0.5–1.5
Assumption
Percentage of experimenters at age 14 years C1 who become established smokers by age 18 years, %
16.2
8.1–24.3
Assumption
Percentage of experimenters at age 18 years C2 who become established smokers by age 22 years, %
6.0
3.0–9.0
Assumption
Percentage of experimenters at age 22 years C3 who become established smokers by age 26 years, %
2.0
1.0–3.0
Assumption
a
95% confidence interval of the base case value This table lists the values and sources of each of these model parameters used in the progression model, including range of values used in sensitivity analyses
To assess the health-related quality of life, National Center for Health Statistics developed a health and activity limitation index for measuring YHL (CDC, 1995). The HRQL was measured on a continuum ranging from death (0.0) to optimal health state (1.0). Its score is derived from answers to two questions asked in the core questionnaire of the NHIS: activity
1689
1690
98
Life Years Saved, Quality-Adjusted Life Years Saved, and Cost-Effectiveness
. Table 98-4 Life years saved per established smoker prevented (discounted at 3%) Femalea
Malea
All
Population number at age 25–29 years Current
3,185,800 3,211,395 6,397,195
Former
1,508,143 1,465,590 2,973,733
Nonsmoker
4,302,456 4,149,613 8,452,069
Life expectancy at age 25 years Current
53.7
45.2
49.4
Former
57.5
51.8
54.7
Nonsmoker
60.6
52.7
56.7
LY lost relative to nonsmokers (lifetime) Current
7.3
Former
2.0
LY lost relative to nonsmokers (lifetime), discounted to age 14 years Current
1.09
Former
0.27
LY saved per established smoker prevented, discounted to age 14 years
0.83
a Values were from Rogers, Powell-Griner (1991) This table shows life expectancy estimates at age 25 years by smoking status, life years lost of former and current smokers relative to nonsmokers, and the discounted average life years saved per established smoker prevented. LY life year
limitations with six response categories (not limited, not limited in major activity but limited in other activities, limited in major activities, unable to perform major activities, or unable to perform instrumental activities of daily living without the help of other persons, and unable to perform self-care activities of daily life without the help of other persons), and self-rated health with five response categories (excellent, very good, good, fair, or poor). The responses to these two questions were used to assign each person to one of 30 health states ranging from 0.1 to 1.0. Using multiple linear regression, we estimated least square-mean (covariate-adjusted means) scores of HRQL by smoking status for men and women aged 25 years and older and controlling for age, sex, race (white, black, and others), and BMI (BMI < 25 and BMI ≥ 25). Because the distribution of HRQL score was skewed, a logarithmic transformation was applied to the data during analysis. This study used SUDAAN statistical software to account for the complex survey design in variance estimates. > Table 98-5 shows HRQL scores and QALYs by smoking status as well as QALYs saved per established smoker prevented. Based on the HRQL scores, we first calculated the QALYs within each age interval (i.e., age 26–29, age 30–34) by smoking status and discounted them to age 14. We then added up the QALYs across all age intervals as the cumulative QALYs after age 26 (16.6 for nonsmokers, 16.2 for former smokers, and 14.7 for current smokers). We estimated QALYs lost of former smokers and current smokers by comparing cumulative QALYs of
Life Years Saved, Quality-Adjusted Life Years Saved, and Cost-Effectiveness
98
. Table 98-5 Health-related quality of life and quality-adjusted life years by smoking status Current Former Nonsmoker HRQL scores 26–29 years
0.870
0.904
0.911
30–34
0.864
0.903
0.905
35–39
0.835
0.893
0.892
40–44
0.805
0.872
0.884
45–49
0.814
0.840
0.853
50–54
0.747
0.821
0.826
55–59
0.729
0.770
0.795
60–64
0.669
0.723
0.754
65–69
0.636
0.705
0.720
70–74
0.724
0.747
0.771
75–79
0.691
0.725
0.745
80+
0.593
0.695
Life expectancy at age 26 years
48.4
53.7
0.701 55.7
QALYs within each age interval, discounted to age 14 years (discounted at 3%) 26–29 years
2.3
2.4
2.4
30–34
2.5
2.7
2.7
35–39
2.1
2.3
2.3
40–44
1.8
1.9
1.9
45–49
1.5
1.6
1.6
50–54
1.2
1.3
1.3
55–59
1.0
1.1
1.1
60–64
0.8
0.9
0.9
65–69
0.7
0.7
0.8
70–74
0.7
0.7
0.7
75–79
0.0
0.6
0.6
80+
0.0
0.0
0.3 16.6
Cumulative QALYs after age 26 years, discounted to age 14 years
14.7
16.2
QALYs lost relative to nonsmokers, discounted to age 14 years
1.9
0.4
QALYs saved per established smoker prevented, discounted to age 14 years
1.4
This table shows health-related quality of life scores and quality-adjusted life years by smoking status as well as quality-adjusted life years saved per established smoker prevented. HRQL health-related qualify of life; QALY quality-adjusted life year. HRQL scores range from the lowest level of function of 0 (dead) to the highest of 1 (not limited, excellent health)
former and current smokers with that of nonsmokers. Based on the QALYs lost of former and current smokers, we estimated the weighted average QALYs saved per established smoker prevented. As shown in > Table 98-4, compared to nonsmokers, current smokers would lose 1.9 QALYs and former smokers would lose 0.4 QALYs after age 26. On average, an estimated
1691
1692
98
Life Years Saved, Quality-Adjusted Life Years Saved, and Cost-Effectiveness
1.4 QALYs would be saved for each established smoker prevented. We calculated the total number of discounted QALYs saved by the intervention by multiplying 1.4 and the number of established smokers prevented.
2.4
Estimating Medical Costs Averted
To estimate the medical costs saved by Project TNT, we needed to know the lifetime medical expenditure associated with becoming a smoker vs. not becoming a smoker. Hodgson’s study (Hodgson, 1992) of the lifetime cost of smoking-related illness had the most suitable estimates for this study. Hodgson used data on the use and costs of medical care and on mortality during each age interval in cross sections of the US population to generate profiles of lifetime health care costs beginning at age 17 years. The profiles, estimated for men and women by age and the number of cigarettes smoked, included the costs of inpatient hospital care, physician services, and nursing home care. Over a lifetime an average male smoker spent $8,638 more than a never smoker for medical care and an average female smoker spent $10,119 more (1990 US $ discounted at 3%). Based on Hodgson’s estimates, the average expected lifetime medical care costs associated with becoming a smoker were $9,379 more than those of not becoming a smoker. We calculated the total medical costs averted by Project TNT as the number of established smokers prevented multiplied by the expected lifetime excess medical care costs per smoker.
2.5
Estimating Cost-Effectiveness of Project TNT
In this study, the cost-effectiveness ratio was calculated as the net cost per LY saved and the net cost per QALY saved. We calculated the net cost by subtracting medical care costs from intervention costs. Most published cost-effectiveness studies of smoking cessation programs for adults do not include medical care cost savings resulting from smoking cessation. Thus, to make the results of this study comparable with these published results, we also presented the cost-effectiveness of Project TNT without medical care costs savings.
3
Sensitivity Analyses
Because the model parameters depended largely on estimates from several studies, we examined the cost-effectiveness ratios for both high and low values of each key parameter in the analysis. Using multivariate and univariate sensitivity analyses to test the robustness of our base-case results and identify parameters that had the most influence on results, we examined 12 key parameters: the hourly pay per health educator, the medical care costs, and the ten parameters (Pc, Qc, A1, A2, B1, B2, B3, C1, C2, C3) that were used to estimate the number of established smokers prevented as presented in > Table 98-3. A multivariate sensitivity analysis was conducted through two steps. First, a computer simulation using SAS (SAS Institute, Cary, NC) was performed to estimate the most and the fewest number of established smokers prevented by varying the values of each of the ten key parameters over a reasonable range. Parameter values for each simulation trial were selected randomly from the two bound values of each parameter. As given in > Table 98-3, for six of
98
Life Years Saved, Quality-Adjusted Life Years Saved, and Cost-Effectiveness
those parameters (Pc, Qc, A1, A2, B1, B2), we assumed that the estimates were normally distributed and used a 95% confidence interval to determine a plausible range for variation. Because no data were available for the other four parameters (B3, C1, C2, C3), we based the lower- and upper-bound estimates on assumptions. Then we estimated the best- and worse-case cost-effectiveness scenarios by varying the estimates of three parameters: the number of established smokers prevented, intervention costs, and medical care costs. Because the hourly pay per health educator in a real-world scenario could be very different from the actual trial scenario, we altered the intervention costs from $16,403 to $36,563 by increasing the hourly pay per health educator from $10 to $30. Although Hodgson’s study (Hodgson, 1992) assessed the medical cost impact of becoming a smoker vs. not becoming a smoker, it did not control for other differences between smokers and never smokers besides smoking that affect medical costs. According to a research reported by Manning et al. (Manning et al., 1989), when other lifestyle choices are controlled, excess lifetime medical costs of smokers compared with nonsmoking smokers (people who are like smokers in age, sex, education, drinking habits, and several other ways, except that they have never smoked) is 87% of the excess lifetime costs of smokers compared with never smokers. For sensitivity analyses, we used Hodgson’s estimate of $9,379 as our upper-bound estimate and used an estimate of $8,160 (87% of $9,379) as our lower-bound estimate.
4
Results of Base-Case Analysis
> Table 98-6 displays the results from both the base-case analysis and the multivariate sensitivity analyses. Under base-case assumptions, at an intervention cost of $16,403 ($13.29 per student), we estimated that the combined intervention would prevent 34.9 students from becoming established smokers. As a result, society could expect to save $327,140 in medical care costs and a total of 29 discounted LYs and a total of 40.5 discounted QALYs. We estimated a
. Table 98-6 Results from base case and multivariate sensitivity analyses Parameters Intervention cost, $ Established smokers prevented Medical care cost saved, $ Discounted LYs saved Discounted QALYs saved Cost per LY saved, $
Base case 16,403.00 34.9
Worst case 36,563.00 19.7
327,139.52 160,991.47 29.0 40.5
16.4 27.6
Best case 16,403.00 51.0 478,329.00 42.3 71.4
10,733.41
7,598.27
10,912.50
7,666.72
4,504.69
6,469.55
Cost per LY saved (excluding medical care costs saved), $
566.59
2,232.73
387.50
Cost per QALY saved (excluding medical care costs saved), $
404.71
1,323.69
229.73
Cost per QALY saved, $
This table displays the cost-effectiveness results from both the base-case analysis and the multivariate sensitivity analyses. LY life year; QALY quality-adjusted life year
1693
1694
98
Life Years Saved, Quality-Adjusted Life Years Saved, and Cost-Effectiveness
cost savings of $10,733 accompanying every LY saved and a cost savings of $7,667 accompanying every QALY saved. When we excluded the medical care costs from the analyses, we estimated that Project TNT would cost $567 per LY saved and $405 per QALY saved.
5
Results of Sensitivity Analyses
On the basis of 1,024 simulation trials in the first step of the multivariate sensitivity analyses, we estimated that the number of established smokers prevented would range from 19.7 to 51.0. From the second step of the multivariate sensitivity analyses, we estimated that the costsavings would range from $7,598 to $10,913 per LY saved and from $4,505 to $6,470 per QALY saved. When medical care costs were excluded from the cost-effectiveness calculation, the estimated cost-effectiveness of the intervention ranged from $388 to $2,233 per LY saved and $230 to$1,324 per QALY saved. These results demonstrated that the cost-effectiveness ratios were robust over a reasonable range of 12 parameter estimates. The intervention can thus be expected to yield net benefits to society under all scenarios considered. > Table 98-7 presents the results of the univariate analysis for each of the 12 key parameters. Although the excess medical costs had slightly more influence than the other parameters on the cost-effectiveness results, they were generally insensitive to the uncertainty in individual parameters.
6
Summary and Implications
This study had some limitations. First, the number of established smokers prevented was modeled rather than directly measured. Second, only one source of data was available for the probabilities of smoking progression by nonsmokers; therefore, we had to use 95% confidence interval estimates for sensitivity analyses. Third, no data were available in the literature to describe the probabilities of experimenters becoming established smokers, so we had to make assumptions for each age interval. Fourth, because the participants of Project TNT were not followed beyond 9th grade, we used the same average transition probability both for the students who received the full prevention program and for the students in the comparison conditions. Hence, it was assumed that the nonsmoking and experimental smoking students at age 14 who had received the full prevention program would have the same progression probabilities as everyone else who did not receive the full intervention. Some data suggest that this assumption may not always be valid if program effects decay rapidly post intervention (i.e., reduced to a non-significant level or disappearing by the end of high school [Flay et al., 1989; Sussman, 2001]). However, a most recent review study of school-based smoking prevention programs suggests that > school-based programs can have significant long-term effects if they: (1) are interactive social influences or social skills programs; (2) involve 15 or more sessions, including some up to at least 9th grade; and (3) produce substantial short-term effects (IOM, 2007). Fifth, this study did not consider the effectiveness of Project TNT on reducing smokeless tobacco use. However, exclusion of such effectiveness should yield conservative estimates of the cost-effectiveness of Project TNT. Sixth, this study did not fully account for all the costs of smoking to society, such as passive smoking, smoking-related fires, and maternal smoking on the health, birth outcomes, and long-term growth of infants.
Life Years Saved, Quality-Adjusted Life Years Saved, and Cost-Effectiveness
98
. Table 98-7 Univariate sensitivity analysis results
Parameter
Parameter rangea
Number of established smokers prevented Lower Upper
Cost-effectiveness ratios (costs per LY saved) Lower
Upper
57.5–62.5a
33.0
36.7
10,761.5
10,701.1
Percentage of students who had become 13.3–16.7a weekly smokers by age 14 years, control group
24.8
45.0
10,860.8
10,503.1
Percentage of nonsmokers at age 14 years 37.9–45.1a who initiate smoking by age 18 years
34.8
34.9
10,733.7
10,732.1
Percentage of nonsmokers at age 18 years 30.7–40.4a who initiate smoking by age 22 years
34.9
34.9
10,733.7
10,733.2
Percentage of nonsmokers at age 14 years 6.2–10.3a who become established smokers by age 18 years
33.8
35.8
10,747.9
10,715.3
Percentage of nonsmokers at age 18 years 1.6–5.2a who become established smokers by age 22 years
34.2
35.3
10,740.2
10,722.1
Percentage of nonsmokers at age 22 years 0.5–1.5 who become established smokers by age 26 years
34.8
35.0
10,735.4
10,732.1
Percentage experimenters at age 14 years 8.1–24.3 who become established smokers by age 18 years
33.7
36.0
10,751.0
10,713.6
Percentage of experimenters at age 18 3.0–9.0 years who become established smokers by age 22 years
34.7
35.0
10,735.4
10,730.5
Percentage experimenters at age 22 years 1.0–3.0 who become established smokers by age 26 years
34.7
35.0
10,735.4
10,730.5
Hourly pay per health educator, $
10.00–30.00
34.9
34.9
10,733.4
10,037.8
Excess medical cost per smoker, $
8,160.00– 9,379.00
34.9
34.9
10,733.4
9,264.8
Percentage of students who initiated smoking by age 14 years, control group
a
95% confidence interval of the base case value This table presents the results of the univariate analysis for each of the 12 key parameters
However, inclusion of the other costs in our study could only improve the cost-effectiveness of Project TNT and will not affect the general conclusions of this study. Even with these limitations, we have been cautious in our approach and have carefully examined the robustness of our results. The sensitivity analyses indicated that the results are
1695
1696
98
Life Years Saved, Quality-Adjusted Life Years Saved, and Cost-Effectiveness
generally robust with respect to all the key sources of uncertainty in the analysis. It is justifiable to conclude that Project TNT is both cost-effective and cost saving under all scenarios considered. The cost-effectiveness of this primary prevention intervention is even more impressive when compared with results of studies of some widely accepted secondary prevention interventions such as breast cancer screening or cervical cancer screening. For example, routine screening for cervical cancer with Papanicolaou testing for all women aged 15–74 years was estimated to cost $22,000 (in 1996 dollars) for every year of life saved, and annual breast cancer screening for women aged 50–69 years was estimated to cost $46,000 (in 1996 dollars) for every year of life saved (Wang et al., 1999). When compared with smoking cessation programs for adults, the cost-effectiveness of Project TNT is still attractive. The cost-effectiveness ratios of $388 to $2,233 per LY saved (excluding medical costs) are generally consistent with those of most smoking cessation programs and, in some cases, more cost-effective. For example, the cost-effectiveness ratio of physicians’ smoking cessation counseling was found to range from $705 to $2,058 (in 1984 dollars, or $1,074–$3,136 in 1990 dollars) per LY saved (Cummings et al., 1989); the cost-effectiveness ratio of nicotine gum as an adjunct to physician’s advice was found to range from $4,113 to $9,473 (in 1984 dollars, or $6,268–$14,436 in 1990 dollars) per LY saved (Oster et al., 1986); and the cost-effectiveness ratio of the nicotine patch with brief physician counseling was found to range from $965 to $2,360 (in 1995 dollars, or $712–$1,742 in 1990 dollars) per LY saved (Wasley et al., 1997). The results of this study suggest that school-based tobacco-use prevention programs can be delivered at a reasonable cost and can be highly cost-effective and cost-saving. Primary prevention programs of this type warrant careful consideration by policy makers and program planners when resource allocation and curriculum decisions are made. The findings of this study also suggest that a school-based primary prevention intervention can be as cost-effective as secondary prevention interventions, such as tobacco-use cessation programs for adults. The recent CDC Best Practices for Comprehensive Tobacco Control Programs provides evidencebased guidance on the levels of investment needed to fund a comprehensive program to prevent tobacco use among youth (CDC, 2007). School-based policies and programs should be part of this comprehensive intervention effort to reduce smoking so substantially that it is no longer a public health problem for our nation (IOM, 2007). In this study, we developed a smoking progression model to project the long-term impact of Project TNT. This model can be used to assess the long-term impact of other smoking prevention programs for young people. Furthermore, such a model can be applied to other risk factors that track from childhood or adolescence to adulthood. For example, a similar model can be developed to assess the long-term impact of obesity prevention programs for adolescents based on the interventions’ short-term impacts. In this study, we used the NHIS data to derive HRQL scores for smokers and nonsmokers. This data set can be used to derive HRQL scores associated with other diseases or conditions. To improve future analyses and to help policy makers with more informed decisions about the prevention of tobacco use among adolescents, additional research into the long-term effects of comprehensive programs to prevent tobacco use among youth, the stages of smoking establishment from adolescence to adulthood, the medical costs of treating smoking-related diseases, and the life expectancy of smokers and nonsmokers is needed. Researchers should routinely include program cost data in their program evaluations so that more economic evaluations of tobacco-use prevention programs can be conducted.
Life Years Saved, Quality-Adjusted Life Years Saved, and Cost-Effectiveness
98
Summary Points School-based tobacco-use prevention programs can be delivered at a reasonable cost and can be highly cost-effective and cost-saving.
School-based tobacco-use prevention programs warrant careful consideration by policy makers and program planners when resource allocation and curriculum decisions are made. School-based primary prevention intervention can be as cost-effective as secondary prevention interventions, such as tobacco-use cessation programs for adults. To reduce overall tobacco use, it is important to increase investment in primary prevention programs for youth. School-based primary prevention programs should be included as part of comprehensive tobacco control programs to significantly reduce the adverse health outcomes of smoking in our society.
References Botvin GJ, Baker E, Dusenbury L, Botvin EM, Diaz T. (1995). JAMA. 273: 1106–1112. CDC. (1989). Reducing the Health Consequences of Smoking: 25 Years of Progress: A Report of the Surgeon General. Washington, DC. CDC. (1990). Interview Survey Part I (Public Use Data Tape Documentation). Hyattsville, MD. CDC. (1994a). Preventing Tobacco Use Among Young People: A Report of the US Surgeon General. Atlanta, GA, pp. 209–292. CDC. (1994b). Morb Mortal Wkly Rep. 43: 1–18. CDC. (1995). Stat Notes. 7: 1–14. CDC. (2000). Reducing Tobacco Use: A Report of the Surgeon General. Washington, DC. CDC. (2005). Morb Mortal Wkly Rep. 54: 625–628. CDC. (2007). Best Practices for Comprehensive Tobacco Control Programs. Washington, DC. Cummings SR, Rubin SM, Oster G. (1989). JAMA. 261: 75–79. Dent CW, Sussman S, Stacy AW, Craig S, Burton D, Flay BR. (1995). J Consult Clin Psychol. 63: 676–677. Dobbins M, DeCorby K, Manske S, and Goldblatt E. (2008). Prev Med. 46: 289–297. Flay BR, Koepke D, Thomson SJ, Santi S, Best JA, Browns KS. (1989). Am J Public Health. 79: 1371–1376. Gold MR, Siegel JE, Russell LB, Weinstein MC. (1996). Cost-effectiveness in health and medicine. Oxford University Press, New York, NY.
Hodgson TA. (1992). Milbank Q. 70: 81–125. IOM. (2007) Ending the Tobacco Problem: A Blue-Print for the Nation. Washington, DC (Appendix D17). Manning WG, Keeler EB, Newhouse JP, Sloss EM, Wasserman J. (1989). JAMA. 261: 1604–1609. Oster G, Huse DM, Delea TE, Colditz GA. (1986). JAMA. 256: 1315–1318. Pierce JP, Choi WS, Gilpin EA, Farkas AJ, Merritt RK. (1996). Health Psychol. 15: 355–361. Rogers RG, Powell-Griner E. (1991). Soc Sci Med. 32: 1151–1159. SAMHSA. (2005) Results from the 2005 National Survey on Drug Use and Health. Rockville, MD. SAMHSA. (2007) Office of Applied Studies, National Survey on Drug Use and Health, 2005 and 2006 (Table 4.5B–4.8B). Sussman S. (2001). Am J Health Behav. 25: 191–199. Sussman S, Dent CW, Burton D, Stacy AW, Flay BR. (1995). Developing School-Based Tobacco Use Prevention and Cessation Programs. Sage, Thousand Oaks, CA. Wang LY, Haddix AC, Teutsch SM, Caldwell B. (1999). Am J Manag Care. 5: 445–454. Wang LY, Crossett LS, Lowry R, Sussman S, Dent CW. (2001). Arch Pediatr Adolesc Med. 155: 1043–1050. Wasley MA, McNagny SE, Phillips VL, Ahluwalla JS. (1997). Prev Med. 26: 264–270.
1697
2
Disease Burdens and Economic Impacts 2.7 Sensory and Musculoskeletal
99 Burden of Disease and Years Lived with Disability Associated with Musculoskeletal Pain M. Suka . K. Yoshida 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1702 2 Population Surveys of the Prevalence of MP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1703 3 Burden of MP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1703 4 Estimation of the YLDs Due to MP in the Japanese Adult Population . . . . . . . . . . . 1707 5 Findings from the Questionnaire Survey on MP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1710 6 YLDs Due to MP in the Japanese Adult Population . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1710 7 Future Prospect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1712 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1715
#
Springer Science+Business Media LLC 2010 (USA)
1702
99
Burden of Disease and Years Lived with Disability Associated with Musculoskeletal Pain
Abstract: > Musculoskeletal pain (MP) is recognized as a problem of global proportions. MP is rarely associated with fatal conditions, however, it causes deterioration in health-related quality of life (QOL). Estimation of the years lived with disabilities (YLDs) due to MP in a population will inform us how much MP deprives the population of health-related QOL. In this chapter, we conduct a review of the literature of population surveys of the prevalence of MP, estimate the YLDs due to MP in the Japanese adult population, and discuss the methodology for the estimation of the burden of MP in the general adult population. Our estimation method, based on the Global Burden of Disease protocol, may be applicable to other populations and may provide a feasible way to make relative comparisons of the burden of MP in different populations. Toward better estimation of the burden of MP in the general adult population, further studies may be required to develop a standard case-finding instrument for MP, to formulate discriminating definitions of acute and chronic MP, to determine the > disability weights for MP in the context of > clustering of MP and pain site, and to update the Global Burden of Disease protocol in an attempt to adequately deal with > comorbidity in the calculation of DALYs. List of Abbreviations: DALY, disability adjusted life year; MP, musculoskeletal pain; QOL, quality of life; YLD, year lived with disability; YLL, year of life lost
1
Introduction
> Musculoskeletal
disorders are a major health problem throughout the world. They have enormous impact on individuals and societies and also on healthcare services and economies. Pain is the most common symptom in people with musculoskeletal disorders. Despite its enormous impact worldwide, musculoskeletal pain (MP) has not been counted among national healthcare priorities in most countries. The Bone and Joint Decade (2000–2010) was launched with the aim of improving the health-related quality of life (QOL) for people with musculoskeletal disorders worldwide. This global campaign contributes to raising awareness of the burden of musculoskeletal disorders, and MP is now recognized as a problem of global proportions (Brooks, 2005; Lidgren, 2003; Woolfe, 2001). Reliable epidemiological estimates of the burden of disease are necessary for evidencebased healthcare. This information is especially helpful in healthcare priority-setting in each country. Various measures have been used to quantify the burden of disease in a population. The Global Burden of Disease Study developed disability adjusted life years (DALYs) as an international measure of the burden of disease (Murray, 1996). It has been adopted by the World Health Organization project. This summary measure indicates the years lost from an ideal life span due to premature mortality (years of life lost; YLLs) and disability (years lived with disability; YLDs). One DALY thus represents the loss of 1 year of equivalent perfect health. MP is rarely associated with fatal conditions; at the same time, it causes deterioration in health-related QOL. Estimation of the YLDs due to MP in a population will inform us how much MP deprives the population of health-related QOL. In this chapter, we describe the estimation of the burden of MP in the general adult population. First, we provide a review of the literature of population surveys on the prevalence of MP. Second, we estimate the YLDs due to MP in the Japanese adult population. Finally, we discuss the methodology for the estimation of the burden of MP in the general adult population.
Burden of Disease and Years Lived with Disability Associated with Musculoskeletal Pain
2
99
Population Surveys of the Prevalence of MP
A health needs assessment is necessary for healthcare priority-setting in each country. The prevalence of disease is the most frequent measure of health needs and enables a nation to identify the most prevalent diseases in the population. Many investigators have reported on the prevalence of MP in the general adult population. > Table 99-1 summarizes the population surveys of the prevalence of MP. Of the nine studies, excluding those that focused on a certain type of MP or included only site-specific data, seven are from European countries and two are from Asia-Pacific countries. There is no study of developing countries. Excluding the study from Norway, all were based on questionnaires, some of which included a pain drawing to report the location of MP. Overall, the prevalence of MP in the general adult population is high with a significant age-dependent increase. The most common sites of MP are back and shoulder, and MP is often present at multiple sites. However, caution is advised when comparing the prevalence between different studies. The definition of MP and the wording of the question varied among the studies. Moreover, the prevalence probably included both acute and chronic MP because of the cross-sectional design of the studies. Even though people reported MP lasting more than 3 weeks, some people may have completely recovered from MP in a short time, whereas others may have suffered MP continuously or recurrently for a long time. We should, therefore, avoid concluding from the results of these studies that Norway has a lower prevalence of MP than other countries.
3
Burden of MP
The prevalence of disease enables us to identify the most prevalent diseases in a population. However, the most prevalent disease does not always have the greatest impact on health-related QOL of people with the disease. For example, in the case of Population X (> Figure 99-1), the prevalence of Disease A is higher than that of Disease B. However, it is questionable whether the burden of Disease A is larger than that of Disease B, because Disease A has a smaller impact on health-related QOL than Disease B. The burden of disease in a population should be measured by the prevalence of the disease in the population and the impact on health-related QOL of people with the disease. DALYs have commonly been used to quantify the burden of disease in a population since the Global Burden of Disease Study reported the first comprehensive estimates of the burden of disease for 107 diseases and injuries (Murray, 1996). The DALYs of a disease in a population are the sum of the YLLs in the population and the YLDs for people with the disease. The basic formula for calculating YLLs is: YLL ¼ N L where N is the number of deaths and L is standard life expectancy at age of death (years). The basic formula for calculating YLDs is: YLDs ¼ I DW L where I is the number of new cases of the disease for a given time interval, DW is disability weights for the disease ranging from 0 (perfect health) to 1 (dead), and L is the average
1703
6,681 (3,252/ 3,374)
1,304 (581/ 723)
5,752 (2,841/ 2,911)
Bromolla and Simrishamn, Sweden
Nationwide, Norway
Regierungsbezirk Karlsruhe, Germany
Greater Manchester, UK
16þ
18–80
16–66
25–74
1,609 (800/ 809)
Setting
Men: 16.1%; Women: 23.9% (crude rates)
49.7% (crude rate)
Prevalence
Questionnaire; Pain in neck, shoulder, elbow, hand, back, hip, or knee for 1 week in the past month
Women Age 16–44, 34.9% Age 45–64, 55.9% Age 65–74, 62.8% Age 75þ, 62.1%
Men Age 16–44, 35.7% Age 45–64, 51.0% Age 65–74, 51.6% Age 75þ, 49.0% Age 75þ, 49.0%
Questionnaire; Prolonged pain in the past 38.3% (crude rate) 6 months
Interview; Pain in neck/upper limb (shoulder, arm, or hand), back, or lower limb (hip or leg) in the past 14 days
Questionnaire; Pain for 3 months
Assessment
Musculoskeletal pain
Andersson (1993)
The most common site of pain Urwin was back, followed by knee and (1998) shoulder. Pain was often present at multiple sites The same research group reported that the prevalence of musculoskeletal pain was higher in ethnic minority populations than in the white population (Allison 2002)
Chrubasik (1998)
The most common site of pain Brage was neck/upper limb, followed (1996) by back and lower limb. Pain was often present at multiple sites
Additional findings
Reference
99
Number (men/ women) Age, y.o.
Study population
. Table 99-1 Population surveys of the prevalence of musculoskeletal pain
1704 Burden of Disease and Years Lived with Disability Associated with Musculoskeletal Pain
3,664 (1,649/ 2,015)
Nationwide, Netherland
Uppland, Sweden 4,506 (2,086/ 2,420)
2,755 (1,304/ 1,451)
Halmstad and Laholm, Sweden
20–64
25þ
20–74
Questionnaire; Pain in head, shoulder, arm, back, or leg in the past 2 weeks
Questionnaire; Pain in neck, higher back, shoulder, elbow, wrist/hand, lower back, hip, knee, ankle, or foot in the past 12 months
Questionnaire; Pain in neck, anterior chest, dorsal chest, shoulder/upper arm, elbow/lower arm, hand, low back, hip/upper leg, knee, or lower leg/foot for 3 months in the past 12 months
Women Age 20–34, 43.5% Age 35–44, 52.7% Age 45–54, 54.2% Age 55–64, 57.3%
Men Age 20–34, 27.5% Age 35–44, 40.1% Age 45–54, 42.5% Age 55–64, 49.6%
Point prevalence: 53.9%; 12-months prevalence: 74.5%; Prevalence of pain lasting 3 months: 44.4% (crude rates)
Men: 30.9%; Women: 38.3% (age adjusted rates)
The most common sites of pain Bingefors were back and shoulder (2004)
The most common site of pain Picavet was lower back, followed by (2003) shoulder and neck. Pain was often present at multiple sites
The most common site of pain Bergman was low back, followed by (2001) shoulder/upper arm and neck.
Burden of Disease and Years Lived with Disability Associated with Musculoskeletal Pain
99 1705
20þ
Niigata, Tsukuba, and Hamamatsu, Japan
Women Age 18–44, 40.0% Age 45–64, 61.5% Age 65þ, 66.7%
Men Age 18–44, 43.3% Age 45–64, 53.3% Age 65þ, 60.0%
Prevalence
Women Age 20–29, 23.0% Age 30–39, 34.9% Age 40–49, 39.2% Age 50–59, 47.5% Age 60–69, 49.7% Age 70þ, 55.2%
Questionnaire; Pain in neck, higher back, Men shoulder, upper arm, elbow, lower arm, Age 20–29, 25.8% wrist, hand, finger, lower back, hip, femur, Age 30–39, 32.6% knee, crus, ankle, or foot for 1 week in Age 40–49, 42.2% the past month Age 50–59, 43.2% Age 60–69, 45.6% Age 70þ, 55.8%
Questionnaire; Pain in neck, shoulder, elbow, hand, back, hip, knee, or foot for 1 week in the past month
Assessment
The most common sites of pain Suka were neck/shoulder and low (2005a) back
The most common sites of pain Tayler were back and shoulder (2005)
Additional findings
Reference
Overall, the prevalence of musculoskeletal pain in the general adult population is high with a significant age-dependent increase. The most common sites of musculoskeletal pain are back and shoulder, and musculoskeletal pain is often present at multiple sites
3,188 (1,956/ 1,232)
18þ
Number (men/ women) Age, y.o.
Musculoskeletal pain
99
North Island, New 289 Zealand (124/ 165)
Setting
Study population
. Table 99-1 (continued)
1706 Burden of Disease and Years Lived with Disability Associated with Musculoskeletal Pain
Burden of Disease and Years Lived with Disability Associated with Musculoskeletal Pain
99
. Figure 99-1 Comparison of the burden of disease in a population. The prevalence of Disease A is higher than that of Disease B, however, Disease A has a smaller impact on health-related QOL than Disease B. The burden of disease in a population should be measured by the prevalence of the disease in the population and the impact on health-related QOL of people with the disease
duration of disability (years). Incidence (I) Duration (L) is equivalent to Prevalence (P). The above YLD formula thus converts into: YLDs ¼ P DW where P is the number of cases of the disease at a given time. Although MP is recognized as a problem of global proportions, to date, few studies have measured the burden of MP in the general adult population according to the Global Burden of Disease protocol. MP causes disability rather than premature mortality. The burden of MP in a population can be predominantly expressed in terms of YLD. As part of the Bone and Joint Decade Monitor Project, we carried out a questionnaire survey on MP in a Japanese population. This population survey gave the first systemic grasp of the prevalence of MP in Japan (Suka, 2005a). Based on findings from this population survey, we estimated the YLDs due to MP in the Japanese adult population (Suka, 2005b).
4 > Figure
Estimation of the YLDs Due to MP in the Japanese Adult Population
99-2 shows an overview of the estimation of the YLDs due to MP in the Japanese adult population. The following information sources were available for the estimation of the YLDs due to MP in the Japanese adult population: the questionnaire survey on MP (Suka, 2005a), the Population Census (Japanese Ministry of Internal Affairs and Communications, Statistics Bureau), and the Global Burden of Disease Study (Murray, 1996). First, the prevalence of MP in the study population was examined. Second, the number of adults with MP was
1707
1708
99
Burden of Disease and Years Lived with Disability Associated with Musculoskeletal Pain
. Figure 99-2 Overview of the estimation of the years lived with disability due to musculoskeletal pain in the Japanese adult population
estimated based on the age- and sex-specific prevalence of MP. Finally, the YLDs due to MP were estimated based on the estimated number of adults with MP. In Japan, all adults are supposed to undergo an annual health examination at a healthcare facility or hospital. A questionnaire survey on MP was conducted in October to November 2003 at three healthcare facilities: Niigata Healthcare Association (Niigata, Niigata prefecture), Tsukuba Multiphasic Health Examination Center (Tsukuba, Ibaraki prefecture), and Seirei Health Examination Center (Hamamatsu, Shizuoka prefecture) (> Figure 99-3). At each healthcare facility, about one thousand participants who had come for health examinations were asked to fill in a questionnaire anonymously. Most of them agreed to complete the survey at the time of their examination. On a drawing with predefined body regions, the participants marked the regions affected by MP for more than 1 week in the past month (> Figure 99-4). A white circle in the parentheses indicated pain that did not interfere with daily activities and a black circle in the parentheses indicated pain that had interfered with daily activities. In addition to the experience of MP, the participants were asked whether they had received treatment for musculoskeletal disorders. Among the total 3,273 respondents, 3,188 eligible adults (1,956 men and 1,232 women) who had information on treatment for musculoskeletal disorders were included in the study population. The age- and sex-specific prevalence of MP was calculated in the context of interference with daily activities and treatment for musculoskeletal disorders.
Burden of Disease and Years Lived with Disability Associated with Musculoskeletal Pain
99
. Figure 99-3 Map of Japan
. Figure 99-4 Pain drawing with predefined body regions used in the questionnaire. On a drawing with predefined body regions, the participants marked the regions affected by musculoskeletal pain for more than 1 week in the past month
1709
1710
99
Burden of Disease and Years Lived with Disability Associated with Musculoskeletal Pain
To estimate the number of adults with MP in Japan, the age- and sex-specific prevalence of MP in the study population was multiplied by the corresponding age- and sex-specific population estimates as of October 2003 (Japanese Ministry of Internal Affairs and Communications, Statistics Bureau; http://www.stat.go.jp/english/data/jinsui/2.htm). To estimate the YLDs due to MP in the Japanese adult population, the estimated number of adults with MP in Japan was multiplied by the corresponding disability weights. The Global Burden of Disease Study shows a table of disability weights. This table covers a wide range of diseases and injuries but does not give values for MP or any other subjective symptoms. Therefore, the values for > osteoarthritis (0.108 for treated cases, 0.156 for untreated cases) were assigned to the cases of MP that interfered with daily activities and the values for > periodontal disease (0.001 for both treated and untreated cases) were assigned to the cases of MP that did not interfere with daily activities. There are arguments for and against the time-discounting and age-weighting health outcomes, so we avoided such processes in our estimation.
5
Findings from the Questionnaire Survey on MP
> Table 99-2 shows the prevalence of MP by anatomical area in the study population. There were considerably higher rates for neck and shoulder area and low back area compared to other areas for both sexes. Generally speaking, the prevalence of MP in each area increased with age up to the age of 60 and then tended to plateau. The exceptions were hip and knee area among men and hip and knee area and ankle and foot area among women, which showed substantial increased rates for ages 60–69 and 70 or older. As a possible explanation, the sensitivity of degenerative changes with aging may differ with the anatomical area. This explanation may be supported by the fact that the age-dependent increase in the prevalence of MP was steeper among women than among men. There was considerable overlap between MP experienced in different areas. Of those who had experienced MP, 27.6% (men 28.5%, women 26.2%), 7.5% (men 6.5%, women 9.0%), and 3.5% (men 2.4%, women 5.4%) had two, three, and four or more areas affected by MP, respectively. The number of areas affected by MP significantly increased with age for both sexes. Overall, 41.4% (men 40.9%, women 42.2%) of the study population reported MP in at least one area, and 20.5% (men 20.3%, women 20.8%) of those who had experienced MP reported interference with daily activities due to the pain. The overall prevalence of MP significantly increased with age among both sexes (> Figure 99-5). > Table 99-3 shows the treatment rates among people with MP in the study population. Despite potential deterioration in health-related QOL, the treatment rates among people with MP were quite low. Even in those who reported interference with daily activities due to the pain, the rate remained 28.1% (men 21.6%, women 38.0%).
6 > Table
YLDs Due to MP in the Japanese Adult Population
99-4 shows the estimated number of adults with MP in Japan. As of October 2003, 42.3 million (41.2%) of the Japanese adult population were estimated to suffer from MP. Among them, 9.1 million (8.8%) were estimated to experience interference with daily activities due to the pain. The estimated prevalence of MP in each area was higher among women than among men.
99
Burden of Disease and Years Lived with Disability Associated with Musculoskeletal Pain
. Table 99-2 Prevalence of musculoskeletal pain (%) by anatomical area in the study population Number
Neck/ shoulder
Elbow/wrist/ hand
Low back
Hip/ knee
Ankle/ foot
3,188
20.3
7.4
19.1
11.1
5.8
20–29
154
10.4
3.9
11.7
3.9
0.0
30–39
497
16.9
3.8
15.5
6.4
3.8
40–49
1,055
21.1
6.0
19.5
9.6
4.7
50–59
1,003
21.8
9.9
19.9
11.4
7.7
60–69
407
21.1
9.6
21.6
17.7
7.1
Age, y.o. All
All
70þ Men
72
16.7
8.3
19.4
34.7
11.1
1,956
19.6
6.6
20.1
9.1
5.5
20–29
93
9.7
6.5
11.8
2.2
0.0
30–39
322
14.0
2.5
17.1
6.2
4.7
40–49
647
21.5
5.9
21.2
7.9
4.8
50–59
599
21.0
8.0
20.9
9.8
7.5
60–69
252
21.8
10.3
23.4
13.9
6.0
43
20.9
7.0
16.3
25.6
4.7 6.3
All
70þ Women All
1,232
21.3
8.6
17.5
14.3
20–29
61
11.5
0.0
11.5
6.6
0.0
30–39
175
22.3
6.3
12.6
6.9
2.3
40–49
408
20.6
6.1
16.9
12.3
4.7
50–59
404
24.5
13.4
20.0
14.6
8.7
60–69
155
20.0
8.4
18.7
23.9
9.0
29
10.3
10.3
24.1
48.3
20.7
70þ
Location of musculoskeletal pain was classified into five anatomical areas: (1) neck and shoulder area (including neck, shoulder, and higher back), (2) elbow, wrist, and hand area (including upper arm, elbow, lower arm, wrist, hand, and finger), (3) low back area (including lower back), (4) hip and knee area (including hip, femur, and knee), and (5) ankle and foot area (including crus, ankle, and foot). There were considerably higher rates for neck and shoulder area and low back area compared to other areas for both sexes. Generally speaking, the prevalence of musculoskeletal pain in each area increased with age up to the age of 60 and then tended to plateau
The YLDs due to MP in the Japanese adult population were estimated at 1,297,843.7 (1263.6 per 100,000 population). After excluding the cases of MP that did not interfere with daily activities, the years remained 1,264,684.4 (1231.3 per 100,000 population). In the previous study, the YLDs due to osteoarthritis and rheumatoid arthritis in Japan were estimated at 65.8 and 42.8 per 100,000 population, respectively (Suka, 2004). For YLD, the burden of MP was 19 and 29 times as large as that of osteoarthritis and rheumatoid arthritis, respectively, in the Japanese adult population. The questionnaire survey on MP was conducted with those who participated in health examinations at three healthcare facilities. Generally speaking, participants in health examinations are rather healthy people, excluding bedridden patients. They are more likely to be aware of their own health and thus be sensitive to subjective symptoms. The number of adults
1711
1712
99
Burden of Disease and Years Lived with Disability Associated with Musculoskeletal Pain
. Figure 99-5 Overall prevalence of musculoskeletal pain in the study population (M men; F women). The cases of musculoskeletal pain that interfered with daily activities are shown as a light-colored bar. The overall prevalence of musculoskeletal pain significantly increased with age among both sexes
with MP and the YLDs due to MP may be somewhat under- or overestimated. However, it is an indisputable fact that MP imposes a substantial burden on the Japanese adult population.
7
Future Prospect
The estimation of the burden of disease is a fundamental part of a health needs assessment. Various measures have been used to quantify the burden of disease in a population. In order to use the measure for healthcare priority-setting in each country, the measurements should be meaningfully compared either within a country or between different countries. The comparison within a country will inform us which health problem should be given the highest priority in the country. The comparison between different countries will inform us whether the burden of disease in one country is larger than that in other countries, and if so, we can try to find out the cause of the discrepancy. The calculation of DALYs is a well-established methodology for the estimation of the burden of disease. The World Health Organization has provided internationally comparable estimates (DALYs) of the burden of disease for more than 130 diseases and injuries. Although MP is recognized as a problem of global proportions, the World Health Organization project has not counted MP among the target diseases and injuries. Reliable epidemiological estimates of the burden of MP are scarce. Presented in this chapter is the first study to measure the burden of MP in YLD. Our estimation method, based on the Global Burden of Disease protocol, may be applicable to other populations and may provide a feasible way to make relative comparisons of the burden of MP in different populations.
99
Burden of Disease and Years Lived with Disability Associated with Musculoskeletal Pain
. Table 99-3 Treatment rates among people with musculoskeletal pain in the study population Interference with daily activity Negative
All
Men
Women
Positive
Age, y.o.
Number
Treated (%)
Number
Treated (%)
All
28.1
1,050
11.5
270
20–29
29
6.9
9
11.1
30–39
141
5.0
25
12.0
40–49
349
7.7
84
25.0
50–59
349
11.7
102
31.4
60–69
152
24.3
40
35.0
70þ
30
23.3
10
50.0
All
638
8.2
162
21.6
20–29
18
11.1
6
16.7
30–39
86
4.7
19
10.5
40–49
222
5.4
51
19.6
50–59
204
9.8
55
23.6
60–69
89
12.4
26
30.8
70þ
19
15.8
5
20.0
All
412
16.7
108
38.0
20–29
11
0.0
3
0.0
30–39
55
5.5
6
16.7
40–49
127
11.8
33
33.3
50–59
145
14.5
47
40.4
60–69
63
41.3
14
42.9
70þ
11
36.4
5
80.0
Despite potential deterioration in health-related quality of life, the treatment rates among people with musculoskeletal pain were quite low
Toward better estimation of the burden of MP in the general adult population, further studies may be required to address the following issues. First, there is no standardized way to determine the prevalence of MP in the general adult population. MP is subjective and is usually described as an event or episode, and thus selfreports are unavoidable for case finding. The prevalence of MP may highly depend on the definition of MP and the wording of the question. However, a variety of questionnaires have been used in the population surveys of the prevalence of MP. The development of a standard case-finding instrument for MP is necessary for the comparison of absolute prevalence between different populations. Second, there is no standardized definition of MP. As indicated in the basic formula for calculating YLDs, the duration of disability is a major determinant of YLDs. The YLDs due to MP may be overestimated when those who completely recover from MP in a short time are counted among the cases of MP. Moreover, it is uncertain whether chronic MP has an equivalent impact on health-related QOL compared to acute MP in terms of disability weight.
1713
1714
99
Burden of Disease and Years Lived with Disability Associated with Musculoskeletal Pain
. Table 99-4 Estimated number (in millions) of adults with musculoskeletal pain in Japan Neck/ shoulder
Elbow/wrist/ hand
18.5 (18.0%)
7.3 (7.2%)
Men
8.9 (17.9%)
3.3 (6.6%)
9.1 (18.4%)
Women
9.7 (22.4%)
4.1 (9.4%)
9.3 (21.6%) 10.4 (24.1%) 4.3 (9.9%) 22.4 (51.9%)
All
Low back
Hip/knee
Ankle/ foot
Any area
18.4 (18.0%) 15.5 (15.1%) 6.6 (6.4%) 42.3 (41.2%) 5.1 (10.3%) 2.3 (4.6%) 19.9 (40.1%)
Values in parentheses indicate the estimated prevalence of musculoskeletal pain (%) in the Japanese adult population. Location of musculoskeletal pain was classified into five anatomical areas: (1) neck and shoulder area (including neck, shoulder, and higher back), (2) elbow, wrist, and hand area (including upper arm, elbow, lower arm, wrist, hand, and finger), (3) low back area (including lower back), (4) hip and knee area (including hip, femur, and knee), and (5) ankle and foot area (including crus, ankle, and foot). The overall prevalence of musculoskeletal pain in the Japanese adult population was estimated at 41.2%. The estimated prevalence of musculoskeletal pain in each area was higher among women than among men
Ideally, the burden of chronic MP should be discussed separately from that of acute MP. In order to formulate discriminating definitions of acute and chronic MP, the course of MP should be thoroughly investigated based on longitudinal data. Third, the Global Burden of Disease Study shows a protocol for calculating DALYs, but it does not cover MP. When we estimated the YLDs due to MP in the Japanese adult population, the disability weights for osteoarthritis and periodontal disease were used as substitutes because of absence of disability weights for MP. MP is often present at multiple sites. Moreover, it is possible that the impact of MP on health-related QOL differs with the pain site. The YLDs due to MP may be underestimated when the number of pain sites is not adjusted. On the other hand, the YLDs due to MP may be overestimated when the YLDs estimated at respective pain sites are simply summed up. The disability weights for MP should be determined in the context of clustering of MP and pain site according to the Global Burden of Disease protocol. To date, few studies have dealt with comorbidity in the calculation of DALYs. Salaffi (2005) reported that comorbidity was significantly associated with poor health-related QOL among people with low back pain. The effect of comorbidity on the health-related QOL was also found in other studies (Bingefors, 2004; Burstrom, 2001; Saarni, 2006; Salaffi, 2005). The YLDs due to MP may be overestimated when comorbidity is not adjusted. The Global Burden of Disease protocol should be updated in an attempt to adequately deal with comorbidity in the calculation of DALYs. Because of high prevalence and great impact on health-related QOL, MP imposes a substantial burden on the general adult population worldwide. MP has received less attention in terms of healthcare allocations than other serious conditions and has not been counted among national healthcare priorities in most countries. However, the control of MP should not be bypassed in efforts to improve the health-related QOL of the general adult population. Picavet (2003) reported that only 30–40% of the Dutch population aged 25 years or older who experienced MP consulted their general practitioner about their pain and 17–27% used medicines for MP. According to our estimates, 5.8 million (64%) of the Japanese adult population who experience interference with daily activities due to MP leave their pain untreated. The sensitivity analysis suggests that the YLDs due to MP in the Japanese adult population may decrease by 20% if all the people who leave their pain untreated receive medical attention. Public education campaigns may be required to help people develop a sense
Burden of Disease and Years Lived with Disability Associated with Musculoskeletal Pain
99
of responsibility for their own health, which may lead to a decrease in the number of people with untreated MP. Health professionals should advance research on the effective treatment of MP. The comparison of the YLDs due to MP may be helpful in assessing the effectiveness of public health measures.
Summary Points Musculoskeletal pain (MP) is rarely associated with fatal conditions, however, it causes deterioration in health-related quality of life.
The prevalence of MP in the general adult population is high with a significant agedependent increase.
Although MP is recognized as a problem of global proportions, to date, few studies have
measured the burden of MP in the general adult population according to the Global Burden of Disease protocol. The burden of MP in a population can be predominantly expressed in terms of year lived with disability (YLD), which is a disability component of disability-adjusted life year (DALY). According to our estimates of YLDs, the burden of MP is 19 and 29 times as large as that of osteoarthritis and rheumatoid arthritis, respectively, in the Japanese adult population. Our estimation method, based on the Global Burden of Disease protocol, may be applicable to other populations and may provide a feasible way to make relative comparisons of the burden of MP in different populations. Further studies may be required to better estimate the burden of MP in the general adult population.
Acknowledgments A series of studies based on the questionnaire survey on MP were supported by the 2003–2004 Health and Labour Sciences Research Grant (Research on Prevention and Treatment for Immune Allergic Diseases) from the Japanese Ministry of Health, Labour, and Welfare. We would like to thank all the staff of Niigata Healthcare Association, Tsukuba Multiphasic Health Examination Center, and Seirei Health Examination Center for their help on the questionnaire survey.
References Allison TR, Symmons DP, Brammah T, Haynes P, Rogers A, Roxby M, Urwin M. (2002). Ann Rheum Dis. 61: 151–156. Andersson HI, Ejlertsson G, Leden I, Rosenberg C. (1993). Clin J Pain. 9: 174–182. Bergman S, Herrstrom P, Hogstrom K, Petersson IF, Svensson B, Jacobsson LT. (2001). J Rheumatol. 28: 1369–1377.
Bingefors K, Isacson D. (2004). Eur J Pain. 8: 435–450. Brage S, Bjerkedal T. (1996). J Epidemiol Commun Health. 50: 166–169. Brooks P. (2005). Rheumatology. 44: 831–833. Burstrom K, Johannesson M, Diderichsen F. (2001). Qual Life Res. 10: 621–635. Chrubasik S, Junck H, Zappe HA, Stutzke O. (1998). Eur J Anaesthesiol. 15: 397–408.
1715
1716
99
Burden of Disease and Years Lived with Disability Associated with Musculoskeletal Pain
Lidgren L. (2003). Bull World Health Organ. 81: 629. Murray CJ, Lopez AD. (ed.) (1996). The Global Burden of Disease: A Comprehensive Assessment of Mortality and Disability from Diseases, Injuries, and Risk Factors in 1990 and Projected to 2020. Harvard University Press, Harvard. Picavet HS, Schouten JS. (2003). Pain. 102: 167–178. Saarni SI, Harkanen T, Sintonen H, Suvisaari J, Koskinen S, Aromaa A, Lonnqvist J. (2006). Qual Life Res. 15: 1403–1414. Salaffi F, De Angelis R, Stancati A, Grassi W. (2005). Clin Exp Rheumatol. 23: 829–839.
Suka M, Yoshida K. (2004). Mod Rheumatol. 14: 285–290. Suka M, Yoshida K. (2005a). Mod Rheumatol. 15: 41–47. Suka M, Yoshida K. (2005b). Mod Rheumatol. 15: 48–51. Taylor W (2005). N Z Med J. 118: U1629. Urwin M, Symmons D, Allison T, Brammah T, Busby H, Roxby M, Simmons A, Williams G. (1998). Ann Rheum Dis. 57: 649–655. Woolf AD, Akesson K. (2001). Br Med J. 322: 1079–1080.
100 Financial Burdens and Disability-Adjusted Life Years in Loss of Vision Due to Trachoma K. D. Frick 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1718
2
Disease Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1719
3
An Elimination Strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1721
4
The Alliance for Global the Elimination of Trachoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1723
5
A Disease of Poverty Leading to More Poverty . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1724
6
Financial Burden of Trachoma Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1725
7
Disability Adjusted Life Years . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1727
8
Other Efforts to Quantify the Burden of Vision Loss Related to Trachoma . . . . . . 1727
9
Cost-Effectiveness of Trachoma Control Efforts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1728
10 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1729 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1730
#
Springer Science+Business Media LLC 2010 (USA)
1718
100
Financial Burdens and Disability-Adjusted Life Years in Loss of Vision Due to Trachoma
Abstract: Trachoma is a disease process that progresses through multiple stages before reaching a disabling stage (> trichiasis) and causing its maximal disability (blindness). At present, the disease affects primarily poor populations in poor countries. The burden can be described either by the financial burden of lost potential productivity among individuals who are disabled as a result of the condition, by the > disability adjusted life years associated with the condition, or the resources necessary to limit the other burdens. Comparisons of the resources necessary to limit the other measures of burden with the amount of burden eliminated are > cost-benefit analyses or > cost-effectiveness analyses. A coordinated strategy to control the burden exists; the strategy involves surgery, antibiotics, face washing, and environmental change and education. The results of the best cost-effectiveness analyses that have been done comparing the individual components of the strategy suggest that, from a > societal perspective, the most cost-effective way of limiting the disability adjusted life years or lost productivity burden associated with trachoma is to treat the later stages through surgery rather than the earlier stages through antibiotics. As areas that are affected at present continue to develop economically, the general improvements in hygiene that accompany these changes are likely to result in decreased infection and transmission. List of Abbreviations: DALY, disability adjusted life year; GDP, > gross domestic product; GET 2020, alliance for the global elimination of trachoma by 2020; HALY, > handicap adjusted life year; ITI, > International Trachoma Initiative; NGOs, > non-governmental organizations; SAFE, surgery, antibiotics, face-washing, education/environment; WHO, World Health Organization
1
Introduction
The burden of trachoma may be characterized using information as simple as the prevalence of active trachoma infection or any of its sequelae. The results of a report by Resnikoff et al. can be adapted to show the proportion of blindness in each > World Health Organization region caused by trachoma (> Figure 100‐1) – a range from nearly zero to nearly 7% (Resnikoff et al., 2004). There have been many reports in the last 5 years on the prevalence of trachoma or related impairments in different regions of the world (Durkin et al., 2007; Goldschmidt et al., 2007; Lansingh and Carter, 2007; Mathenge et al., 2007a, b) (> Table 100‐1 lists the areas.). However, there are more nuanced ways of measuring the burden of trachoma – in particular focusing on the disability adjusted life years (DALYs) associated with the condition or the financial burden of the condition. The financial burden of the condition can be characterized as the monetary value of the productivity that is lost because of the condition or the money that is required to treat the condition and avoid the productivity losses and other aspects of activity limitations. All of these are eventually summarized in the tradeoffs that are encapsulated in a cost-benefit or cost-effectiveness analysis. This chapter will summarize the disease process and international efforts to eliminate the burden of trachoma that results from blindness, various measures of the financial burden of the disease, the financial burden of treating the disease, DALYs associated with the condition, other attempts to summarize the value of the disease, and the cost-effectiveness of different ways of trying to limit the financial and disability-related burden associated with trachoma.
Financial Burdens and Disability-Adjusted Life Years in Loss of Vision Due to Trachoma
100
. Figure 100‐1 Proportion of blindness in World Health Organization regions caused by trachoma. Proportion of individuals who are blind in each World Health Organization region who are blind because of trachoma. Based on data from Resnikoff et al. (2004)
. Table 100‐1 Examples of countries in which the prevalence of trachoma has been documented in the past 5 years Country
World Health Organization region
Kenya
African region-E
Mexico
Region of the Americas-B
Myanmar
South-east Asia region-D
Rwanda
African region-E
Countries of the world for which the prevalence of trachoma has been studied in the past 5 years, along with the World Health Organization region in which each country is located. The African region is more strongly represented than other regions
2
Disease Process
Vision loss related to ocular infection with Chlamydia trachomatis is an avoidable end result of a progressive process (sometimes labeled as chronic) that occurs in only a fraction of the world’s countries at this point in time (West, 2004). Many of these countries are in the
1719
1720
100
Financial Burdens and Disability-Adjusted Life Years in Loss of Vision Due to Trachoma
sub-Saharan region of Africa. The only country with an established market economy in which trachoma remains endemic is Australia (Tellis et al., 2007), specifically among the Aboriginal population. To summarize, this is a disease that primarily affects poor areas of the world (Wright et al., 2007). . Figure 100‐2 Disease progression. The stages of disease are shown with indications of different points at which the disease progression may stop, which stages of the disease can be subject to intervention, which stages are not subject to intervention but not disabling, and which stages are disabling
The disease process begins with infection, primarily among children (> Figure 100‐2 shows the progression of the disease and makes clear the number of steps between active infection and blindness.). The infected child is likely to have inflammation of the upper eyelid and may have ocular discharge. The two levels of infection are characterized by the intensity of the inflammation. While identification of this inflammation is used in rapid assessment methodologies, there have been multiple papers demonstrating that the inflammation does not always occur when there is an infection and that the inflammation sometimes remains after infection. These two facts make interventions that treat infected children and the members of their households less effective than might be the case if the presence of inflammation were a more sensitive and specific indicator of infection status. It is financially prohibitive to consider the possibility of using a clinical test for identification of the condition to treat only those with infection. The infection itself, and the associated inflammation and ocular discharge, have not been shown to have an associated disabling or financial burden of productivity loss. However, preventing infection is likely to be a key aspect of any potential strategy to limit the burden of trachoma because avoiding infection is the first step in avoiding the remainder of the process that leads to burden through trichiasis, visual acuity impairment, and eventually blindness.
Financial Burdens and Disability-Adjusted Life Years in Loss of Vision Due to Trachoma
100
Over time, individuals who have multiple infections, particularly multiple severe infections, develop trachomatous scarring. The scarring of the upper eyelid has not been shown to be associated with a disabling burden and there is nothing that can be done to treat the scarring. The scarring is simply an intermediate step in the process that can be avoided by avoiding earlier stages and that represents a risk factor for further progress in the disease process. The disease process continues as individuals with scarring develop trichiasis. Trichiasis involves at least one inturned eyelash that rubs on the globe of the eye. The severity of trichiasis is characterized by the number of lashes that are rubbing on the globe of the eye. When an eyelash rubs on the cornea, the cornea becomes scarred. An eyelash rubbing against the globe of the eye is painful and is associated with photophobia. Inturned eyelashes can be treated surgically, and the recommended surgery is bilamellar tarsal rotation (West and Taylor, 1999). This procedure can be performed by a well trained ocular nurse or other personnel who are not medical doctors but who are well trained for the low cost and relatively simple and surgical procedure. When surgery is not available, some individuals with trichiasis resort to epilation, or manual plucking of the eyelash, although this has not been shown to be as effective (West et al., 2006). Despite the relative simplicity and low cost of the trichiasis surgery, there is a lack of surgery available in trachoma endemic countries and there have been reports of high levels of recurrence after surgery. The eyelash rubbing on the cornea has been demonstrated to be associated with a burden even before the individual has impaired visual acuity. Without treatment, the person will eventually have impaired visual acuity that cannot be corrected as the cornea is permanently scarred. With sufficient scarring, the person’s visual acuity will become sufficiently impaired to be considered blind. > Table 100‐2 provides a few facts about trachoma. . Table 100‐2 Facts about trachoma Facts Caused by Chlamydia trachomatis Stages include inflammation, scarring, trichiasis (inturned eyelashes), and corneal opacity Spread to Europe in the 1800s (West, 2004) No longer a problem in developed countries (West, 2004) Affects 300–500 million individuals (Thylefors et al., 1995) Blindness prevalence related to trachoma is approximately 5.9 million cases (Thylefors et al., 1995) Risk factors for trachoma infection include age, gender, lack of access to clean water, personal hygiene, flies, a lack of latrines, the presence of cattle, crowding in households, and nutritional deficiency (West, 2004) Strategy for control includes surgery, antibiotics, face washing, and environmental change and education
3
An Elimination Strategy
Different points of potential intervention were described above. One comprehensive strategy for the elimination of blindness related to trachoma is called the SAFE strategy (Frick and West, 2001). The acronym stands for surgery, antibiotics, face washing, and environmental change and education.
1721
1722
100
Financial Burdens and Disability-Adjusted Life Years in Loss of Vision Due to Trachoma
Surgery has already been described. Not only has it been shown to be relatively low cost, but it has also been shown to be the most cost-effective single intervention by Baltussen et al. (2005) (as will be discussed later). This is largely because there is no burden (as measured by disability or a monetary value) that has been associated directly with the infectious or scarring stage of the trachoma disease process. If a sufficient quantity of surgical supplies and a sufficient number of trained surgeons were available, this would be an important finding for public policy planning. From a public health perspective, this finding is somewhat counterintuitive, as it is usually presumed that preventing infection would be the best approach to reducing the burden and there is no desire to see anyone unnecessarily having an infection and suffering through the difficulties of ocular discharge that may occur. However, given that a relatively small fraction of those who ever are infected eventually reach the trichiasis stage of progression toward trachomatous blindness, policy makers reasonably may consider only the surgery as long as the surgery has a high rate of initial success, a low rate of long-term failure as indicated by recurrence of trichiasis, and is available soon after the onset of trichiasis. The original antibiotic of choice for trachoma was tetracycline (Burton et al., 2002). The difficulty with this antibiotic was that this was an ointment that had to be applied to a patient’s eyes for 6 weeks. Applying any ointment to an eye is difficult, and patients had difficulty complying with 6 weeks of applying tetracycline ointment. At present, azithromycin has been shown to be relatively effective (Burton et al., 2002). Azithromycin can be effective when used to treat infected children and their families (Holm et al., 2001), and when used in mass treatment approaches (West et al., 2005). While there is often concern about the use of antibiotics in mass treatment settings, it has been demonstrated that there is little development of resistance to azithromycin (Solomon et al., 2005). One problem with mass treatment approaches is that the effectiveness is limited if migration continues to occur; if anyone who is infected migrates into an area that has previously received mass treatment, the cycle of infection can begin again. Another difficulty with azithromycin as a choice for antibiotic (mass treatment or otherwise) is its expense. The expense contributes to the finding of a lack of cost-effectiveness when azithromycin is considered as the sole strategy for trachoma control. A non-governmental organization called the International Trachoma Initiative (ITI) was formed in 1998 by grants for the Edna McConnell Clark Foundation and Pfizer (Kumaresan and Mecaskey, 2003). Pfizer has committed to donate azithromycin for the countries that are being directly served by ITI and already has been responsible for the donation of millions of doses (Mabey and Solomom, 2008). However, given that the list price of azithromycin in the United States is higher than the total cost of surgery in most regions of the world (Baltussen et al., 2005), it is difficult to envision national governments being able or willing to use their own resources to adopt azithromycin as a primary part of a trachoma control strategy – particularly for a mass treatment. Face washing has the potential to be a useful component of a strategy to control trachoma and the resulting burden. Face washing is important because the ocular discharge associated with infection facilitates disease transmission. Ocular discharge contributes to transmission because children touch their eyes with their hands and then touch other children with their hands. Ocular discharge also can contribute to transmission in endemic areas with large numbers of eye-seeking flies, Musca sorbens, that have been studied as a vector of transmission (Emerson et al., 2000). Flies contribute to disease transmission by landing first on the faces of individuals with the infection and later on the faces of those without. Persons whose faces have been landed on are at higher risk for disease because when they rub their eyes they may
Financial Burdens and Disability-Adjusted Life Years in Loss of Vision Due to Trachoma
100
introduce Chlamydia trachomatis into the ocular environment. Despite its potential efficacy, face washing is difficult to implement in many of the trachoma endemic areas because of a lack of clean water. Even with clean water, achieving consistent face washing can be difficult as health habits are generally difficult to change. Studies of interventions to increase face washing have not shown a particularly high level of effectiveness. The types of environmental changes begin with moving livestock away from the general living area. The presence of livestock is associated with a general lack of cleanliness and a higher prevalence of flies. Moving livestock away from living quarters can be difficult because many at-risk families own very small houses and very little land. A second component of environmental changes is using covered pit latrines instead of allowing individuals to defecate in uncontrolled areas. This is also aimed at controlling fly populations. Finally, environmental changes include making clean water more accessible. Clean water can be used to wash faces and improve the general cleanliness of the environment. The educational component focuses on the need for appropriate face washing using water efficiently, the need to implement the environmental changes, and the usefulness of surgery over epilation for treating trichiasis. No economic evaluation of the complete SAFE strategy has been performed. Economic analyses have assessed the way in which the burden would change if specific components of the strategy were adopted – usually antibiotics and surgery. Different surgical settings and different types of antibiotic administration have been compared. Combinations of surgery and antibiotic have also been analyzed as a set. Little has been done to assess the cost of achieving improved face washing. Assessments of environmental change are difficult because of the difficulty of assessing the cost of environmental change and the difficulty of characterizing all effects of environmental change that extend well beyond trachoma control. Another complicating factor is that at least one study (West et al., 2006) has reported that when the strategies are implemented together, the effects may be less than additive. Specifically, insecticide treatment after mass treatment with azithromycin helped to control the fly population but did little to incrementally limit trachoma prevalence. The finding regarding insecticide and trachoma control after mass treatment with azithromycin is in contrast to the report by Emerson et al. (2004) indicating that insecticide treatment alone was effective at both reducing the number of flies on children’s faces and at reducing the prevalence of infections. > Table 100-3 summarizes the components of the SAFE strategy that have been subjected to willingness to pay (assessing the value placed on the component by the local population), cost, cost-effectiveness (with cases or disability adjusted life years as the outcome), and costbenefit (with dollars as the outcome) evaluations.
4
The Alliance for Global the Elimination of Trachoma
Despite the existence of a well-defined public health approach to trachoma control, vision loss related to trachoma remains a prominent cause of blindness. The avoidable nature of this particular cause of vision loss has led to its having its own initiative called the Alliance for the Global Elimination of Trachoma by 2020 (GET 2020) (Mabey and Solomon, 2008). GET 2020 includes a diverse array of organizations and individuals. The World Health Organization (WHO) is a key organization in the alliance and most frequently the site of the meetings. Non-governmental organizations (NGOs) that promote interventions related to eye
1723
1724
100
Financial Burdens and Disability-Adjusted Life Years in Loss of Vision Due to Trachoma
. Table 100‐3 Components of the SAFEa strategy subjected to various cost-outcome analyses Component
Willingness to pay
Surgery Antibiotic
X
Cost
Costeffectiveness
Costbenefit
X
X
X
X
X
Face washing Environmental change and education a
SAFE – Strategy to control trachoma including surgery, antibiotics, face washing, and environmental/educational changes This table shows that most economic evaluations and descriptions of potential changes in the burden have focused on surgery and antibiotics rather than face washing or environmental change and education
health are also key members of the alliance. One of the NGOs is the International Trachoma Initiative (ITI). Pfizer is also a member and has donated millions of doses of azithromycin in a program to help in efforts to break the cycle of trachoma transmission. Country representatives from endemic areas in Africa, Southeast Asia, and Latin America are also key members of the organization as no new initiatives to control trachoma can be implemented without agreement from countries in which trachoma remains a problem and without considering the needs and concerns of the target population. Finally, academic researchers are also members of the alliance; these researchers include physicians, epidemiologists, and researchers focusing on the flies that are thought to be a transmission mechanism and on the economics of the condition. Over time this group of stakeholders and researchers has helped to enhance the control of trachomatous blindness. GET 2020 members have discussed issues including rapid assessment, the burden of disease, the cost-effectiveness of efforts to control the disease, and a variety of efforts to use public health interventions to control the disease. During the existence of the alliance, several countries have moved close to the goal of eliminating trachoma as a cause of blindness, or at least eliminating it as a cause of sufficient blindness to be listed as anything more than a small component of the other causes of blindness which is the most realistic goal for the elimination of trachoma (Mabey and Solomon, 2008). The remainder of the paper will describe what is known about the financial and disability adjusted life years consequences related to vision loss due to trachoma.
5
A Disease of Poverty Leading to More Poverty
The nations where trachoma remains a problem are among the poorest in the world when measured by the gross national income per capita. Even within the poor populations within the countries in which trachomatous vision loss remains endemic, the poor are at greater risk of the disease (Wright et al., 2007). The poor are more vulnerable than their well off counterparts because they are more likely to live in close quarters with livestock, less likely to use pit
Financial Burdens and Disability-Adjusted Life Years in Loss of Vision Due to Trachoma
100
latrines, and less likely to have access to clean water. All of these factors are associated with generally poor hygiene. The link between poor hygiene and the transmission of trachoma was described above. The exact financial burden (in terms of vision loss from trachoma making a contribution to keeping people in poverty or leading people into poverty) has not been well documented. However, there are reasons to suspect that those who are suffering from stages of vision loss prior to blindness will have difficulty with work and be unable to lift themselves out of poverty or fall into deeper poverty. Frick et al. (2001c) documented that even those who have trichiasis but who do not have a severe visual acuity impairment are at greater risk of having problems with daily functions that are necessary for life in villages in Tanzania. There is no reason to suspect that individuals in other regions of Africa or in other areas around the world in which trachoma remains endemic would not have similar difficulty with functions including work for pay and other subsistence tasks. The financial burden in this case would not only affect the individual but would also affect the entire family unit as other individuals within the family may need to perform tasks that had been performed by the person who has trichiasis or someone else may need to be hired to perform the tasks. Further, the person experiencing trichiasis may also need direct care that can be provided by a family member and that will take a family member away from other activities that could be used to sustain the family. There has been little direct study anywhere around the world regarding the exact amount of care that is provided to individuals who are blind or visually impaired by other individuals in their community. A figure that is commonly used in economic assessments it that one-tenth of the time of an otherwise productive individual is taken up by caring for an individual who is blind. However, this figure comes from a single study in India in which it was basically assumed.
6
Financial Burden of Trachoma Control
There is an important financial burden associated with trichiasis surgery that is borne by the patients and their families that does not always enter into the calculations of the burden or the cost-effectiveness. One study referred to it as an ‘‘indirect cost’’ of obtaining surgery, although other studies might refer to the costs as direct but unrelated. The term ‘‘indirect cost’’ is sometimes used in accounting to refer to overhead. The term ‘‘indirect costs’’ in health economics is generally used to refer to productivity costs. Melese et al. (2004) used it to refer to costs that were related to obtaining care other than the cost of the care itself (i.e., transportation, food costs, etc.). The authors of this study did not directly measure the burden of these costs related to obtaining care, but reported that for individuals with trichiasis in Ethiopia, this was the most commonly cited reason for not obtaining surgery. This factor cannot be ignored in planning for efforts to relieve the burden of trachoma. Individuals and families with any health condition must assess whether the burden associated with the disease is greater than the burden of relieving the disease. Governments and non-governmental organizations who hope to reduce the burden of trachoma must focus on reducing the burden of obtaining the treatment. Frick et al. (2001a) reported that the cost of surgery including providing transportation to the village in The Gambia was $6.13. While this is a low cost for surgical procedures, this is still a substantial amount to pay for a single medical procedure in a trachoma endemic country. However, it is critical to note that even at no charge for the surgery, individuals would have
1725
1726
100
Financial Burdens and Disability-Adjusted Life Years in Loss of Vision Due to Trachoma
had to spend $0.22 to travel locally to surgery and that only two-thirds of eligible patients obtained surgery even at this essentially zero cost (this follows from the work of Bowman et al. (2000) that preceded the work of Frick, Keuffel, and Bowman). This finding of a lack of uptake even at very low out-of-pocket costs is consistent with the finding reported in Ethiopia that the costs of getting care other than the cost of surgery itself is an important barrier to the receipt of care. Frick et al. (2001b) reported on the cost of implementing a mass treatment and the cost of screening in Nepal. They found that a public health program provider cost per child would have been 3.05 Nepali rupees (US$0.04) without the cost of azithromycin from an Indian provider, 29.39 rupees (US$0.39) with the cost of azithromycin, and that the > societal cost would have been 31.17 rupees (US$0.42). The costs for targeted treatment (which resulted in treating older children the greater amounts of azithromycin) were 4.13 (US$0.06), 61.16 (US$0.82), and 63.20 (US$0.85). The obvious difference for the public health program without the cost of azithromycin included is the need to perform the screening to target the treatment. Schemann et al. (2007) reported on the cost of different strategies for antibiotic administration in Mali. The three different strategies considered included mass community-based treatments (all residents), treatment of all children under age 11 and all women aged 15–50 (an approach targeting those most at risk of having active trachoma), and treatment targeted to households with at least one child with active disease. The cost of the first strategy is much higher than the other two and does not depend on prevalence. For a district of 500,000 individuals, the cost of the mass treatment of the entire community was estimated at over $90,000 even without the cost of azithromycin and nearly $250,000 with the cost of azithromycin included. At a prevalence of 35%, the societal cost of treatment of all children under age 11 and all women of childbearing age is lower than the cost of the targeted program. At 20% prevalence or lower, the costs of the targeted program are the lowest. Baltussen et al. (2005) analyzed the cost-effectiveness of different strategies to reduce the burden of trachoma from a societal perspective and those results are reported below. Baltussen et al. did not report specifically on the per person costs of administering any of the interventions, but instead summarized the 10 year costs within each of the global regions. However, it is useful to examine the prices that were reported for certain components of the treatment strategy or the costs of surgery in the African region with a low level of health that are most directly comparable with the results reported by Frick, Keuffel, and Bowman from The Gambia. Baltussen et al. reported on the average cost of a tube of tetracycline and a dose of azithromycin in > international dollars (i.e., dollars adjusted to have purchasing power parity with the United States). The base cost of a tube of tetratcycline was only I$0.12. For a 250 mg capsule of azithromycin, the base cost was I$7.50. For a bottle of suspension with 1,200 mg of azithromycin, the base cost was I$29.51. These are not the total cost of providing these. Interestingly, Baltussen et al. found that even in the least healthy and least developed regions of Africa, the cost for trichiasis surgery was calculated at I$17.71, a much higher figure than that reported earlier by Frick et al. (2001a). Both Melese et al. (2004) and Frick et al. (2001a) also note that the costs of the surgery itself is an important factor in the lack of uptake of surgery. This prolongs the burden of trichiasis and presents a different type of burden in order to relieve the burden of trichiasis. Rabiu and Abiose (2001) did not measure the cost of surgery itself but found that respondents in Nigeria also noted that this was a major deterrent to obtaining surgery. Other authors have similarly found cost to be a barrier (Oliva et al., 1997–1998). O’Loughlin et al. (2006) reported the average amount spent on building latrines in Amhara, Ethiopia. The mean cost among all individuals with latrines was $0.80 although
Financial Burdens and Disability-Adjusted Life Years in Loss of Vision Due to Trachoma
100
there was a large range of costs from those who had spent nothing on building the latrine (other than their own time) to an average of US$4.0 among those who had paid anything. Rodgers et al. (2007) reported that it would be necessary for all individuals within a trachoma endemic community to have covered latrines for latrines to be an effective intervention for the purposes of limiting trachoma. In trachoma endemic areas that tend to rely on subsistence agriculture and are in countries with relatively low gross national income per capita, an expenditure of US$0.80 for a family that wants to have a covered latrine is not negligible. However, it is a sufficiently low cost that the local population, the national government, or a non-governmental organization may be able to afford to fund an intervention to build latrines for an entire community. No effort has been made to characterize the cost of insecticide, the costs of providing clean water, or the costs of face washing interventions.
7
Disability Adjusted Life Years
Frick et al. (2003a) projected the total number of disability adjusted life years (DALYs) experienced each year and the lifetime DALYs experienced due to years of life with disability (assigning nothing for years of life lost due to trachoma). The work of these authors suggested a total of 3.6 million DALYs experienced each year and 38.9 million DALYs experienced over the lifetimes of the population with prevalence low vision or blindness due to trachoma. Baltussen et al. (2005) did not calculate total DALYs associated with trachoma, but only calculated the DALYs that would be reduced by the effects of interventions.
8
Other Efforts to Quantify the Burden of Vision Loss Related to Trachoma
Individuals are able to express their perspective on the burden of trachoma by indicating their willingness to pay for goods and services that can help them to avoid the conditions caused by trachoma (i.e., blindness). Frick et al. (2003c) studied Tanzanian adults’ willingness to pay for a second azithromycin treatment for their children. Interestingly, these authors found that 38% of adults would have been willing to have their children receive a second dose but were willing to pay nothing. This demonstrates one of the difficulties of using willingness to pay to make inferences about the value of prevention of complications from a disease in a developing country setting. While the willingness to pay was logically related to the perception of a benefit from the antibiotic, it was also related to the availability of cash resources and the populations affected by a condition like trachoma do not tend to have a large amount of cash resources available. Of course, in this case, it may also be a function of the fact that the relationship between a single dose of azithromycin and the reduction in the future risk of visual impairment due to trachoma is hardly clear to researchers and may be totally unclear to the individuals who were being surveyed. Thus, while preference elicitation methods may be useful for assessing the value of avoiding the burden (and conversely the value of the burden) as perceived by the affected population in some cases, it does not seem to be a useful measure in the context of trachoma control. Trichiasis has been shown to have a burden prior to the development of visual impairment. Frick et al. (2001c) compared responses to a questionnaire that focused on daily activities of
1727
1728
100
Financial Burdens and Disability-Adjusted Life Years in Loss of Vision Due to Trachoma
life in an African among individuals who had trichiasis but no visual acuity impairment, visual acuity impairment but no trichiasis, both conditions, and neither condition. The authors found that both men and women were able to eat, bathe, and dress regardless of their visual acuity or their trichiasis status. However, particularly for women, having trichiasis without visual impairment was associated with a number of limitations in other tasks and greater difficulty with those tasks. The list of tasks included visiting neighbors, walking outside the village, attending market, recognizing faces, helping with the farm, weeding, housework, gathering firewood, cooking, and caring for children. All these tasks are critical to life in a village in most trachoma endemic countries. To date, while there are disability weights for low vision and blindness and the quality of life associated with a variety of ocular conditions has been assessed in more developed countries, no weight has been developed to characterize the loss of ability associated with trichiasis but not visual acuity impairment. Frick et al. (2003b) applied a dollar value of the burden associated with blindness and low vision associated with trachoma around the world. The figure was US$5.3 billion. The largest proportion of the total figure was accounted for by US$2.2 billion assuming a 60% productivity loss for blindness. Allowing for the possibility that blindness results in a person being completely unproductive (particularly in a developing country setting in which there is little accommodation for blindness), this would add another US$1.5 billion to productivity loss. Assigning a 24.5% loss to productivity for individuals whose visual acuity was at the level of low vision rather than blindness resulted in an additional US$1.3 billion dollars of loss. Finally, the remaining US$0.4 billion was attributable to the loss of productivity due to individuals providing informal care to those who were blind. There is one older piece of literature that characterized the burden associated with trachoma in terms of handicap adjusted life years. This was a concept that grew out of the distinction between disability and handicap in the International Classification of Impairment, Disease, and Handicap that was in use at the time that the concept of DALYs was developed. This system separated the concepts of disability and handicap by differentiating whether the social context of the disability was being considered. Disability was intended to be a ‘‘social context free’’ measure while handicap was the disability within a particular social context. Only one paper ever reported on the burden of trachoma using the concept of HALYs (Evans and Ranson, 1995–1996). This paper provided a point estimate of 80 million HALYs but gave a range from 15 million HALYs to 500 million HALYs. The authors of this paper pointed out that the uncertainty of the estimate of the HALYs associated with trachoma, due largely to the wide range of estimates of the prevalence of trachoma (Ranson and Evans, 1995–1996), could make any comparison of the burden associated with trachoma and the burden associated with other conditions misleading. Despite the fact that Frick et al. (2003b) did not report a confidence interval for their estimate of the dollar value of the burden related to trachoma, the work of Evans and Ranson (1995–1996) should be taken as an indication of how difficult it is to obtain a precise estimate of the burden associated with trachoma.
9
Cost-Effectiveness of Trachoma Control Efforts
This section will address results reported by Evans et al. (1996) focusing on Burma; Frick et al. (2001b) focusing on Nepal; Frick et al. (2001a) focusing on The Gambia, and Baltussen et al. (2005) reporting results around the world.
Financial Burdens and Disability-Adjusted Life Years in Loss of Vision Due to Trachoma
100
Evans et al. (1996) reported on 30 years of experience with a trachoma control program in Burma. They found that the program cost $54 per case of visual impairment prevented. Their results suggest that surgery cost more per case averted ($193) than non-surgical interventions ($47) although this was in a time prior to the availability of azithromycin. They converted results to handicap adjusted life years and calculated a cost of $4 per HALY avoided for the entire program with the range from $10 per HALY averted for surgery and $3 per HALY averted for non-surgical procedures. These results suggest that it is possible to avoid the burden associated with trachoma at a fairly reasonable price but the results are difficult to compare with other studies as the outcome units used are unique to the work of Evans and colleagues. Frick et al. (2001b) reported an incremental cost-effectiveness analysis of mass treatment and targeted household treatment in Nepal. They found that the targeted household treatment was no more effective and definitely more expensive than the mass treatment. As a result, if these were the only two choices, the mass treatment would be recommended for relieving the burden associated with trachoma in Nepal. Frick et al. (2001a) reported that on the positive net benefit of the village-based surgery program in The Gambia using a > human capital approach to assess the benefit of the program. The human capital approach places a value on health that is related to the potential productivity of the individual when he or she is healthy. This result argues that a surgical program would be an economically justifiable way to reduce the burden associated with trachoma in The Gambia. Baltussen et al. (2005) reported on results in seven regions of the world defined by geography and by the local levels of health: Africa-D and Africa-E (the countries in ‘‘E’’ having worse health status), America-B (the United States is in America-A), Eastern Mediterranean-B and Eastern-Mediterranean-D, Southeast Asia-D, and Western Pacific-B. In all areas, surgery has the most favorable cost-effectiveness ratio. In all regions, mass treatment with tetracycline is less cost-effective than mass treatment with azithromycin. In all regions, mass treatment with azithromycin is more cost-effective than targeted treatment with azithromycin. In no region, is mass treatment with azithromycin less costly per DALY avoided than three times GDP per capita, which has been set as an indicator of a relatively cost-effective service. Baltussen et al.’s results suggest that surgery should be adopted to reduce the burden associated with trachoma. The results also suggest that programs combining surgery and antibiotics may be viewed as relatively cost-effective but that the cost may still be prohibitive.
10
Conclusion
Trachoma is a long-term disease process for which a sizeable fraction of the world’s population in less developed countries is at risk and that has been the target of global efforts to eliminate the blindness that can result from this condition. Despite this, the burden is small when measured by prevalence or by disability adjusted life years or dollars. The burden is small because despite the prevalence of active disease, the burden of the active disease has never been quantified in terms of dollars or disability adjusted life years. The monetary or DALY result has only been characterized for visual impairment. Clearly, the symptoms prior to the onset of visual acuity impairment cause some burden. Effective and cost-effective ways of reducing the burden associated with trachoma exist. Despite the elegance of the SAFE strategy that has been promoted by the World Health Organization, it is clear that the primary cost-effective method
1729
1730
100
Financial Burdens and Disability-Adjusted Life Years in Loss of Vision Due to Trachoma
of reducing the burden associated with trachoma is by training surgeons and ensuring sufficient surgical supplies to make surgery available at the onset of trichiasis to minimize the discomfort and activity limitations associated with trichiasis even prior to visual acuity impairment and to prevent visual acuity impairment from occurring.
Summary Points Trachoma is a progressive condition affect poor individuals in poor parts of the world. Trachoma has been shown to have a disabling effects when the disease progresses to trichiasis (inturned eyelashes rubbing on the globe of the eye) and continues to progress to corneal opacity and blindness. The annual economic burden associated with lost productivity resulting from the condition has been shown to be in the billions of dollars. The single most cost-effective approach to limiting the disability and financial burden has been shown to be surgery after the disease has progressed to trichiasis. While mass treatment with azithromycin is not the most cost-effective strategy, this strategy is facilitated by donations from Pfizer that are distributed through the International Trachoma Initiative for mass treatment, which has been shown to lead to little excess antibiotic resistance.
References Baltussen RM, Sylla M, Frick KD, Mariotti SP. (2005). Ophthalmic Epidemiol. 12: 91–101. Bowman RJ, Soma OS, Alexander N, Milligan P, Rowley J, Faal H, Foster A, Bailey RL, Johnson GJ. (2000). Trop Med Int Health. 5: 528–533. Burton MJ, Frick KD, Bailey RL, Bowman RJ. (2002). Expert Opin Pharmacother. 3: 113–120. Durkin SR, Casson RJ, Newland HS, Aung TH, Shein WK, Muecke JS, Selva D, Aung T. (2007). Ophthalmology. 114: e7–e11. Emerson PM, Bailey RL, Mahdi OS, Walraven GE, Lindsay SW. (2000). Trans R Soc Trop Med Hyg. 94: 28–32. Emerson PM, Lindsay SW, Alexander N, Bah M, Dibba SM, Faal HB, Lowe KO, McAdam KP, Ratcliffe AA, Walraven GE, Bailey RL. (2004). Lancet. 363: 1093–1098. Evans TG, Ranson MK. (1995–1996). Int Ophthalmol. 19: 271–280. Evans TG, Ranson MK, Kyaw TA, Ko CK. (1996). Br J Ophthalmol. 80: 880–889. Frick KD, Basilion EV, Hanson CL, Colchero MA. (2003a). Ophthalmic Epidemiol. 10: 121–132. Frick KD, Hanson CL, Jacobson GA. (2003b). Am J Trop Med Hyg. 69(Suppl 5): 1–10.
Frick KD, Keuffel EL, Bowman RJ. (2001a). Ophthalmic Epidemiol. 8: 191–201. Frick KD, Lietman TM, Holm SO, Jha HC, Chaudhary JS, Bhatta RC. (2001b). Bull World Health Organ. 79: 201–207. Frick KD, Lynch M, West S, Munoz B, Mkocha HA. (2003c). Bull World Health Organ. 81: 101–107. Frick KD, Melia BM, Buhrmann RR, West SK. (2001c). Arch Ophthalmol. 119: 1839–1844. Frick KD, West SK. (2001). Ophthalmic Epidemiol. 8: 205–214. Goldschmidt P, Vanzzini Zago V, Diaz Vargas L, Espinoza Garcia L, Morales Montoya C, Peralta B, Mercado M. (2007). Rev Panam Salud Publica. 22: 29–34. Holm SO, Jha HC, Bhatta RC, Chaudhary JS, Thapa BB, Davis D, Pokhrel RP, Yinghui M, Zegans M, Schachter J, Frick KD, Tapert L, Lietman TM. (2001). Bull World Health Organ. 79: 194–200. Kumaresan JA, Mecaskey JW. (2003). Am J Trop Med Hyg. 69(Suppl 5): 24–28. Lansingh VC, Carter MJ. (2007). Surv Ophthalmol. 52: 535–546. Mabey D, Solomon AW. (2008). JAMA. 299: 819–821.
Financial Burdens and Disability-Adjusted Life Years in Loss of Vision Due to Trachoma Mathenge W, Kuper H, Limburg H, Polack S, Onyango O, Nyaga G, Foster A. (2007a). Ophthalmology. 114: 599–605. Mathenge W, Nkurikiye J, Limburg H, Kuper H. (2007b). PLoS Med. 4: e217. Melese M, Alemayehu W, Friedlander E, Courtright P. (2004). Trop Med Int Health. 9: 426–431. Oliva MS, Munoz B, Lynch M, Mkocha H, West SK. (1997–1998). Int Ophthalmol. 21: 235–241. O’Loughlin R, Fentie G, Flannery B, Emerson PM. (2006). Trop Med Int Health. 11: 1406–1415. Rabiu MM, Abiose A. (2001). Ophthalmic Epidemiol. 8: 181–190. Ranson MK, Evans TG. (1995–1996). Int Ophthalmol. 19: 261–270. Resnikoff S, Pascolini D, Etya’ale D, Kocur I, Pararajasegaram R, Pokharel GP, Mariotti SP. (2004). Bull World Health Organ. 82: 844–851. Rodgers AF, Ajono LA, Gyapong JO, Hagan M, Emerson PM. (2007). Trop Med Int Health. 12: 772–782.
100
Sche´mann JF, Guinot C, Traore L, Zefack G, Dembele M, Diallo I, Traore A, Vinard P, Malvy D. (2007). Acta Trop. 101: 40–53. Solomon AW, Mohammed Z, Massae PA, Shao JF, Foster A, Mabey DC, Peeling RW. (2005). Antimicrob Agents Chemother. 49: 4804–4806. Tellis B, Keeffe JE, Taylor HR. (2007). Commun Dis Intell. 31: 366–374. Thylefors B, Ne´grel AD, Pararajasegaram R, Dadzie KY. (1995). Bull World Health Organ. 73: 115–121. West ES, Munoz B, Imeru A, Alemayehu W, Melese M, West SK. (2006). Br J Ophthalmol. 90: 171–174. West ES, Munoz B, Mkocha H, Holland MJ, Aguirre A, Solomon AW, Bailey R, Foster A, Mabey D, West SK. (2005). Invest Ophthalmol Vis Sci. 46: 83–87. West S, Taylor HR. (1999). Surv Ophthalmol. 43: 468. West SK. (2004). Prog Retin Eye Res. 23: 381–401. West SK, Emerson PM, Mkocha H, McHiwa W, Munoz B, Bailey R, Mabey D. (2006). Lancet. 368: 596–600. Wright HR, Turner A, Taylor HR. (2007). Clin Exp Optom. 90: 422–428.
1731
101 The Economic Burden of Rheumatoid Arthritis: Asia/Thailand Perspective M. Osiri . A. Maetzel 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1734
2
Prevalence of RA in Developing Countries in Asia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1735
3
Healthcare Burden of RA in Asia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1736
4
Economic Burden of RA in Asia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1737
5
Healthcare Coverage Systems in Thailand . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1741
6
Drugs used in Rheumatic Diseases Listed in the National List of Essential Drugs of Thailand . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1742
7
Problems in RA Management in Thailand that Leads to Economic Burden of this Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1742 7.1 Delayed Diagnosis and Management/Referral . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1742 7.2 Suboptimal Treatment with DMARDs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1744 7.3 Additional Costs of Managing AEs from NSAIDs/GC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1744 8
Cost-effectiveness Analyses of DMARD Treatment of RA in Thailand . . . . . . . . . . 1745
9
Future Research Agenda Focusing on RA Burden in Thailand . . . . . . . . . . . . . . . . . . 1747
10
Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1748 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1749
#
Springer Science+Business Media LLC 2010 (USA)
1734
101
The Economic Burden of Rheumatoid Arthritis: Asia/Thailand Perspective
Abstract: > Rheumatoid arthritis (RA) is a worldwide disease that affects the working-age population. Evidence from industrialized countries has shown that RA poses a significant impact on the patients and societies in terms of healthcare resource utilizations, work productivity, physical disability, psychological effects and mortality. Compared with extensive information from Western countries, data on this disease in developing countries are scarce. Several factors contribute to the unawareness of RA burden and the estimated low prevalence: (1) a shortage of rheumatologists and other physicians trained to recognize the disease and (2) a perception of the disease symptoms as being benign relative to other more serious illnesses and thus not soliciting medical attention. > Non-steroidal anti-inflammatory drugs and > glucocorticoids (GC) abuses are common. Although novel therapy of RA with > biologic agents is considerably prescribed in Western countries, the access to these agents in Asian developing countries is difficult due to their high costs and lack of long-term cost effectiveness data. In Thailand, an economic study of RA has shown that RA incurred considerable expenditures for society. A > cost effectiveness analysis (CEA) has indicated that methotrexate-based combination > disease modifying antirheumatic drugs (DMARDs) are cost-effective in the treatment of RA. Further studies in every aspect of RA conducted in developing countries would help in directing resource allocation to properly improve the quality of RA care in these countries. List of Abbreviations: COPCORD, Community Oriented Program for the Control of Rheumatic Diseases; CSMBS, Civil Servant Medical Benefit Scheme; DMARDs, disease modifying antirheumatic drugs; GBP, British pound; GC, glucocorticoids; GP, general practitioner; HAQDI, > Health Assessment Questionnaire Disability Index; HRQoL, Health related quality of life; ICER, > incremental cost effectiveness ratio; ILAR, International League of Associations for Rheumatology; MTX, methotrexate; NLED, National List of Essential Drugs of Thailand; NSAIDs, non steroidal anti-inflammatory drugs; OTC, over-the-counter; PCU, Primary care unit; PPP, > Purchasing Power Parity; PT, Physical therapy; QALY, > quality-adjusted life-year; RA, rheumatoid arthritis; RCT, randomized controlled trial; SD, standard deviation; SF-36, Short-Form 36; SSS, Social Security Scheme; Thai FDA, Food and Drug Administration of Thailand; TNF, Tumor necrosis factor; UCS, Universal Coverage Scheme; UK, United Kingdom; USA, United States of America; USD, United States dollar; WHO, World Health Organization
1
Introduction
Rheumatoid arthritis (RA) is a chronic inflammatory disease affecting mainly the working-age population. This disease results in long-term joint damage, chronic debilitating pain, functional disability, and premature mortality (see key features of RA in Box 101‐1). The average prevalence of RA in Europe and North America is 1% (Alamanos et al., 2006). Thus, the healthcare burden of RA in these countries is tremendous. However, there is little evidence of RA burden in Asia, where billions of people reside. In Asia, RA is most likely not considered an important healthcare problem because it does not result in immediate death. Healthcare policy and resource allocation instead focuses on ‘fatal’ diseases with significant impact on the population, such as cancer, coronary heart disease, Human Immunodeficiency Virus (HIV) infection and acquired immunodeficiency syndrome (AIDS). Disease modifying anti-rheumatic drugs (DMARDs) are considered standard treatment for RA and now also include biologic agents, a new class of drugs that are highly effective in suppressing inflammation and retarding joint damage. However, the high cost of biologic
The Economic Burden of Rheumatoid Arthritis: Asia/Thailand Perspective
101
agents precludes their extensive use even in Europe and North America. A number of cost analyses of biologic agents were conducted to evaluate the economic benefit of these agents in Europe and North America (Bansback et al., 2004; Brennan et al., 2004; Choi et al., 2000, 2002, 2004; Kobelt et al., 2003; Tanno et al., 2006; Wailoo et al., 2008; Wong et al., 2002). However, there are substantial differences in healthcare financing among countries. The available data on economic burden and cost analysis from Europe and North America may, therefore, not be applicable to Asian countries. In this article, two objectives are addressed: (1) to review the data on the prevalence and economic burden of RA; and (2) to assess the economic value of RA treatment in Thailand and other countries in Asia.
. Box 101-1 Key features of rheumatoid arthritis RA is a systemic disease of unknown cause characterized by chronic inflammation and progressive destruction of multiple joints. The prevalence of RA is approximately 1%. The average age of onset is 30–40 years, with a female to male ratio of 3–4 to 1. The clinical features of RA include articular and extra-articular manifestations. Articular manifestations are chronic, progressive inflammation of small joints of hands and feet, wrists, elbows, shoulders, hips, knees, and ankles. This disease results in joint pain, swelling, loss of function and deformity. Extra-articular manifestations include vasculitis of skin and nerve, causing chronic skin ulcers and muscle weakness; inflammation and thinning of sclera; enlargement of spleen and low neutrophils; and subluxation of cervical spine. Patients with RA have 3–12 year-reduced life expectancy. The main cause of death is cardiovascular disease. The main treatment of RA is the early administration of disease modifying antirheumatic drugs, with methotrexate as a preferred agent. Non-steroidal anti-inflammatory drugs and glucocorticoids are used as adjunctive treatment to control pain and inflammation. Biologic agents, such as anti tumor necrosis factor, are newer class of drugs for controlling inflammation and retarding joint damage. These agents may be used alone or together with methotrexate. The content of this box describes the key facts of rheumatoid arthritis, including epidemiology, clinical features, outcomes and treatment. RA rheumatoid arthritis
2
Prevalence of RA in Developing Countries in Asia
The population of Asia in 2008 is approximately 3.8 billion, which accounts for more than 60% of the world population. Of these, about 200 million inhabit in more developed countries, i.e., Hong Kong, Japan, Singapore, South Korea and Taiwan. The remaining 3.6 billion live in the developing nations of Asia (Anonymous, 2006). In 1980s, the World Health Organization (WHO) collaborated with the International League of Associations for Rheumatology (ILAR) to investigate the rheumatic complaints and prevalence of common rheumatic diseases outside of Europe and North America. This program, Community Oriented Program for the Control of Rheumatic Diseases (COPCORD),
1735
1736
101
The Economic Burden of Rheumatoid Arthritis: Asia/Thailand Perspective
. Table 101‐1 Prevalence of rheumatoid arthritis in Asia Authors, year Darmawan et al., 1993
Country Indonesia
Hameed et al., 1995
Pakistan
Dans et al., 1997
Philippines
Type of community
Prevalence (%)
Rural
0.2
Urban
0.3
Urban poor
0.1
Urban affluent
0.2
Urban
0.17
Chaiamnuay et al., 1998
Thailand
Rural
0.12
Chopra et al., 2001
India
Rural
0.5
Minh Hoa et al., 2003
Vietnam
Urban
0.28
Haq et al., 2005
Bangladesh
Rural
0.7
Urban slum
0.4
Urban affluent
0.2
Veerapen et al., 2007
Malaysia
Semirural suburban
0.15
Zeng et al., 2008
China (mainland)
Rural
0.3–0.5
Urban
0.2–0.35
Taiwan
Rural
0.26
Urban
0.78–0.93
This table shows the prevalence of rheumatoid arthritis (RA) in Asian countries. RA prevalence in Asia is lower than that in Europe and North America. Urban communities generally have higher prevalence of RA than rural communities
provided the data on the prevalence of RA in Asia (Muirden, 2005). In addition, individual studies on the prevalence of RA in Asian countries have been conducted. In these studies, the prevalence of RA is estimated to vary between 0.2 and 0.5%, lower than that in Europe and North America. It was also found that the prevalence of RA increases in more urban settings, which may be the results of increased living standards with people becoming on the one hand less tolerant of the pain inflicted by RA and on the other hand having easier access to specialized medical care. The estimated point prevalences of RA in Asia are shown in > Table 101‐1. From > Table 101‐1, the prevalence of RA in Asian countries varied among different ethnic groups in different countries. Generally, the prevalence of RA in urban communities is higher than that in rural areas in South East Asia. Asian countries with developed societies, such as Taiwan and Hong Kong, had a comparable prevalence of RA with that in Western countries. Although the percentage of RA prevalence seems to be less than that in Europe and North America, the total number of RA patients in Asia is approximately 10 million, of which poses a significant burden on the developing societies.
3
Healthcare Burden of RA in Asia
The burden of RA encompasses functional disability and premature mortality, with an economic impact that is increasing everywhere in the world (Lundkvist et al., 2008).
The Economic Burden of Rheumatoid Arthritis: Asia/Thailand Perspective
101
Although previous studies in Asia and other developing countries showed RA to be milder with fewer extra-articular manifestations than in Europe and North America (Kalla,Tikly, 2003), the progressive westernization of developing societies may increase the recognition of RA at the early stage by better trained healthcare professionals, especially in the urban communities. Several prospective studies investigating the impact of RA on patients’ functional capacity, quality of life and mortality were conducted in Europe (Kvien, 2004) and North America (Allaire et al., 2008; Wolfe et al., 1994). These studies confirmed the substantial healthcare burden of RA on these societies. The data on the impact of RA in Asia are scant. Our group studied a 1-year change in functional ability of patients with established RA using different DMARDs with the pre-validated Thai version of the Health Assessment QuestionnaireDisability Index (HAQ-DI) (Osiri et al., 2006). Functional decline in all patients was observed regardless of the types of DMARDs prescribed. RA Patients receiving methotrexate (MTX) together with antimalarial drugs had the lowest reduction in functional ability while patients taking non-MTX based DMARDs had a more rapid progression in functional decline. The results of our study confirmed the findings from studies in Europe and North America on the advantage of using combination DMARDs with MTX as the anchor drug. Health-related quality of life (HRQoL) in patients with RA was also extensively studied in Europe and North America. HRQoL reflects physical, mental and social well-being of the patients (Kobelt, 2006). Almost all of the studies assessing HRQoL in RA patients used questionnaires measuring generic quality of life, such as Short-Form 36 (SF-36) or measured > utilities. Utility is defined as a preference that individuals have for a given state of health, ranged from 0 (death) to 1 (full health), which is measured as quality-adjusted life year (QALY). QALYs are calculated by weighting length of life with its quality (Torrance, 1986). QALYs are used for comparing HRQoL between RA and other diseases or among different treatment options. Reduction in life expectancy for patients with RA was also estimated to be considerable, as studies from Europe and North America showed that the average life expectancy of RA patients decreased by 3–12 years (Kvien, 2004; Wolfe et al., 1994). The most common causes of death in these patients were cardiovascular diseases, indicating that chronic inflammation may play a role in the development of premature atherosclerosis. The standardized mortality ratio (SMR) of patients with RA was estimated to vary between 2 and 3, indicating that RA patients were two to three times more likely to die than non-RA controls (Riise et al., 2001; Symmons et al., 1998; Wolfe et al., 1994; Yelin et al., 2002). To date, no study has estimated the mortality rate of RA patients in Asian countries. However, some authors suggest that infections may be the leading cause of death for RA patients in Asian countries (Kalla and Tikly, 2003). More evidence is needed to support this conclusion.
4
Economic Burden of RA in Asia
The economic burden of RA has also been assessed in many studies from Europe and North America. A systematic review reported the annual costs of RA to vary between USD 1,503 and USD 16,514 (Rosery et al., 2005). This large variation of the economic impact of RA may be due to the difference in the healthcare system among regions and/or countries, the category of cost included, and the viewpoint taken. The point of view in cost of illness studies may be that from the patient, payer, or society, with the societal viewpoint generally being the preferred approach.
1737
1738
101
The Economic Burden of Rheumatoid Arthritis: Asia/Thailand Perspective
In most cost of illness studies, costs are categorized into two to three groups: > direct costs, > indirect costs and > intangible costs (Kavanaugh, 2007; Pugner et al., 2000). Direct costs are those that are directly related to RA and are usually classified into two types: direct medical costs and direct non-medical costs. Direct medical costs are costs of RA treatment, e.g., prescribed medicines, laboratory tests, radiographs, out-patient consultations, hospitalizations, nursing and ancillary services, aids/device, physical therapy programs and modalities, over-the-counter drugs and co-payments, and alternative medicines. Direct nonmedical costs are not related to treatment and include costs incurred for home carers and helpers, daycare services, transportations, private health insurance premiums and social security payments. Indirect costs comprise productivity loss, work absenteeism, early retirement and premature death due to RA. Intangible costs are the costs of reduced quality of life, including lost leisure opportunities, lost social opportunities, pain, anxiety, and depression. This type of cost is generally difficult to define in monetary term. We conducted a prospective study to assess the annual direct and indirect costs of RA in Thai patients attending the RA clinic at a medical school hospital in Bangkok, Thailand (Osiri et al., 2007a). Direct costs accounted for almost 80% of total costs of RA. Hospitalization and RA medication costs contributed the largest proportion of the direct costs in our study. Details on cost categories contributed to the costs of RA are present in > Figure 101‐1. Indirect costs are the major cost component in studies from Europe and North America. Approximately 50% of costs fall into this category, since productivity loss, sick leave, premature retirement or death pose a significant financial burden in Europe and North America (Guillemin et al., 2004; Jacobsson et al., 2007; Maetzel et al., 2004; Meenan et al., 1978; Pe´ntek et al., 2007; Ruoif et al., 2003; Yelin, 1996). Earlier studies have shown a significant discrepancy . Figure 101‐1 Percentage of cost categories contributed to the total costs of rheumatoid arthritis (Osiri et al., 2007a)
The Economic Burden of Rheumatoid Arthritis: Asia/Thailand Perspective
101
between direct and indirect costs. The cost of RA studies conducted after 1999 in Europe and North America has shown a higher proportion of direct costs due to the introduction of biologic agents, particularly anti tumor necrosis factor (TNF) therapy. These agents, although prescribed in a small number of patients, accounted for a substantial proportion of RA-related medical costs (Guillemin et al., 2004; Jacobsson et al., 2007; Maetzel et al., 2004; Ruoif et al., 2003). On the other hand, direct costs were the major cost component in economic studies of RA in countries from Asia (Aggarwal et al., 2006; Osiri et al., 2007a). Both studies were conducted before the availability of anti-TNF agents. Should the costs of biologic agents be included, direct medical cost will be expected to constitute a much larger proportion of total costs of RA than traditional DMARDs. > Table 101‐2 shows the percentage of direct and indirect costs from RA cost of illness studies from developed countries in North America and Europe and developing countries in Asia. The data from > Table 101‐2 suggest that differences in living standards and economic wealth mainly influence the proportion of costs of RA. The high proportion of direct costs for RA in Asian countries may be explained by the selection bias of RA patients with moderate to severe disease, since both studies were conducted in tertiary care referral hospitals; and the dependence on imported medicines and technology from developed countries. Another factor that may contribute to this high direct cost was the delay in DMARD prescriptions. This may be due to the unavailability of rheumatologists in developing countries and the lack of
. Table 101-2 Percentage of direct and indirect costs in economic analyses of rheumatoid arthritis from different countries % of total cost Country, author, year of publication
Direct cost
Indirect cost
USA, Meenan et al., 1978
25.4
74.6
USA, Yelin, 1996
28.3
71.7
55.1
44.9
44.9
55.1b
21.7
78.3c
France, Guillemin et al., 2004a
59.3
40.7
Sweden, Jacobsson et al., 2007a
41
55
Hungary, Pe´ntek et al., 2007
45.2
54.8
India, Aggarwal et al., 2006
69.3
30.7
Thailand, Osiri et al., 2007a
79.6
20.4
Developed countries
Canada, Maetzel et al, 2004
a
Germany, Ruoif et al., 2003a
Developing countries
The percentages of direct and indirect costs of rheumatoid arthritis from different countries in Europe and North America are compared with those from Asian developing countries. Indirect costs were the major cost component in Western countries while direct costs were the main component in Asian developing countries. Inclusion of biologic agent costs increases the direct costs in Western countries a The costs of biologic agents (tumor necrosis factor (TNF) inhibitors) were included in direct medical costs b Indirect cost calculated from sick leave c Indirect cost calculated from work disability and early retirement
1739
1740
101
The Economic Burden of Rheumatoid Arthritis: Asia/Thailand Perspective
internists’ knowledge in proper treatment of RA, which led to the abuse of glucocorticoids (GC) and/or nonsteroidal anti-inflammatory drugs (NSAIDs). The relatively low proportion of indirect costs for RA in Thailand compared with those in Europe and North America may be explained by the lower income of the Thai population. In Thailand, the majority of the studied patients were housewives, manual-labor workers, farmers, and low-rank government officers. Therefore, productivity loss and sick leave owing to RA was the lowest compared with those in Western countries and India. Underestimation of indirect costs in Thailand and India may also be because other cost components were not included in this cost category. For example, the patient’s reduced work performance without absenteeism and increased workload to co-workers might not be accounted for productivity loss in both studies. When the data on the studied patients and cost categories were considered, there were several differences in the cost of RA studies from India and Thailand, as shown in > Table 101‐3.
. Table 101‐3 Cost of illness studies of RA in Thailand and India Variables
Thailand (Osiri et al., 2007a)
India (Aggarwal et al., 2006)
158
101
Demographic data Number of patients Female:male (%)
151: 7 (95.6:4.4)
90:11 (89:11)
MeanSD age (years)
53.2 12.2
43.2 11.7
MeanSD disease duration (years)
10.3 7.8
8.1 5.6
MeanSD HAQ-DI
0.69 0.75
0.97 0.69
DMARDs treatment (%)
152 (96.2)
96 (95)
6,475 5,101
17,670 9,489
84 (53.2)
10 (9.9)
Direct costs
2,135 3,601
1,482 1,328
Direct medical cost
1,923 3,556
MeanSD annual income (USD)a Number with full healthcare coverage (%) Cost of RA (meanSD in USD)a
Direct non-medical cost
213 617
Indirect costs
547 1449
656 2,198
2,682 4,637
2,138 2,832
Total costs
This table compares the studied patients characteristics, economic status and cost components of RA between Indian and Thai studies. Although the demographic data of the patients and costs of RA were not significantly different, the annual income of patients from India was significantly higher than that of the Thais. The percentage of healthcare coverage, however, was higher in Thai patients. RA rheumatoid arthritis; : standard deviation; HAQ-DI Health Assessment Questionnaire Disability Index; DMARDs disease modifying anti-rheumatic drugs; USD United State dollar a Purchasing power parity estimates for the year 2000 were used for currency conversion, Thailand: 11.88 baht = 1 USD, India: 7.84 rupees = 1 USD
The Economic Burden of Rheumatoid Arthritis: Asia/Thailand Perspective
101
From > Table 101‐3, the average annual costs of RA in Thailand and India were almost similar although the Thai RA patients were older and had longer disease duration. However, Thai RA patients were less functionally disabled than patients from India. Direct costs from the Indian study comprised mainly direct medical costs, while direct nonmedical costs were included in the indirect cost category. Thus, average direct cost of RA in the Thai study was higher than that in the Indian study. The dependence of imported medications (DMARDs and NSAIDs) and technology (monitoring costs) in Thailand may be another reason for higher average direct cost compared with India. The high annual income of the Indian RA patients indicated that more affluent patients might have better access to proper treatment in tertiary care centers as very few patients had full healthcare reimbursement. If the average total cost of RA from Thailand is used to calculate the overall economic burden of RA in developing countries in Asia, the annual cost of RA would approximate USD 27 billion, not accounting the costs of biologic treatment. Several models used for calculating the lifetime cost of RA have been reported from developed countries. The suggested models from Europe and North America may not be directly applicable to Asia due to the discrepancies in the nature of RA itself, healthcare coverage systems, and the availability of high cost treatments. Although the disease progression seems to be milder in Asian RA patients, the delay or inadequate treatment with DMARDs and the lack of social support in these patients would result in a worse outcome compared with RA patients in Europe or North America. Biologic agents are currently not listed in the National List of Essential Drugs (NLED) of Thailand due to their high costs. Very few patients are able to get access to this kind of treatment, usually at the later stage of the disease. Therefore, the cost offsets due to the effectiveness of biologic agents may not be observed.
5
Healthcare Coverage Systems in Thailand
To better understand the impact of RA and its treatment on the Thai patients and the nation, healthcare coverage systems in Thailand should be considered. The current healthcare coverage systems in Thailand are classified into three groups (Hughes and Leethongdee, 2007; Sreshthaputra and Kaemthong, 2001): 1. Civil Servants Medical Benefit Scheme (CSMBS) 2. Social Security Scheme (SSS) and Workmen’s Compensation Scheme 3. Universal Coverage scheme The CSMBS provides full healthcare coverage to government officers, state enterprise employees and their family members in public hospitals for both ambulatory and in-patient care. High cost drugs not included in the NLED may be reimbursed by the government after patients make their own advance payment. Partial coverage for in-patient healthcare services in private hospitals is also provided. The CSMBS covers a total number of 5 million people – approximately 8% of the population in Thailand – with expenditures increasing dramatically every year. The SSS provides healthcare coverage to formal private-sector employees and some publicsector employees. All employees covered in this scheme as well as their employers contribute 5% of their gross wage for healthcare insurance in the Social Security Fund. Healthcare services have to be delivered by pre-specified providers. Healthcare services in other hospitals
1741
1742
101
The Economic Burden of Rheumatoid Arthritis: Asia/Thailand Perspective
are not reimbursed except for emergency cases or formal referral. The SSS covers the expenses of drugs listed in the NLED only; otherwise patients have to pay by themselves. This scheme covers approximately 8 million people (13%) in Thailand. The UC scheme was introduced in 2001 to replace the Healthcare Card Scheme and Health Welfare for the low income and underprivileged groups. This scheme provides basic coverage for the majority of the Thai population, about 50 million people or 75% of the population. For each visit or admission, patients were required to pay a fee of 30 baht (USD 2.50) but this fee is already waived. Similar to the SSS, people covered by the UC scheme are required to attend a pre-specified primary care unit (PCU) or hospital. Drugs listed in the NLED are available without any copayment but those not listed in the NLED are not covered.
6
Drugs used in Rheumatic Diseases Listed in the National List of Essential Drugs of Thailand
Whether a drug should be included in the NLED is proposed by working groups of subspecialty experts. The choices of the drugs depend on their efficacy, safety, and costs. The final decision will be made by the committee from the Food and Drug Administration of Thailand (Thai FDA). In the past, the NLED was designed as a ‘minimal’ list of drugs that should be available to satisfy basic healthcare needs in developing countries, according to WHO. The revised version of the NLED has now become a ‘maximal’ list for government hospitals (Anonymous, 1999). However, the list of drugs for musculoskeletal and joint diseases seems not to change (Anonymous, 2008). As shown in > Table 101‐4, DMARDs available in community (primary care) hospitals are antimalarial drugs (chloroquine and hydroxychloroquine) only. Methotrexate (MTX), sulfasalazine (SSZ), parenteral gold salt, azathioprine and cyclosporine are limited in provincial or general hospitals and tertiary care hospitals. Despite their effectiveness, leflunomide and biologic agents are currently not listed in the NLED due to their high costs. Attempts have been made to include these agents into a special list of the NLED that requires patient registration and close monitoring. However, the latest released NLED (Feb 15, 2008) in the Thai FDA website is similar to that for the year of 2004.
7
Problems in RA Management in Thailand that Leads to Economic Burden of this Disease
Similar to countries in Asia and other continents, problems encountered in the management of RA in Thailand include the delay in the diagnosis and treatment/referral, suboptimal treatment with DMARDs, and management of adverse events and complications from long-term NSAIDs and/or GC use. These problems pose a significant economic burden on the nation but this impact has not received adequate attention by government and healthcare decision makers.
7.1
Delayed Diagnosis and Management/Referral
Thai patients developing the symptoms of polyarthritis are usually self-treated by NSAIDs and/or GC bought from local pharmacies. After the disease progresses or complications
The Economic Burden of Rheumatoid Arthritis: Asia/Thailand Perspective
101
. Table 101‐4 List of drugs used in rheumatic diseases in the National List of Essential Drugs of Thailanda Drugs
Categoryb
1. Non-steroidal anti-inflammatory drugs Aspirin
A
Diclofenac sodium
A
Indomethacin
A
Naproxen
A
Piroxicam
B
2. Disease-modifying antirheumatic drugs Chloroquine phosphate
B
Hydroxychloroquine sulfate
B
Azathioprine
C
D-Penicillamine
C
Sulfasalazine
C
Cyclosporin
C
Methotrexate
C
Sodium aurothiomalate
C
a
Updated Feb 15, 2008 There are five categories of the drugs listed in the NLED: A: Standard drugs for all classes of healthcare units, should be used first according to the indications B: Drugs used in specific diseases as substitutes of drugs in the ‘A’ list C: Drugs used in specific diseases by experts due to their potential toxicity or cost-ineffectiveness D: Drugs, usually high cost ones, used in specific indications and prescriptions are limited to experts. Drug utilization evaluation system is required to monitor the use of these drugs E: Drugs for special projects of government organizations or high cost drugs for specially-needed patients that required patient registration system
b
develop, they would then seek help from a hospital. This may take weeks, months, or even years before they are given the right treatment. Early RA is generally treated with alternating NSAIDs or short-course GC. Once typical features of RA occur, especially joint deformity, DMARDs are started. This delay may be partly owing to the patients themselves but the other important reason is the lack of rheumatologists in Thailand. To date, there are approximately 80 board-certified rheumatologists in Thailand. The proportion of rheumatologists per population (64 million) is about 1 per 800,000. In addition, the availability of rheumatologists is not evenly distributed throughout the country. More than half reside in Bangkok, mainly in private hospitals and medical schools. Fewer than 10% of general hospitals in each province have rheumatologists. The reasons why there are few rheumatologists in Thailand may be because this subspecialty is relatively new compared with other subspecialties in Internal Medicine. Very few internists are interested in rheumatology training as this field contains no procedure that can indicate that they are truly a ‘specialist’. For example, a gastroenterologist can perform endoscopy but any internist can perform arthrocentesis. Rheumatic diseases are usually considered chronic, benign but incurable diseases. The impact of these diseases is not comparable to vehicle accidents, coronary heart diseases, diabetes mellitus, cancer and
1743
1744
101
The Economic Burden of Rheumatoid Arthritis: Asia/Thailand Perspective
HIV/AIDs. In addition, the Ministry of Public Health does not recognize the benefit of having a rheumatologist in a provincial hospital. As a result, grants for rheumatology training are usually not provided as rheumatic diseases are considered diseases for orthopedists. The delay in the diagnosis and treatment of RA will result in a more severe disease with advanced joint damage, the need of more potent DMARDs or biologic agents to control the disease, the patient’s disability and complications from treatment, and finally, the economic burden to the society and nation.
7.2
Suboptimal Treatment with DMARDs
The restriction of MTX availability in primary care hospitals decreases the opportunity of Thai patients with early RA to be offered the chance of remission. From the National Drug Committee’s point of view, MTX is chemotherapy and too dangerous to be prescribed by general practitioners (GPs) or internists. Thus, the only DMARD available to RA patients in community hospitals are antimalarials, mainly chloroquine. Antimalarial drugs are usually prescribed together with GC. Although GC has good evidence for its ability to delay joint damage, improve inflammation and functional ability in short term, long-term GC use results in serious adverse events and complications, which incur a substantial amount of expenses. On the other hand, there has been no evidence for antimalarials efficacy in delaying joint damage. Prolonged use of antimalarials without scheduled eye monitoring may result in visual impairment or even blindness. In all primary care hospitals, basic laboratory tests, i.e., complete blood count, creatinine, liver function test, are available. Monitoring of MTX adverse effects may be much easier than comprehensive eye examination by an ophthalmologist in this situation. Some RA patients with active and advanced disease do not respond to MTX alone or MTX-based combination of DMARDs listed in the NLED. These patients may require treatment with the drugs not listed in the NLED or surgical intervention. Patient referral to secondary, tertiary care hospitals or medical schools for a proper management of RA requires the transfer of budget from PCUs to these hospitals. Prescription of biologic agents or leflunomide for these patients at higher-level hospitals is not allowed since the PCUs would pay for the drugs listed in the NLED only. Otherwise the patients themselves or tertiary care hospitals have to be responsible for this expense. Thus, the opportunity that an RA patient covered by the UC or SSS scheme would receive these high cost drugs is to participate in a clinical trial, negotiate for compassionate provision of product by pharmaceutical companies, or encourage the National Drug Committee to include these agents in the NLED.
7.3
Additional Costs of Managing AEs from NSAIDs/GC
RA patients in Thailand are usually taking NSAIDs and/or GC for a long time before DMARDs are initiated. A number of patients are unable to discontinue NSAIDs and/or GC despite adequate doses of DMARDs. Long-term exposure to NSAIDs and/or GC results in significant adverse events and complications, e.g., gastrointestinal bleeding/perforation/stricture, cardiovascular complications, renal insufficiency, osteoporosis and fracture, cataracts, glaucoma, diabetes, dyslipidemia, infections. These complications incur a substantial
The Economic Burden of Rheumatoid Arthritis: Asia/Thailand Perspective
101
economic impact on the patients, families, societies, and nation. Therefore, early and timely treatment of RA with appropriate DMARDs will decrease long-term use of NSAIDs and/or GC and their complications. Cost effectiveness analyses of DMARDs conducted in Thailand are necessary because they will provide important data in choosing the right DMARDs for Thai patients with RA at acceptable expense.
8
Cost-effectiveness Analyses of DMARD Treatment of RA in Thailand
Cost effectiveness analysis is a tool for policy makers to assess which interventions provide the highest ‘value for money’ as resources is limited and there are a lot of healthcare interventions to choose from. RA is among the first diseases with extensive studies on the cost effectiveness of its treatments. It must be emphasized that cost analyses are country-specific since healthcare coverage systems are different among countries. Thus, cost effectiveness analyses using data from industrialized countries may not be applicable in developing countries without adjustment (Kobelt, 2006). In a cost effectiveness analysis, the incremental cost effectiveness ratio of a certain treatment is calculated from the cost incurred from this new treatment minus the cost incurred from standard treatment divided by the differences between the effectiveness of the new and standard treatment, as shown in the formula below: Incremental cost Cost of new treatment cost of standard treatment ¼ Effectiveness of new treatment effectiveness of standard treatment effectiveness ratio ðICERÞ
Several ways to measure the effectiveness of treatment for the ICER are available. If the effectiveness is measured as standard outcome measures for response to treatment in a disease, such as the HAQ-DI, > Disease Activity Score (DAS) in RA, this type of economic evaluation is called cost effectiveness analysis. If the effectiveness is measured as utility or QALY, the study is called cost utility analysis. When the effectiveness is defined in monetary term, which has the same unit as cost, this kind of economic evaluation is called cost benefit analysis. Treatment of RA with a new therapeutic option is expected to be more effective at a higher cost compared with standard treatment. The acceptable threshold of ICER of a treatment option is USD 50,000, 50,000 Euro, or GBP 30,000 per QALY, if utility is used for measuring effectiveness (Kavanaugh, 2007). This means a new treatment of RA will be considered costeffective if it incurs an additional cost of not more than this limit. To date, there is no proposed limit of ICER in developing countries in Asia, although WHO has suggested the use of three times national income per capita (Kobelt, 2006). Numerous cost effectiveness analyses of DMARDs and biologic agents, especially TNF inhibitors, have been conducted in industrialized countries (Wong et al., 2002; Kobelt et al., 2003; Bansback et al., 2004; Brennan et al., 2004; Chiou et al., 2004). Although anti-TNF agents have been proven for their efficacy in randomized controlled trials (RCTs) and openlabel extension studies of these RCTs, their uses as first-line agents instead of traditional DMARDs are not recommended. This is partly because of their unclear long-term effectiveness since many patients develop resistance to these agents after a period of response. The other important reason is because of their high costs. In Europe or North America, the annual costs of TNF inhibitors usually exceed USD 10,000. Although the use of these agents may help decrease the direct medical costs incurred from other treatments, e.g., treatment of complications from NSAIDs or GC, costs of traditional DMARDs and analgesics, these savings cannot
1745
1746
101
The Economic Burden of Rheumatoid Arthritis: Asia/Thailand Perspective
compensate the high costs of these agents (Kobelt, 2006), especially when considering the short-term cost effectiveness of RA treatment with biologic agents. Since RA is a chronic progressive disease, the cost components of the cost effectiveness studies should include the lifetime costs of this disease and the cost savings from treatments. Similarly, the effectiveness of new treatments of RA should consider the suppression of joint inflammation and delay in disease progression, especially functional disability and deformity. Thus, several models of long-term costs and effectiveness of these agents have been proposed. These models use data on costs and effectiveness of TNF inhibitors from different sources, mostly RCTs and their extension studies. These data may not be accurate enough to estimate the long-term cost and effectiveness of these agents but the data from inception cohorts or meta-analyses of RCTs are not sufficient. Therefore, these models inevitably have to rely on assumptions of several parameters. This results in different conclusions and is subject to criticism as evident from the cost effectiveness models of anti-TNF agents for RA treatment reported from Europe and North America (> Table 101‐5) (Kobelt, 2006; Kavanaugh, 2007). As shown in > Table 101‐5, the ICER for each QALY gained from using anti TNF agents was approximately USD 30,000. This is considered cost effective because the ICER does not exceed the limit of USD 50,000. However, the National Institute of Clinical Excellence of the United Kingdom has proposed a model of lifetime cost effectiveness analysis of TNF inhibitors and their results have shown that TNF inhibitors are cost-effective only in patients who fail at least two traditional DMARDs (one of them has to be MTX) (Chen et al., 2006). In Thailand, the cost effectiveness analysis of biologic agents for RA treatment has never been reported. Only one cost effectiveness study of traditional DMARDs was conducted in a tertiary care and teaching hospitals (Osiri et al., 2007b). Compared with antimalarial drugs alone for the treatment of RA, MTX plus antimalarials are more effective in improving the HAQ-DI score and cost less, that is, this combination dominates antimalarial drugs. MTX alone, MTX plus other DMARDs, and leflunomide are more effective than antimalarials but incur higher costs. DMARDs other than MTX or non-MTX based combination DMARDs are not cost-effective. The ICERs of each DMARD treatment option are shown in > Table 101‐6. . Table 101‐5 The incremental cost effectiveness ratio of tumor necrosis factor (TNF) inhibitors in the cost utility analysis modelsa Agents
Comparator
Author, year
Country
Cost per QALY (USD)
Infliximab + MTX MTX
Wong et al., 2002
USA
30,500
Infliximab + MTX MTX
Kobelt et al., 2003
Sweden and UK
34,320
Etanercept
Brennan et al., 2004
UK
29,394
DMARDs
Etanercept
Anakinra
Chiou et al., 2004
USA
13,387
Adalimumab
Anakinra
Chiou et al., 2004
USA
91,927
Adalimumab
MTX, DMARDs Bansback et al., 2004 Sweden
28,924
This table shows the incremental cost effectiveness ratio as USD of different TNF inhibitors per one unit of QALY gained compared with MTX, other DMARDs, and interleukin-1 receptor antagonist (anakinra). Anti-TNF treatments, either with or without MTX, are cost effective when compared with traditional DMARDs, because the cost per QALY did not exceed USD 50,000. TNF tumor necrosis factor; MTX methotrexate; QALY quality-adjusted life-year; USD United States dollars; USA United States of America; UK United Kingdom; DMARDs disease modifying antirheumatic drugs a Adapted from Kavanaugh, 2007
The Economic Burden of Rheumatoid Arthritis: Asia/Thailand Perspective
101
. Table 101‐6 Incremental cost effectiveness ratio of different DMARDs compared with antimalarial drugs for RA treatment in Thailand (Osiri et al., 2007b) Mean societal cost (USD)
Mean change in HAQ-DI scorea
2,285
0.68
Reference
MTX
3,080
0.29
2,061
MTX + AM
1,841
0.15
834 (dominates)b
MTX + SSZ
2,572
0.22
625
AM + SSZ
3,692
0.58
14,482
MTX + AM + SSZ
2,993
0.1
1,222
Leflunomide
3,310
Other DMARDs
3,507
DMARDs Antimalarials (AM)
0.25 0.6
ICER
1,104 16,016
Incremental cost effectiveness ratio of DMARDs compared with AM in a Thai RA study is shown. Combined MTX and AM was more effective but cost less than AM. MTX alone, other MTX-based combination DMARDs, and leflunomide were more effective and also cost more. Non-MTX based DMARDs were not cost effective. DMARDs: disease modifying antirheumatic drugs, RA rheumatoid arthritis; USD United States dollar; HAQ-DI Health Assessment Questionnaire Disability Index; ICER incremental cost effectiveness ratio; AM antimalarial drugs; MTX methotrexate; SSZ sulfasalazine a Negative value represents deterioration and positive value represents improvement of mean change of HAQ-DI b MTX + AM costs less but delay functional disability better than AM alone
Our results supported the findings from those in Western countries that MTX-based combination DMARDs were more effective than single agents. However, the lack of data on the cost effectiveness of biologic agents in Thailand precludes policy makers and reimbursement authorities from considering these agents essential drugs in the National Drugs list. Reimbursement authorities concern only the high costs of these agents that will blow up the healthcare budget of the country, while the potential long-term benefits from indirect cost savings are excluded. As the annual costs of TNF inhibitors in Thailand are far exceeded three times the gross national income per capita, these agents are thus considered cost ineffective in the view of decision makers.
9
Future Research Agenda Focusing on RA Burden in Thailand
As mentioned earlier, RA is unrecognized by policy makers, reimbursement authorities, and even healthcare providers in Thailand and other developing countries in Asia because of its low prevalence and insignificant immediate impact on the society. Thus, the burden of this disease is undervalued and novel therapy with biologic agents is considered cost-ineffective. To increase public awareness of RA and its burden as well as to improve the quality of care, several steps have to be undertaken as the following: 1. Educational programs for the population, community leaders and healthcare providers regarding the symptoms and signs of RA, the need for early treatment with appropriate DMARDs, and early referral to rheumatologists, and the complications emerging from abuse of NSAIDs and/or GC.
1747
1748
101
The Economic Burden of Rheumatoid Arthritis: Asia/Thailand Perspective
2. Support the formation of RA patient network to update the knowledge of this disease and its treatment, and help their members in need. 3. Promote the widespread and early use of MTX, mount the education campaigns about the manageable profile of MTX and monitor the use of MTX by RA patients, especially the early use. 4. Standardization of the RA patient information forms. This is the initiation of inception cohorts conducted in hospitals with rheumatologists. Long-term data on the disease progression, functional status, complications, and costs may be achieved by this method. These data will then provide important information required for the appropriate DMARDs use in the country and monitoring methods. This may lead to a clinical practice guideline of RA specifically generated for the country. 5. Rheumatology training grants should be provided to internists. Rheumatologists should be evenly distributed throughout the country. The availability of rheumatologists in every provincial hospital would facilitate the early diagnosis and treatment of RA patients and thus delay functional disability and improve quality of life in the patient population. 6. Occasional surveys on the prevalence of RA and its impact on the patients and their families. This is to observe the trend of RA prevalence over time and assess the burden of this disease in patients seen by GPs, internists or other subspecialists. 7. Lobbying the government and healthcare funders to increase awareness of RA and its burden on the nation. This is to perform after the availability of economic and quality of life data from long-term prospective studies. Indirect costs which are saved from improvement of RA have to be emphasized over the concern of drug acquisition costs alone. Funding to support RA-related educational programs and treatments should be initiated. 8. Initiation of patient registration system for RA patients who require treatment with biologic agents. Patient registration system would provide the information on treatment response and complications, in addition to limit the use of these high cost drugs to those really in need. As biologic agents, especially TNF antagonists, have been used in RA for not more than two decades, their long-term safety is still unclear. However, their short-term adverse events are known; particularly increases in the incidence of tuberculosis (TB). Since Thailand is an endemic area of TB, extensive studies and guidelines are required to better prevent and handle the infections that may occur after the initiation of these agents. 9. Researches on the efficacy of herbal medicines as complimentary treatments should be encouraged. Many plants may exhibit the anti-inflammatory and/or immunomodulatory effects. This may help decrease the amount of imported medications and the country expenditure.
10
Conclusions
RA is not perceived as one of the important diseases in Asian developing countries. In fact, the economic burden of RA in Asia is substantial because of the high population in this continent, inadequate knowledge and support from the government, complications from treatment drug abuse, and productivity losses resulted from progressive disability and premature deaths. The last factor is difficult to estimate and usually ignored by government and policy makers.
The Economic Burden of Rheumatoid Arthritis: Asia/Thailand Perspective
101
Increase awareness and educations in RA will improve the quality of care and decrease disability. Prospective research studies in the field of RA will provide essential data on the appropriate DMARDs prescribed and monitoring tests, costs incurred and saved from different treatment options, and cost effectiveness of each DMARD and novel agent. These data will help guide the government, policy makers and reimbursement authorities to suitably allocate the resources to improve the standard of care for all RA patients in the country.
Summary Points RA affects approximately 10 million people in Asian developing countries. Several factors contribute to poor outcomes in these patients: delay in diagnosis and
treatment, inadequate dosages and choices of DMARDs, NSAIDs and GC abuse, lack of multidisciplinary collaboration, and inaccessibility to novel biologic treatment. Economic evaluation of RA in Asian developing countries has shown that direct medical costs constitute the major cost component of total costs of RA. MTX-based combination DMARDs have shown to be most effective in delaying functional decline in Thai patients with RA. MTX-based combination DMARDs, especially MTX plus antimalarial drug, are the most cost effective options in RA treatment in Thailand. Early recognition and treatment of RA with MTX-based combination DMARDs is crucial for the suppression of inflammation and retardation of functional disability. Due to their high acquisition costs, biologic agents are usually inaccessible for RA patients from developing world, even they are fully indicated for these agents. Cost effectiveness studies using long-term data from developing countries are required for resource allocation to RA treatment. Data from developing countries may not be applicable in developing countries.
References Aggarwal A, Chandran S, Misra R. (2006). Natl Med J India. 19:187–191. Alamanos Y, Voulgari PV, Drosos AA. (2006). Semin Arthritis Rheum. 36:182–188. Allaire S, Wolfe F, Niu J, Lavalley MP. (2008). Arthritis Rheum. 59:474–480. Anonymous. (1999). Consumer Project on Technology. access date April 4, 2008. Anonymous. (2006). German Foundation for World Population (Deutsche Stiftung Weltbevo¨lkerung). DSW-data report 2006. Anonymous. (2008). National List of Essential Drugs of Thailand. Food and Drug Administration of Thailand. Last update (in Thai) February 15, 2008.
Bansback NJ, Brennan A, Sengupta N, Pang F. (2004). Ann Rheum Dis. 63 (supplement I):512–513. Brennan A, Bansback N, Reynolds A, Conway P. (2004). Rheumatology. 43:62–72. Chaiamnuay P, Darmawan J, Muirden KD, Assawatanabodee P. (1998). J Rheumatol. 25:1382–1387. Chen Y-F, Jobanputra P, Barton P, Jowett S, Bryan S, Clark W, Fry-Smith A, Burls A. (2006). Health Technol Assess. 10:1–227. Chiou CF, Choi J, Reyes CM. (2004). Expert Rev Pharmacoeconomics Outcomes Res. 4:307–315. Choi HK, Seeger JD, Kuntz KM. (2000). Arthritis Rheum. 43:2316–2327. Choi HK, Seeger JD, Kuntz KM. (2002). J Rheumatol. 29:1156–1165. Chopra A, Patil J, Billempelly V, Relwani J, Tandle HS, WHO-ILAR COPCORD Study. (2001). J Assoc Physicians India. 49:240–246.
1749
1750
101
The Economic Burden of Rheumatoid Arthritis: Asia/Thailand Perspective
Dans LF, Tankeh-Torres S, Amante CM, Penserga EG. (1997). J Rheumatol. 24:814–1819. Darmawan J, Muirden KD, Valkenburg HA, Wigley RD. (1993). Br J Rheumatol. 32:537–540. Guillemin F, Durieux S, Daure`s J-P, Lafuma A, Saraux A, Sibilia J, Bourgeois P, Sany J. (2004). J Rheumatol. 31:1297–1304. Hameed K, Gibson T, Kadir M, Sultana S, Fatima Z, Syed A. (1995). Br J Rheumatol. 34:252–256. Haq SA, Darmawan J, Islam MN, Uddin MZ, Das BB, Rahman F, Chowdhury MA, Alam MN, Mahmud TA, Chowdhury MR, Tahir M. (2005). J Rheumatol. 32: 348–353. Hughes D, Leethongdee S. (2007). Health Affairs. 26:999–1008. JacobssonLTH, Lindroth Y, Marsal L, Juran E, Bergstro¨m U, Kobelt G. (2007). Scand J Rheumatol. 36:179–183. Kalla AA, Tikly M. (2003). Best Practice Res Clin Rheumatol. 17:863–875. Kavanaugh A. (2007). Best Practice Res Clin Rheumatol. 21:929–942. Kobelt G. (2006). Scand J Rheumatol. 35:415–425. Kobelt G, Jonsson L, Young A, Eberhardt K. (2003). Rheumatology. 42:326–335. Kvien TK. (2004). Pharmacoeconomics. 22(supplement I):1–12. Lundkvist J, Kastang F, Kobelt G. (2008). Eur J Health Econ. 8(supplement 2):S49–S60. Maetzel A, Li LC, Pencharz J, Tomlinson G, Bombardier C, the Community Hypertension and Arthritis Project Study Team. (2004). Ann Rheum Dis. 63:395–401. Meenan RF, Yelin EH, Henke CJ, Curtis DL, Epstein WV. (1978). Arthritis Rheum. 21:827–833. Minh Hoa TT, Darmawan J, Chen SL, Van Hung N, Thi Nhi C, Ngoc An T. (2003). J Rheumatol. 30:2252–2256. Muirden KD. (2005). Curr Opin Rheumatol. 17:153–156. Osiri M, Deesomchok U, Tugwell P. (2006). Clin Rheumatol. 25:721–727. Osiri M, Kamolratanakul P, Maetzel A, Tugwell P. (2007). Rheumatol Int. 27: 1063–1069 (Osiri et al, 2007b).
Osiri M, Maetzel A, Tugwell P. (2007). J Rheumatol. 34:57–63 (Osiri et al, 2007a). Pe´ntek M, Kobelt G, Czirja´k L, Szekanecz Z, Poo´r G, Rojkovich B, Polga´r A, Genti G, Kiss CG, Brodszky V, Ma´jer I, Gula´csi L. (2007). J Rheumatol. 34:1437. Pugner KM, Scott DI, Holmes JW, Hieke K. (2000). Semin Arthritis Rheum. 29: 305–320. Riise T, Jacobsen BK, Gran JT, Haga HJ, Arnesen E. (2001). Clin Rheumatol. 20: 123–127. Rosery H, Bergemann R, Maxion-Bergemann S. (2005). Pharmacoeconomics. 23: 243–257. Ruoif J, Hu¨lsemann JL, Mittendorf T, Handelmann S, von der Schulenburg JM, Zeidler H, Merkesdal S. (2003). Ann Rheum Dis. 62:544–549. Sreshthaputra N, Kaemthong I. (2001). Asian-Pacific Health Economics Network (APHEN), access date April 2, 2008. Symmons DP, Jones MA, Scott DL, Prior P. (1998). J Rheumatol. 25:1072–1077. Tanno M, Nakamura I, Ito K, Tanaka H, Ohta H, Kobayashi M, Tachihara A, Nagashima M, Yoshino S, Nakajima A. (2006). Mod Rheumatol. 16:77–84. Torrance G. (1986). J Health Econ. 5:1–30. Veerapen K, Wigley RD, Valkenburg H. (2007). J Rheumatol. 34:207–213. Wailoo AJ, Bansback N, Brennan A, Michaud K, Nixon RM, Wolfe F. (2008). Arthritis Rheum. 58:939–946. Wolfe F, Mitchell DM, Sibley JT, Fries JF, Bloch DA, Williams CA, Spitz PW, Haga M, Kleinheksel SM, Cathey MA. (1994). Arthritis Rheum. 37:481–494. Wong JB, Singh G, Kavanaugh A. (2002). Am J Med. 113:400–408. Yelin E. (1996). J Rheumatol. 23 (supplement 44):47–51. Yelin E, Trupin L, Wong B, Rush S. (2002). J Rheumatol. 29:1851–1857. Zeng QY, Chen R, Darmawan J, Xiao ZY, Chen SB, Wigley R, Le Chen S, Zhang NZ. (2008). Arthritis Res Ther. 10:R17.
Part 3
3
Quality of Life Measures and Indices 3.1 General Aspects, Pathologies and Metabolic Disorders
102 Quality of Life-Related Concepts: Theoretical and Practical Issues A. A. J. Wismeijer . A. J. J. M. Vingerhoets . J. De Vries 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1754
2
Health Status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1754
3 3.1 3.2 3.3
Other QOL-Related Concepts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1755 Illness-Related Stressors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1756 Illness Impact . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1757 Suffering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1760
4
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1765 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1765
#
Springer Science+Business Media LLC 2010 (USA)
1754
102
Quality of Life-Related Concepts: Theoretical and Practical Issues
Abstract: Quality of Life (QOL) is a complex and multidimensional concept, which is related to a host of health concepts. In addition, there are several other concepts which seem to be related to quality of life. The focus of this chapter is to present an overview of, and introduction into QOL-related concepts, and to discuss how these concepts are measured and used in clinical practice. This concerns the following constructs: health status, illness related stressors, illness intrusiveness, impact of disease and suffering. List of Abbreviations: HS, health status; QOL, quality of life; WHO, world health organization; WHOQOL, World Health Organization Quality of Life Questionnaire
1
Introduction
Quality of life (QOL) is an increasingly important theme both in scientific research and patient care. This increasing popularity can be explained by the broad recognition that QOL is a highly relevant outcome measure in medical treatment, forming an important addition to the traditional biomedical endpoints such as limitations in the functioning of organs, senses, or limbs or mortality (De Vries, 2001). The fact that such biomedical parameters are often only weakly associated with the subjective well-being of the patient emphasizes its additional value. QOL refers to (dis)satisfaction with various aspects of life. The WHOQOL group (1994, 1995a, 1998a) has defined QOL as someone’s perception of his/her position in life in relation to his/her goals, expectations, standards, values and cares. Two aspects of this definition are particularly important. First, it shows that QOL is a subjective concept that refers to positive as well as negative aspects of life, and second, QOL appears to be a broad and multidimensional concept. Health care professionals thus should pay special attention to not only the influence of the disease or handicap on the everyday functioning of their patients (= health status (HS)), but also to their patients’ satisfaction with their physical, psychological, and social functioning (= QOL) (De Vries, 2001; De Vries and Drent, 2007). The link between QOL and HS is clear, but there are also other concepts which very likely are associated with QOL, but which nevertheless concern distinct constructs. What these concepts have in common with QOL is that they relate to the impact of disease: To what extent is the daily life of a patient affected by his or her disease and what are the consequences for the patient? Below we will discuss some of these concepts.
2
Health Status
Health status (HS) is a first related concept, with notable and distinctive characteristics. The definitions of the QOL and HS concepts stress that both are multidimensional. In line with the definition of health by the WHO (1958), QOL studies and questionnaires focus on the physical, mental and social domain. These domains usually include a number of aspects. For example, the physical domain includes pain; the mental domain includes cognitive function and self-image, whereas social support belongs to the social domain. In addition, particularly in clinical trials the focus is often on illness and treatment-related symptoms. The multidimensionality of both concepts is important because the diversity of experiences can not be captured with a questionnaire that assesses only one dimension, for example,
Quality of Life-Related Concepts: Theoretical and Practical Issues
102
the physical dimension (Fitzpatrick et al., 1992). The fact that a total score can be computed for a number of HS questionnaires is not compatible with the proposed multidimensionality. Such a total score only contains very limited information and is not easily interpretable: a given total score on a multidimensional questionnaire can be the result of the sum of very different subscores (De Vries and Drent, 2007). For example, one patient may have problems mainly at the social level; another suffers mainly from mental problems, whereas the third patient has predominantly physical limitations. Nevertheless, their total scores may be similar. Unfortunately, the concepts HS (also known as health-QOL) and QOL are often used interchangeably, as synonyms, which may easily induce confusion. Many studies suggest in their titles that they assess QOL, while actually HS has been measured. More precisely, for HS, the main issue is to what extent one’s functioning is limited. For example, can someone still walk, dress him/herself, go shopping, or maintain social contacts. QOL, in contrast, focuses on how (dis)satisfied the individual is with what s/he is still able to do. Research has clearly demonstrated that physical limitations or malfunctioning not always imply a poor QOL. Individual expectations about health, ambitions that can no longer be achieved, the (in)ability to cope with restrictions, the tolerance for discomfort and self-efficacy regarding disease are important determinants of one’s QOL. For instance, two people with similar limitations in functioning (HS) may evaluate their QOL very differently. A low score on HS can go along with a high score on a corresponding QOL domain and vice versa (De Vries, 2001; De Vries and Drent, 2006). Take as an example the impact of a problem with a knee for a middle-aged office manager and for a 22-year-old talented soccer player. Another difference between the two concepts is that QOL addresses both positive and negative aspects of one’s functioning. This is also reflected in the items included in QOL questionnaires: not only do they address one’s limitations and how one feels about those limitations, but they also address what one still can do and to what extent one is satisfied. This is a contrast with HS questionnaires that mainly focus on the limitations of the respondent, which can easily lead to a negative response tendency. It is often stated that HS questionnaires are in fact also a measure of QOL because they are completed by the patients themselves, and thus are subjective. However, there are two meanings of the word subjective in this context. First, it means that QOL and HS measures both are filled out by patients. However, the second meaning refers to the fact that patients report their own evaluation of their health status, which is a major difference with HS questionnaires. This distinction between HS and QOL is important because they can lead to different results and recommendations. Two frequently applied health status questionnaires are the Sickness Impact Profile (SIP; Jacobs et al., 1990) and the Medical Outcomes Study Short Form – 36 (SF-36; Ware et al., 1993). From the Medical Outcomes Study several disease-specific HS questionnaires have been derived, such as for hypertension, diabetes, thyroid disease, etc (e.g., Brooks et al., 1982).
3
Other QOL-Related Concepts
In addition to HS, there are some more health psychology concepts that seem to be in some way associated to QOL and therefore deserve adequate attention. For example, the concept of illness-related stressors, the consequences of illness and disability for one’s life goals (‘‘illness intrusiveness’’), the impact the disease has on several aspects of life, disease burden and suffering all seem to have some conceptual overlap with QOL. For all these concepts,
1755
1756
102
Quality of Life-Related Concepts: Theoretical and Practical Issues
assessment methods have recently become available. Below we will briefly introduce some of these concepts.
3.1
Illness-Related Stressors
The term stressor refers to a stimulus, situation, event or condition that has the potential to evoke stress reactions. Many of the stressors one is exposed to concern random events that just happen; its occurrence is not related to the functioning of the individual. Being confronted with them is merely a matter of bad luck, such as, for example, being at the wrong place at the wrong moment. However, stressors also may be rather closely associated to one’s personality, functioning and health status (Vingerhoets et al., 1989). People create to a large extent, aware or unaware, their own environment by avoiding certain situations and actively searching for others. A simple illustration of this idea is that individuals will not be hospitalized and undergo painful medical procedures, unless they are suffering from a serious health problem. That is, the exposure to these stressors depends fully on one’s health status (Prugh and Thompson, 1990; Schechter and Leigh, 1990). Also other stressors may be related to one’s functioning and/or health status, such as the loss of one’s job, having financial problems, loss of the capacity to engage in certain hobbies, being forced to move, etc. This is nicely demonstrated by Blokhorst et al. (2002), who found that whiplash patients, compared to healthy controls, obtained equal scores on the so-called person-independent stressors, but scored significantly higher on the person-dependent stressors of the Everyday Problems Checklist (EPCL; Vingerhoets and Van Tilburg, 1994). Interestingly, the patients rated the impact of both categories of stressors higher than the healthy controls, which may be interpreted as an indication that the patient group is more vulnerable to stress. The EPCL contains 114 items representing everyday events that have the potential to provoke a variety of negative emotions, including fear, anger, disappointment, guilt, regret and embarrassment. It has adequate psychometric features and a unique characteristic concerns the above described two subscales with, respectively, person-dependent and person-independent items. These scales were based on the ratings of each item by behavioral scientists and clinicians how likely it is that the occurred event described by the item can be attributed to the person. As support for the validity of this distinction, it was found that individuals scoring high on neuroticism also especially score higher on person-dependent items. Example items of this checklist are shown in > Table 102-1. Especially the patients’ score on the person-dependent items may thus be interpreted as an indication of the impact of the disease. A move into a nursery home or a hospitalization, in particular when it implies that one has to undergo painful medical procedures, is generally experienced as rather stressful. Some studies have specifically focused on the assessment of stressors in the medical context (e.g., Koenig et al., 1995). This research revealed that the interaction with health professionals and especially the lack of information stand out as sources of stress. In addition, the hospital environment (the noise, rigid routines, lack of privacy), worrying about the home situation, homesickness, being stigmatized and discriminated by health care providers or fellow patients, and the fear or losing one’s independency and autonomy as well as loss of control are chief determinants of patient stress. Finally, as said before, painful procedures and interventions that may be a threat to one’s physical or psychological integrity may contribute to the stress experienced by hospitalized patients. Specific questionnaires have been designed to evaluate
Quality of Life-Related Concepts: Theoretical and Practical Issues
102
. Table 102-1 Examples of items from the Everyday Problems Checklist (EPCL; Vingerhoets and van Tilburg, 1994) Person-dependent items 1. Important possessions were lost 6. You wanted things you were not able to afford financially 78. You failed to accomplish tasks that you thought you were capable of doing 104. You could not be yourself 110. You unintentionally insulted someone Person-independent items 8. Your sleep was disturbed 16. You were unemployed or temporarily laid off 62. People around you behaved irresponsibly 74. Your favorite team suffered defeat and/or humiliation 105. You witnessed a traffic accident or criminal offense
the impact of having to undergo stressful medical procedures. In addition, there is ample attention for parents and family members of children with serious health conditions that are hospitalized in intensive care units (Board and Ryan-Wenger, 2003; Spear et al., 2002). To summarize, having to live with a disease not only implies the experience of symptoms, but not exceptionally it has major consequences such as hospitalization, undergoing intensive medical treatments, as well as influence on work and relationships.
3.2
Illness Impact
Disease and disability can strongly interfere with one’s life goals, and hence affect well-being. Future plans need to be drastically revised, because they can not be realized. Carefully planned careers are thwarted and dramatic adaptations are required. Some researchers focus especially on the relationship between the attainment of life goals and well-being (e.g., Schmuck and Sheldon, 2001). Their research reveals that a global distinction can be made between individuals whose main aim is to avoid negative situations, whereas for others the focus is more on seeking pleasure. Until now, little is known about the effects of disease on the well-being of these two groups of people. Illness Intrusiveness is assessed with the Illness Intrusiveness Rating Scale (IIRS; Devins et al., 2001), consisting of 13 items. Implicit in this concept is that the disruption of lifestyles and activities attributable to constraints imposed by chronic disease and its treatment has a major impact on one’s well-being. The respondents have to indicate to what extent their disease has an impact on 13 dimensions of their life, including relationships, work, trust in one’s body, etc. However, this measure only asks for the extent of the impact; there is no specification whether the impact is considered as negative or positive. Generally, negative associations are reported with QOL, whereas there is a positive connection with depression (Schimmer et al., 2001). What we consider a major disadvantage of this instrument is that the respondent thus has no possibility to indicate that his/her health problem also may have a positive influence on one’s life.
1757
1758
102
Quality of Life-Related Concepts: Theoretical and Practical Issues
Inspired by the IIRS and being aware of the above limitation, we have recently developed the Perceived Disease Impact Scale (PDIS; Mols et al., 2007; van Gestel et al., 2007) which differs from the IIRS in two important aspects. First, we added eight items, resulting in a total of 21 items. > Table 102-2 shows some examples of the items. And, second, the respondents had to indicate
. Table 102-2 Examples of items from the Perceived Disease Impact Scale (PDIS; Mols et al., 2007) To what degree did your disease and/or the treatment affect your current: 1. Physical health 9. Financial situation 15. Personal development 17. Sense of involvement to what happens in the world 20. Confidence in your body
whether they perceived the impact of their disease on the listed aspects of life as neutral, positive or negative. This proved to be very important because, in particular with long term cancer survivors, but also among women with fertility problems, it appeared that the effects of the disease were not just negative. That is, patients also indicated positive effects of their disease on several aspects of their life. While the effects of (breast) cancer are almost always negative in the immediate aftermath of diagnosis, years later these effects apparently are not only negative. Whereas approximately 30% of breast cancer survivors still experience specific complaints, there is also evidence that many long-term survivors experience a good quality of life 5 years or more after diagnosis. Some recent studies have provided evidence that cancer patients engage in what is referred to as ‘‘benefit finding.’’ Well-known examples of these benefits are a greater appreciation of life and a change in life priorities. In addition, the literature also suggests that after cancer, patients may experience posttraumatic growth. Posttraumatic growth refers to the success with which individuals, coping with the aftermath of trauma, reconstruct or strengthen their perceptions of self, others, and the meaning of events (see > Figure 102-1). Whereas benefitfinding may start immediately after diagnosis, research suggests that posttraumatic growth specifically develops first after a process of rumination and restructuring that occurs in the weeks, months, and even years following the trauma. A striking illustration of these processes is provided by multiple Tour de France-winner Lance Armstrong: "
‘‘When I was 25, I got testicular cancer and nearly died,’’ writes Armstrong in his 2001 memoir It’s Not About the Bike: My Journey Back to Life. ‘‘I was given less than a 40% chance of surviving, and frankly, some of my doctors were just being kind when they gave me those odds.’’ (. . .) ‘‘There are two Lance Armstrongs, pre-cancer and post. Everybody’s favourite question is ‘‘How did cancer change you?’’ The real question is how didn’t it change me? I left my house on October 2, 1996 as one person and came home another. . . . The truth is that cancer was the best thing that ever happened to me. I don’t know why I got the illness, but it did wonders for me, and I wouldn’t want to walk away from it. Why would I want to change, even for a day, the most important and shaping event of my life?’’
Until now, (long-term) adaptation to (breast) cancer has been measured traditionally by means of questionnaires about well-being or (health-related) QOL. However, such measures
Quality of Life-Related Concepts: Theoretical and Practical Issues
102
. Figure 102-1 Schematic representation of a model with health complaints, suffering and posttraumatic growth
typically fail to offer any insight into precisely which specific domains of life are (no longer) positively or negatively affected. These measures also lack the capacity to assess the positive effects of cancer and/or its treatment on the different life-domains. For this reason, we decided to determine which life domains were positively or negatively affected in long-term breast cancer survivors. We also wanted to know which patient or tumor characteristic, in terms of age, stage and treatment, was associated with the least negative or most positive effects of cancer and its treatment on their lives. In > Table 102-3 we summarize the findings obtained in different patient groups. The results are striking in at least two ways. First, there is the remarkable correspondence among different groups of cancer patients/survivors. Second, compared to cancer survivors other patient groups apparently report far less positive effects of their disease. However, what is most important, is that researchers and clinicians as well should be careful to infer conclusions about the impact of a disease. Prejudices may easily lead to wrong perceptions, implying that the patients do not receive the care and attention that they really need. An interesting example in this respect is hair loss in cancer patients after being treated with chemotherapy. Mulders et al. (2008) showed that oncology nurses and physicians underestimated the relevance of this side effect of chemotherapy. Using a specially designed so called psychophysical scaling method, these investigators compared the perceived impact of several cancer and cancer treatment related effects. Whereas there was a close correspondence between nurses and physicians, both groups grossly overestimated and underestimated various issues. For example, the effects on relationships with partners and children were greatly overestimated
1759
1760
102
Quality of Life-Related Concepts: Theoretical and Practical Issues
. Table 102-3 Reported amount of negative, neutral and positive consequences of disease, for different patient groups # of negative consequences
# of neutral consequences
# of positive consequences
Breast cancer (2–3 years; DCIS)
2.9
10.1
6.1
Breast cancer (2–3 years; invasive)
4.7
8.1
6.4
Breast cancer 5–10 years
2.3
7.1
8.3
Patient group
Prostate cancer 5–10 years
2.7
7.4
9.7
Morbid obesitas
7.8
7.7
4.5
5.0
8.5
6.1
10.9
5.7
3.3
Infertility (women) Addiction
by nurses, while physicians underestimated hair loss the most. The authors concluded that this observed lack of correspondence between patients and health-care providers may result in inappropriate provision of attention and health care. They argue that methods have to be developed to assess easily the main needs and worries of individual patients, which is an essential condition to be able to provide optimal care. In conclusion, there is little doubt that chronic and life threatening disease may have a major impact on the patient. However, not all aspects of life are necessarily negatively influenced, and there may be great individual differences in what is perceived as most threatening and stressful. Health care providers cannot assume that every patient experiences his/her disease in the same way and that all patients can be treated equally. In order to provide optimal care, careful questioning the patient either face-to-face or with a structured questionnaire is needed in order to pay adequate attention to the specific concerns and worries of that individual patient.
3.3
Suffering
The alleviation of suffering is a chief objective of medicine, especially in the care of terminal patients. However, medicine also seem to apply interventions (e.g., chemotherapy) which increase suffering, rather than reducing it. Suffering can only be treated if it can be recognized and diagnosed. It involves not only mere symptoms or the process that threatens the patient, but also is related to issues like the perceived meaning of the symptoms and concerns about the future. The meanings and fears may be very personal and individual, so that even if two patients have the same symptoms, their suffering is likely to be very different. Cassell (1982) wrote a seminal theoretical contribution about the nature of human suffering. The essence of his description is that suffering is ‘‘experienced by persons, not merely by bodies, and has its source in challenges that threaten the intactness of the person as a complex social and psychological entity’’ (Cassell, 1982, p. 639). Cassell also asserted that suffering may include pain but is not limited to it. His comparisons of pain versus suffering
Quality of Life-Related Concepts: Theoretical and Practical Issues
102
and his exploration of the concept meaning emphasize the importance of taking into account the whole person, when attempting to diagnose and operationalize suffering. This author also feels that because nursing and medicine have become highly technical and often quite depersonalized, there is hardly adequate attention for this aspect. The challenge therefore also is to design a valid method to assess suffering, taking into account the personal meaning an individual gives to the threats to his or her ‘‘personhood’’ and recognizing this aspect is critical to adequately understanding human illness and suffering. This personhood can be seen as a complex intermingling set of person features like personality and character, past life experiences, social environment, and cultural background. Consequently, the extent to which a disease affects the individual depends on this melting pot of person and environmental features. Therefore, it may not come as a surprise that there is a lot of individual variability in the degree of suffering given a certain specified objective disease severity. To illustrate this, Cassell refers to a young woman with breast cancer: ‘‘This young woman had severe pain and other physical symptoms that caused her suffering. But she also suffered from some threats that were social and from others that were personal and private. She suffered from the effects of the disease and its treatment on her appearance and abilities. She also suffered unremittingly from her perception of the future’’ (Cassell, 1982). Being aware of the relevance of suffering for medicine is one thing, the development of valid ways to diagnose or assess it is another. As Kleinman (1982) wrote: ‘‘Clinical and behavioral science research also possess no category to describe suffering. Symptom scales, survey questionnaires and behavioral checklists quantify functional impairment and disability, yet about suffering they are silent.’’ An interesting and possibly relevant model in this respect is the so called enmeshment model introduced by Pincus and Morley (2001). According to these authors, individuals have cognitive representations (schemata) of their self, their illness and their pain. They theorized that the experience of chronic pain related to the degree to which the three schemas of pain, self and illness over-lapped, leading, ultimately, to a form of enmeshment where the activation of elements from one would influence the other. A pathway, via the self, was hypothesized between the physical and psychological dimensions of chronic pain and it was the degree to which chronic pain disrupted the aspects of the person’s schema of the self that determined the focus and degree of enmeshment: ‘‘the degree to which the chronically activated pain schema ‘traps’ negative aspects of the self. As a consequence, the pain experience is viewed in terms of its behavior and affective implications for the self and not just its sensory characteristics’’ (Pincus and Morley, 2001). To put it differently, ‘‘healthy’’ adjustment or adaptation involves separation of the schema for Self, Illness and Pain, whereas distress arises when the schemata for Pain and Illness become enmeshed with the Self-Schema (see > Figure 102-2). The graphical representation of the enmeshed model inspired Bu¨chi and Sensky (1999) to design a simple visual method to assess coping and adjustment to disease: the Pictorial Representation of Illness and Self Measure (PRISM; Bu¨chi et al., 2002; Bu¨chi and Sensky, 1999). The PRISM was originally believed to assess coping. However, pilot qualitative research demonstrated that this instrument was probably measuring suffering, which is something more complex than just coping. This measure breaks with tradition in that it does not contain any questions, but rather consists of a rectangular (A4 size) metal board, with a fixed yellow circle (7 cm in diameter) in the bottom right-hand corner. Patients are asked to imagine that the white board represents their current life and the yellow circle their ‘‘self.’’ They subsequently receive a magnetic red
1761
1762
102
Quality of Life-Related Concepts: Theoretical and Practical Issues
. Figure 102-2 Schematic example of normal enmeshment versus distressed enmeshment (based on Pincus and Morley, 2001)
disk (5 cm in diameter), which represents their illness (illness-disk), and are asked to place the illness-disk on the board to represent the place of the illness in their current life (see the upper part of > Figure 102-3). The patient receives oral standard instructions explaining the task. The PRISM is hypothesized to produce a graphical summary of the relationships between illness, self, and ‘‘life at the moment’’ which cannot be reduced to a single dimension. This summary is difficult to capture with questions because each individual will give different weights to a wide variety of factors which determine suffering. The quantitative measure derived from this application, the distance between the centers of both disks, is referred to as Self Illness Separation (SIS). It is assumed that patients have cognitive representations of their ‘‘self ’’ and their illness and that healthy adjustment to the disease implies that there is a separation between the ‘‘self ’’ and the illness schema. Comments of patients completing the task revealed that the SIS measure is associated predominantly with patients’ perception of the intrusiveness of the illness, its controllability, and the interference of the illness with salient aspects of everyday life. > Table 102-4 shows some examples of patients’ comments patients corresponding to either low or high SIS. In essence, PRISM is a visual representation of the relationship between the person’s ‘‘self ’’ and his/her illness. Recently, Wouters et al. (2008) have developed two modifications of the original PRISM task, the PRISM-R1 and PRISM-R2, because patients commented that the ‘‘size’’ of their disease could be perceived very differently. In addition, it was noticed that patients occasionally considered the middle of the A4 sheet as the central point in their life, rather than the self-disk. The major change of the first modification (PRISM-R1) therefore involved giving patients a choice of three different sized illness-disks (see the middle part of > Figure 102-3). This modification resulted in an additional quantitative variable, referred to as Illness Perception Measure (IPM). The second modification (PRISM-R2) involved placing the self-disk in the middle of a large printed circle, rather than in one corner of a rectangular sheet as in the original PRISM to symbolize the centrality of the self-disk (see the lower part of > Figure 102-3). In three separate studies, the feasibility and psychometric qualities of the two revised versions of the PRISM was explored (Wouters et al., in press). Two studies were carried out with PRISM-R1. In the first one, the potential use of the PRISM-R1 as a generic measure for suffering by comparing results of five different patient groups was explored, and its validity
Quality of Life-Related Concepts: Theoretical and Practical Issues
102
. Figure 102-3 Schematic example of the PRISM, PRISM-R1 and PRISM-R2. Reprinted with the kind permission of Heldref Publications, publisher of Behavioral Medicine
was examined. It was found that whiplash patients and women with fertility problems indicated higher suffering than lung, cancer, and psoriasis patients. In the second study the sensitivity to change of the PRISM-R1 was tested by comparing pre- and post treatment data of a group of whiplash patients participating in a multidisciplinary intervention program. As expected, after the intervention significant changes were detected, suggesting less suffering. PRISM-R2 was evaluated in a third study involving the collection of additional qualitative and quantitative data among morbidly obese patients seeking bariatric surgery. This latter study was designed to investigate the content validity and the convergent and divergent construct
1763
1764
102
Quality of Life-Related Concepts: Theoretical and Practical Issues
. Table 102-4 Examples of patient comments on their choice for positioning the ‘‘self’’ disk (SIS) Themes
Low SIS statements: SIS 52 mm overlap with ‘‘self’’ no overlap with ‘‘self’’
Impact of the patient’s medical problem on daily life
My obesity interferes with everything: movement, work, mood (SIS 0 mm)
Impact of the medical problem on health status
For me it is very important to lose My medical problem doesn’t cause weight because of my back and me psychological health problems, feet complaints (SIS 0 mm) I am not ashamed of my appearance (SIS = 59 mm)
Attributed origin of the medical problem (self, other/something else)
I am the problem myself (SIS = 2 mm) Overweight is a problem beyond my power (SIS 29 mm) It is part of myself and part of others (harassments) (SIS 47 mm)
My obesity is something close to me. I carry it around all day. However, it doesn’t control my life (SIS = 64 mm)
validity of the SIS and IPM. SIS and IPM showed overlap, but also tapped specific, unique aspects of the perceived burden of disease. Further research is needed to unravel the specific elements addressed by both variables, and currently a study is being designed that focuses on the cognitive processes that respondents apply when making their choices concerning the disk size and the location. That information might contribute importantly to our understanding of this intriguing measure. This measure is currently applied in a large multi-center study among diabetes patients. A pilot study learned that this measure was met with much enthusiasm by the involved diabetes nurses. All patients completed a computerized version of the PRISM-R2 and the results were used to start a conversation with the patient, asking him/her to explain why s/he choose for that specific size of the illness disk and why it was attached at that specific location. The nurses also asked what the ideal situation would be and what was perceived as main barriers to reach that ideal situation. Such information might not only very helpful to identify those patients that need extra support and attention, it will also contribute to a better understanding of the interpretation of this measure. A final interesting theoretical issue concerns the precise the relationship between suffering and QOL. It is probably too simple to say that suffering is merely a very low QOL. In line with the ideas formulated by Cassell (1982, 1991), we come to the following preliminary hypothesis: in case of severe physical problems and limitations, the QOL will be determined to a large extent by being able to give meaning and to perceive benefits of one’s condition. If a patient fails to meet these demands necessary for successful adaptation to the disease, a very low QOL, to be labeled as suffering is most likely the result. In contrast, when, in the same condition, the patient is able (and many patients appear to have this capacity!) to give meaning and to find benefits, the condition is set for a rather good QOL and maybe even personal growth. Further research is needed to investigate the validity of this model.
Quality of Life-Related Concepts: Theoretical and Practical Issues
4
102
Conclusion
There can be little doubt that the QOL concept takes a central and increasingly important position in health care. This is evidenced, among others, by the strong increase in the number of publications on this issue. QOL is a main outcome when evaluating health care interventions. However, the term QOL is often mistakenly used interchangeably with related terms such as health status. In addition, for a better understanding of the dynamic processes that determine one’s QOL, it seems useful to examine concepts such as illness-related stressors, illness impact and suffering. We also want to plea for assessment of QOL and/or the here discussed related concepts, in order to prevent health care providers from not fitting the care to be provided to the needs of the patients.
Summary Points The concepts HS (also known as health-QOL) and QOL are often used interchangeably which may easily induce confusion.
Physical limitations or malfunctioning not always imply a poor QOL. In particular long term cancer survivors also report many positive effects of their disease on several aspects of their life.
In order to understand the dynamics of the factors contributing to QOL, more research is needed with concepts like illness related stressors, illness intrusiveness or disease impact, and suffering. Failure to assess adequately the QOL and/or related concepts in patients may result in the provision of inadequate and not needed care, at the cost of attention for issues that seriously bother the patients.
References Blokhorst MGBG, Lousberg R, Vingerhoets AJJM, Winter FAM, Zilvold G. (2002). Int J Rehab Res. 25: 1–8. Board R, Ryan-Wenger N. (2003). J Pediatr Nurse. 3: 195–202. Brooks RH, Lohr KN, Goldberg GA. (1982). Conceptualisation and Measurement of Health for Adults, vol. 9: Thyroid Disease. Rand Corporation, Santa Monica, CA. Bu¨chi S, Sensky T. (1999). Psychosom. 40: 314–320. Bu¨chi S, Buddeberg C, Klaghofer R, Russi EW, Brandli O, Schlosser C, Stoll T, Villiger PM, Sensky T. (2002). Psychother Psychosom. 71: 333–341. Cassell EJ. (1982). N Engl J Med. 18: 639–645. Cassell EJ. (1991). The Nature of Suffering and the Goals of Medicine. Oxford University Press, New York. Devins GM, Dion R, Pelletier LG, Shapiro CM, Abbey S, Raiz LR, Binik YM, McGowan P, Kutner NG,
Beanlands H, Edworthy SM. (2001). Med Care. 10: 1097–1104. De Vries J. (2001). In: Vingerhoets AJJM (ed.) Assessment in Behavioural Medicine. Brunner-Routledge, Hove, UK, pp. 353–370. De Vries J, Drent M. (2007). Semin Respir Crit Care Med. 28: 121–127. De Vries J, Drent M. (2006). In Baughman RP (ed.) Sarcoidosis Taylor & Francis, New York, pp. 463–478. Fitzpatrick R, Fletcher A, Gore S, Jones D, Spiegelhalter D, Cox D. (1992). Br Med J. 305: 1074–1077. Jacobs HM, Luttik A, Touw-Otten FWMM, De Melker RA. (1990). Dutch J Health Care. 134: 1950–1954. Kleinman A. (1982). Cult Med Psychiatry. 6: 117–190. Koenig HG, George LK, Stangl D, Tweed DL. (1995). Int J Psychiat Med. 25: 103–122. Mols F, Vingerhoets AJJM, Coebergh JWW, van de PollFranse LV. (2008). Psychol Health. 1–14.
1765
1766
102
Quality of Life-Related Concepts: Theoretical and Practical Issues
Mulders M, Vingerhoets AJJM, Breed W. (2008). Eur J Oncol Nurs. 12: 97–102. Pincus T, Morley S. (2001). Psych Bull. 127: 599–617. Prugh DG, Thompson TL II. (1990). In: Noshpitz JD, Coddington RD (eds.) Stressors and the Adjustment Disorders. Wiley, New York, pp. 60–142. Schechter JO, Leigh, L. (1990). In: Noshpitz JD, Coddington RD (eds.) Stressors and the Adjustment Disorders. Wiley, New York, pp. 143–159. Schimmer AD, Elliott ME, Abbey SE, Raiz L, Keating A, Beanlands HJ, McCay E, Messner HA, Lipton JH, Devins GM. (2001). Cancer. 92: 3147–3154. Schmuck P, Sheldon KM. (2001). Life Goals and WellBeing: Towards a Positive Psychology of Human Striving. Hogrefe and Huber, Kirkland, WA. Spear ML, Leef K, Epps S, Locke R. (2002). Am J Perinatol. 19: 205–213. Van Gestel YRBM, Voogd AC, Vingerhoets AJJM, Mols F, Nieuwenhuijzen GAP, Repelaer van Driel OJ,
van Berlo CLH, van de Poll-Franse LV. (2007). Eur J Cancer. 43: 549–556. Vingerhoets AJJM, Jeninga AJ, Menges LJ. (1989). Behav Healthc. 17: 10–17. Vingerhoets AJJM, Van Tilburg MAL. (1994). Alledaagse Problemen Lijst – Manual. Swets & Zeitlinger, Lisse, The Netherlands. Ware JE Jr, Snow KK, Gandek B. (1993). SF-36 Health Survey. Manual and Interpretation Guide. The Health Institute, New England Medical Center, Boston MA. WHO. (1958). The First Ten years of the World Health Organisation. WHO, Geneva, Switzerland. WHOQOL group. (1994). Int J Ment Health. 23: 24–56. WHOQOL group. (1995a). Soc Sc Med. 41: 1403–1409. WHOQOL group. (1998a). Soc Sci Med. 46: 1569–1585. Wouters EJM, Reimus JLM, Van Nunen AMA, Blokhorst MGBG, Vingerhoets AJJM. (2008). J Behav Med. 34: 65–76.
103 Alternative Therapies and Quality of Life J. X. Zhang 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1768 2 The Challenges in Research on CAM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1769 3 CAM Modalities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1771 4 Prevalence of CAM Use in Cancer Treatments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1771 5 CAM Use and QOL in Patients with Breast Cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1772 6 CAM Use and QOL in Patients with Musculoskeletal Disease . . . . . . . . . . . . . . . . . . . . 1774 7 Effect of CAM Use on QOL in Treatment of Patients with Other Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1776 8 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1777 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1779
#
Springer Science+Business Media LLC 2010 (USA)
1768
103
Alternative Therapies and Quality of Life
Abstract: Nearly half of adults in the United States use > complementary and alternative medicine (CAM) every year, and Americans are visiting CAM practitioners more frequently than primary care physicians. Worldwide, the prevalence rates of CAM use in adults ranged from 9 to 76%. Despite the significant growth in CAM use in the past decade, there exists a lack of consistency in the body of CAM literature as to precisely what CAM is. There is also a paucity of evidence on the efficacy of CAM. The research on the efficacy of CAM use is faced with many challenges that are different from those associated with conventional medicine. Historically, the common drawbacks in CAM research include poor study design, small sample size, and inadequate controls. In recent years, high-quality research articles have begun to emerge in leading medical journals. However, the evidence on the efficacy of CAM use on quality of life (QOL) is not convincing. Even for those areas of CAM which have been studied with several relatively large trials, it is unclear whether the observed effect is due to true CAM intervention or placebo effect. List of Abbreviations: CAM, complementary and alternative medicine; CES-D, Center for Epidemiologic Depression Scale; CHF, congestive heart failure; HADS, Hospital Anxiety and Depression Scale; IBS, irritable bowel syndrome; NCCAM, National Center for Complementary and Alternative Medicine; OA, osteoarthritis; PSS, perceived stress scale; PST, periosteal stimulation therapy; RCTs, randomized controlled trials; SF-36, 36-item short-form health survey; SPPB, short physical performance battery; TCA, traditional Chinese acupuncture; WOMAC, Western Ontario and McMaster Universities Osteoarthritis Index
1
Introduction
The proportion of the general population in the United States that reports using complementary and alternative medicine (CAM) has increased from 33.8% in 1990 to 42.1% by 1997 (Eisenberg et al., 1998), and the National Health Interview Survey indicated a continued growth of CAM use to 62.1% in 2002 (Barnes et al., 2004). The increased use of CAM was the result of a secular trend across all major socio-demographic sectors in the general population (Kessler et al., 2001). Corroborating evidence also suggested that nearly half of adults in the United States use CAM every year, and that Americans are visiting CAM practitioners more frequently than primary care physicians (Kelly et al., 2005; Nahin et al., 2005; Ruggie, 2005; Kelly et al., 2005, and Scharf et al., 2006). The growth in CAM use has been significant in the US in the past decade. Worldwide, a systematic review found that the prevalence rates of CAM use in general population ranged from 9 to 65% in developed economies with German having the highest prevalence rate, based upon various definitions of CAM (Ernst, 2000). A more recent study in Singapore, a city-state in south-eastern Asia, reported an even higher 76% prevalence rate (Lim et al., 2005). Patients seek CAM for a variety of reasons including treating their cancer, relieving symptoms and boosting their immune system, as well as to gain hope and a sense of control over their health condition (Astin 1998; Richardson et al., 2004; Downer et al., 1994; Richardson 2004; Verhoef et al., 2005). In addition to CAM use in adults, CAM has also been used with high prevalence in children with chronic illness, such as carcinoma (42%), juvenile rheumatoid arthritis (70%), and inflammatory bowel disease (72%), and in general adolescent populations cared for through hospital inpatient and outpatient services (41%). (Crawford et al., 2006). There are substantial interests from various stakeholders in the effects of CAM on patient outcomes such as quality of life (QOL) measures. However, the purpose of this chapter
Alternative Therapies and Quality of Life
103
is not to suggest whether patients should or should not take certain > CAM modalities for certain conditions; rather, it is to weigh the evidence revealed by the recent published research and determine whether or not there exists corroborating evidence suggesting a unified finding on the effect of CAM on QOL.
2
The Challenges in Research on CAM
While CAM use has been experiencing an substantial growth in recent years, there exists a lack of consistency in the body of CAM literature as to precisely what CAM is. The National Center for Complementary and Alternative Medicine (NCCAM), a branch of the National Institutes of Health in the US, has defined complementary and alternative medicine (CAM) as “a group of diverse medical and health care systems, practices, and products that are not presently considered to be part of conventional medicine” (NCCAM, 2002). Earlier studies had also defined CAM as “diagnosis, treatment, and/or prevention which complements mainstream medicine by contributing to a common whole, satisfying a demand not met by orthodox medicine, or diversifying the conceptual framework of medicine”(Ernst et al., 1995); other studies have defined CAM by its difference from what has been taught in medical schools or is generally available in hospitals, as well as whether it is generally accepted by the dominant group of medical practitioners in a society (> Table 103-1). . Table 103-1 Definition of CAM Author (year) NCCAM (2002)
Definition of CAM A group of diverse medical and health care systems, practices, and products that are not presently considered to be part of conventional medicine
Ernst et al. (1995) Diagnosis, treatment, and/or prevention which complements mainstream medicine by contributing to a common whole, satisfying a demand not met by orthodox, or diversifying the conceptual framework of medicine Eisenberg et al. (1993)
Interventions neither taught widely in medical schools nor generally available in hospitals
Gevitz (1988)
Practices that are not accepted as correct, proper, or appropriate or are not in conformity with the beliefs or standards of the dominant group of medical practitioners in a society
Source: Author’s adaptation from the respective publications
The diverse definition of CAM is a reflection of the fact that CAM has been defined not only for scientific reasons, but also for political and social reasons (Jonas, 2002). At the meantime, the curricula in CAM education in medical schools and pharmacy schools have progressed continuously and substantially in the United States in recent years. For example, an overwhelming majority of medical schools (98) in the US were offering courses in CAMrelated topics during the 2002–2003 academic year (Barzansky and Etzel, 2003), up from 45 of 125 schools in the 1997–1998 academic year (Wetzel et al., 2003). A recent survey also indicated that over 97% of pharmacy schools have taught herbal medicine. Other CAM modalities, such as homeopathy, Chinese herbal medicine, megavitamins, and acupuncture, were taught in over fifty percent of schools. Overall, 73% of pharmacy schools were offering
1769
1770
103
Alternative Therapies and Quality of Life
courses in CAM (Dutta et al., 2003). Thus, with the increasing acceptance of CAM in medical educational programs, the definition of CAM has to be broadened to include a wide spectrum of CAM modalities different from conventional medicine. Besides the lack of a consistent definition of CAM, there is also a paucity of cumulative evidence on the efficacy of CAM. This is perhaps partially attributable to the fact that CAM was historically not a dominant practice in contrast to conventional medicine, and sometimes was not even considered “correct” medicine (Gevitz, 1988). The lack of research infrastructure, such as a federal agency in the US to fund and support research of CAM, may have also contributed to the lack of research development in the efficacy of CAM use. This situation started to change after the creation of NCCAM in 1998, a branch of the National Institutes of Health. Since then the funding for CAM research has been steadily growing (> Figure 103-1). . Figure 103-1 NIH funding for CAM research. Source: Author’s adaptation based upon statistics reported in Alternative and Complementary Medicine in the United States by Institute of Medicine Committee on Complementary and Alternative Medicine, 2005. In $ million. *Estimated funding
In addition to the lack of consistent definition and funding, the research on the efficacy of CAM use is faced with many other challenges that are different from those associated with conventional medicine among which, one is the lack of a biomedical foundation to delineate the mechanisms of how the various CAM modalities work. Many CAM modalities, despite of being practised in various cultures for hundreds of years, have never been subject to the scrutiny of basic clinical science that the conventional medicine has been used to. There is also controversy surrounding whether or not the observed effect of CAM, if any, is a true effect as CAM claims or merely a placebo effect (de la Fuente-Fernandez et al., 2001, Moerman and Jonas, 2002). In addition, CAM practices often apply an individualized treatment approach which makes standardization and generalization difficult. Such difficulty is further compounded by the fact that the treatment may involve multiple chemicals (e.g., herbals), or CAM modalities which make the distinction of a single effect virtually impossible. Despite the tremendous challenges in CAM research, the body of literature on the efficacy of CAM published in peer-reviewed journals has been growing. Historically, the common drawbacks in CAM research include poor study design, small sample size, and inadequate controls. In recent years, high-quality research articles have begun to emerge in leading
Alternative Therapies and Quality of Life
103
medical journals. There is a strong interest in the efficacy of CAM in improving the quality of life (QOL) in areas where conventional medicine is inadequate, such as cancer treatment and the treatment of musculoskeletal disease. In cancer treatment, of particular interest is the effect of CAM used by patients with breast cancer; in musculoskeletal disease, new evidence from a series of high-quality studies using randomized trials has been published in leading medical journals in recent years. The discussion below will thus be focused on the evidence of the effect of CAM on QOL in these two major conditions.
3
CAM Modalities
CAM covers a wide range of different modalities from herbs to acupuncture. NCCAM includes five different modalities in its categorization of CAM, which include (1) alternative medical systems, (2) mind-body interventions, (3) biologically-based treatments, (4) > manipulative and body-based methods, and (5) energy therapies. Examples of these five modalities are listed in > Table 103-2. Alternative classifications of CAM modalities also exist. For example, one descriptive taxonomy categorized CAM into a more prominent, “mainstream” CAM and a more culture-bound, “parochial” unconventional medicine (Kaptchuk and Eisenberg, 2001). To maximize the consistency in this chapter, we adopted the classification system proposed by NCCAM. . Table 103-2 CAM modality classifications CAM modalities
Examples
Alternative medical systems
Homeopathic medicine, naturopathic medicine, traditional Chinese medicine, and Ayurveda
Mind-body Medicine
Patient support groups, cognitive-behavioral therapy, meditation, prayer, mental healing, and therapies that use creative outlets such as art, music, and dance
Biologically based practices
Herbs, foods, vitamins, dietary supplements, herbal products, and the use of other so-called natural but as yet scientifically unproven therapies, for example, using shark cartilage to treat cancer
Manipulative and bodybased method
Chiropractic or osteopathic manipulation, and message
Energy medicine
Biofield therapies, such as qi gong, Reiki, and therapeutic touch; and bioelectromagnetic-based therapies, such as pulsed fields, magnetic fields, or alternating-current or direct-current fields
Source: NCCAM “What are the major types of complementary and alternative medicine?” http://nccam.nih.gov/ health/whatiscam/ accessed on 15 February 15, 2008
4
Prevalence of CAM Use in Cancer Treatments
Studies suggest the prevalence of CAM use among cancer patients often exceeds that of the general population. In 2000, a study from MD Anderson Cancer Center quantified patient use of CAM at 68.7% – even after excluding spiritual counseling and psychotherapy (Richardson
1771
1772
103
Alternative Therapies and Quality of Life
et al., 2000) – and 88.2% of advanced cancer patients enrolled in phase I clinical trials at the Mayo Clinic reported using CAM (Dy et al., 2004). Among breast cancer patients, CAM use has been studied extensively and as much as 84% report CAM use (DiGianni et al., 2002). Similar rates have also been found within prostate cancer and head and neck cancer populations (Diefenbach et al., 2003; Kao and Devine, 2000; Warrick et al., 1999). While some of the differences in the prevalence of CAM use are due to different or wide- ranging definitions of CAM within specific studies, these researches clearly described CAM’s widespread and common use by significant and diverse populations of cancer patients.
5
CAM Use and QOL in Patients with Breast Cancer
CAM use is particularly popular in patients with breast cancer. A diverse set of CAM modalities has been used including dietary supplements, mind-body approaches, and acupuncture, and the reasons cited for using CAM were to boost the immune system, improve quality of life, prevent recurrence of cancer, provide control over life, and treat breast cancer and the side effects of treatment (Nahleh and Tabbara, 2003). The studies are often limited in design quality (Nahleh and Tabbara, 2003). More evidence from clinical trials has emerged in recent years, which has been mixed, showing both favorable and unfavorable results of CAM use in managing breast cancer treatment (> Table 103-3). In the area of breast biopsy, a randomized pilot study to determine the efficacy of Reiki in reducing psychological distress including anxiety and depression associated with breast biopsy using self-reported questionnaires for 32 women did not suggest evidence of efficacy. Besides the small sample size (N = 32), there existed selection bias in this trial as the baseline anxiety level was comparatively low (Potter, 2007). Another randomized, patient-blinded, controlled trial to determine whether therapeutic touch administered at the time of stereotactic core biopsy of suspicious breast lesions, using visual analog scales to measure change in pain and anxiety, found no significant differences between the study arms regarding post biopsy pain, anxiety, fearfulness, or physiological parameters (Frank et al., 2007). In terms of acupuncture, a qualitative study on the opinions of sixteen women having ear acupuncture for hot flushes and night sweats associated with adjuvant hormonal treatment for breast cancer found that acupuncture was helpful and relaxing; the women reported reductions in hot flush frequency, as well as improvements in overall emotional and physical wellbeing (Walker et al., 2007). A large randomized trial involving 19 community clinical oncology programs on the effect of acupressure for chemotherapy-induced nausea and vomiting, with balanced intervention and a control group consisting of 160 women under chemotherapy, revealed no significant differences in acute nausea or emesis by treatment group. With delayed nausea and vomiting, the acupressure group had a statistically significant reduction in the amount of vomiting and the intensity of nausea over time when compared with the placebo and usual-care groups (Dibble et al., 2007). Recently, results of two trials on the effect of lifestyle interventions on quality of life have been published in the Journal of Clinical Oncology. One study examining the impact of yoga, including physical poses, breathing, and meditation exercise, on Functional Assessment of Cancer Therapy quality of life measures such as fatigue, distressed mood, and spiritual wellbeing among a multiethnic sample of 128 patients found the control group had greater decrease in social well-being compared with the intervention group after controlling for baseline social well-being and covariates. Favorable outcomes were also detected for the
Stereotactic core biopsy
Hot flushes and night sweats associated with adjuvant hormonal treatment
Chemotherapyinduced nausea and vomiting
Breast cancer therapy
Breast cancer patients initiating adjuvant chemotherapy
Breast cancer An integrated patients undergoing yoga program radiotherapy
Frank et al., (2007)
Walker et al., 2007
Dibb et al., (2007)
Moadel et al., (2007)
Courneya et al. (2003)
Banerjee et al. (2007)
Anxiety and depression
QOL domain
Hospital Anxiety and Depression Scale (HADS), and Perceived Stress Scale (PSS)
Observational study, Better outcomes in HADS scores stratified by Yoga and PSS in the yoga group intervention and control group. N = 68
A multi-center randomized controlled trial N = 242
CAM use is particularly popular in patients with breast cancer. The reasons cited for using CAM were to boost the immune system, improve quality of life, prevent recurrence of cancer, provide control over life, and treat breast cancer and the side effects of treatment. The studies are often limited in design quality (Nahleh and Tabbara, 2003). More evidence from clinical trials has emerged in recent years, which has been mixed, showing both favorable and unfavorable results of CAM use in managing breast cancer treatment
Psychological stress and radiation-induced genotoxic stress
Functional Assessment of Cancer Therapy – Anemia scale
No statistically significant difference
Deduction in delayed nausea and vomiting
Improvements in overall emotional and physical well being
Physical Aerobic and resistance exercise functioning, cancer-specific QOL
Randomized trial, N = 160
focus groups, Qualitative study on transcripts were the opinions, N = 16 analyzed using grounded techniques Observed frequency of nausea and vomiting
Results did not suggest evidence of efficacy
Main findings
Randomized, Patient- No significant differences blinded controlled between the arms trial, N = 82
Randomized trial, N = 32
Study design and sample size
Improvements in overall quality of life, emotional well-being, social well-being, spiritual wellbeing, and distressed mood
Acute nausea or emesis, delayed nausea and vomiting
Overall emotional and physical well being
Visual analog scale
Self-reported questionnaires
QOL instruments
Yoga, including Fatigue, distressed Functional Randomized trial, physical poses, mood, and spiritual Assessment of Cancer multiethnic sample, breathing, and well-being Therapy N = 128 meditation exercise
Acupuncture
Acupuncture
Krieger-Kunz Pain and anxiety Therapeutic touch
Reiki
Breast biopsy
Potter (2007)
CAM modalities
Targeted treatment
Author
. Table 103-3 CAM modalities and QOL domains and instruments in the recent studies in breast cancer management
Alternative Therapies and Quality of Life
103 1773
1774
103
Alternative Therapies and Quality of Life
intervention group in overall quality of life, emotional well-being, social well-being, spiritual well-being and distressed mood in a secondary analysis for those not receiving chemotherapy (Moadel et al., 2007). The other study, a multicenter randomized controlled trial with a sample size of 242 breast cancer patients initiating adjuvant chemotherapy in Canada, evaluated the relative merits of aerobic and resistance exercise in blunting the unfavorable changes in physical functioning, body composition, psychosocial functioning, and quality of life measured by the Functional Assessment of Cancer Therapy – Anemia scale. The study reported that changes in cancer-specific QOL, fatigue, depression, and anxiety favored the exercise groups but did not reach statistical significance (Courneya et al., 2007). In terms of the strength of evidence of these two studies, a common limitation appeared to be a lack of adequate controls, such as group context, instructor attention, number of contacts, and other confounding factors in the design of the lifestyle interventions (Denmark-Wahnefried, 2007). Recently, additional evidence has been reported on the effect of an integrated yoga program in modulating perceived stress levels, anxiety and depression levels as measured by the Hospital Anxiety and Depression Scale (HADS) and the Perceived Stress Scale (PSS) involving 68 breast cancer patients undergoing radiotherapy in Singapore. The study reported a significant decrease in the HADS scores in the yoga intervention group. Mean PSS was decreased in the yoga group, whereas there was no change in the control group (Banerjee et al., 2007). A summary of recent findings on the effect of various CAM modalities on QOL can be found in > Table 103-3. In terms of Tai Chi, a systematic review was performed of 27 studies on the effect of Tai Chi in breast cancer treatment, among which three were randomized controlled trials (RCTs) and one a non-randomized controlled trial. Two of the RCTs reported significant differences in psychological and physiological symptoms compared to a psychosocial support control. Most trials suffered from methodological flaws such as small sample size, inadequate study design, and poor reporting. The author concluded that the evidence is not convincing enough to suggest Tai Chi is an effective supportive treatment for breast cancer (Lee et al., 2007). Broadly, a systematic review of randomized clinical trials on the effect of CAM on the treatment of breast cancer was done, by searching the electronic databases Cochrane Central Register of Controlled Trials, MEDLINE, EMBASE, Allied and Complementary Medicine, Scirus, BIOSIS, CanderLit and CINAHL, and including the findings of ongoing trials at the MetaRegister and the National Research Register. The review examined 13 RCTs using treatments including various methods of psychosocial support such as group support therapy, cognitive behavioral therapy and cognitive existential group therapy, a combination of muscle relaxation training and guided imagery, the Chinese herbal remedy Shi Quan Da Bu Tang, thymus extract, and melatonin. The review revealed that in most cases the methodological quality of these studies is low, and there existed little convincing evidence to support CAM use to improve QOL (Ernst et al., 2006).
6
CAM Use and QOL in Patients with Musculoskeletal Disease
There is a growing body of literature on the efficacy of CAM in the treatment of musculoskeletal disease. Of particular interest is the effect of CAM on treating osteoarthritis (OA). OA affects 20 million Americans and is the sixth leading cause of disability in developed countries (Marray and Lopez, 1996). The consequences of OA are massive: patients suffer from limitations in physical function, postural instability, sleep disturbance, and psychosocial disability
Alternative Therapies and Quality of Life
103
(Weiner et al., 2007). In the US, Canada, the UK, France, and Australia, the direct costs of medical treatment and indirect costs of social welfare due to OA account for 1–2.5% of gross domestic product (March and Bachmeier, 1997). Between 2004 and 2006, three studies on the effects of CAM in treating OA have been published in leading medical journals including Annals of Internal Medicine and The Lancet, reflecting a growing interest in this subject and improving quality of study design in CAM research. Annals of Internal Medicine published a study in 2006 on a randomized controlled trial on the efficacy of traditional Chinese acupuncture (TCA) compared with sham acupuncture (needling at defined nonacupuncture points) and conservative therapy in 1,007 patients who had chronic pain for at least six 6 months due to OA of the knee. The main outcome was measured by the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) score. The study results suggested that compared with physiotherapy and as-needed antiinflammatory drugs, addition of either TCM or sham acupuncture led to greater improvement in WOMAC at 26 weeks. No statistical significant difference was observed between TCA and sham acupuncture. The limitations of this study include the lack of blinding between acupuncture and conservative therapy, and no monitoring of acupuncture compliance with study protocol (Scharf et al., 2006). The study by Scharf et al. provided further evidence from a larger sample, in addition to the finding of another study published by Annals of Internal Medicine 2 years earlier in 2004 on the same subject, in which the study suffered from high drop-out in the follow-up period. In that study, Berman and colleagues studied 570 patients with OA of the knee who were randomized into a true acupuncture intervention group, a sham acupuncture and an education control group. Primary outcomes were measured by WOMAC pain and function scores. The secondary outcome included patient global assessment and physical health scores of the 36-item Short-Form Health Survey (SF-36). The study results showed that participants in the true acupuncture group experienced greater improvement in WOMAC function scores than the sham acupuncture group at eight 8 weeks, but not in WOMAC pain score or the patient global assessment. At 26 weeks, the true acupuncture group experienced significantly greater improvement than the sham group in the WOMAC function score, WOMAC pain score, and patient global assessment. The study was limited in that 43% of the participants in the education group and 25% in each of the true and sham acupuncture groups were not available for analysis at 26 weeks (Berman et al., 2004). The Lancet published one study in 2005 on a randomized controlled trial with 300 patients with chronic OA of the knee, which has a smaller sample size and less positive results when compared to the two studies published in Annals of Internal Medicine described above. Witt and colleagues developed a three-arm study design of acupuncture, minimal acupuncture (superficial needling at non-acupuncture points), and a waiting list control. The study aimed to assess the effectiveness of acupuncture on patients with chronic OA, using the same outcome measure of WOMAC score. The study reported that after eight 8 weeks of treatment, pain and joint function improved more with acupuncture than with minimal acupuncture or no acupuncture, but such difference was not significant when measured after 52 weeks. The authors reasoned that the beneficial effect of acupuncture decreases over time (Witt et al., 2005). These three studies by Scharf and colleagues, Berman and colleagues, and Witt and colleagues collectively suggest that there was some effect associated with acupuncture in OA treatment measured by WOMAC, at least in the short time period. In the meantime, since the study by Berman and colleagues suffered from higher drop-out rates, the longer-term result in that study is less convincing. Witt and colleagues found no longer-term effect. Scharf and
1775
1776
103
Alternative Therapies and Quality of Life
colleagues suggested that there was some effect at 52 weeks; however, since there was no statistically significant difference between the acupuncture and sham acupuncture groups and there was no blinding in the study design, the effect observed from the acupuncture group could very well be the placebo effect, which was not related to the pressure point although the principles of acupuncture would suggest otherwise. Such a possibility of placebo effect was further reinforced by a more recent trial published in British Medical Journal in 2007. Foster and colleagues conducted a study on the benefit of adding acupuncture to a course of advice and exercise delivered by physiotherapists for pain reduction in patients with OA in a multi-center, randomized controlled trial involving 352 adults aged 50 or more, in three arms: advice and exercise, advice and exercise plus true acupuncture, and advice and exercise plus non-penetrating acupuncture. The outcome measures included WOMAC pain subscale, physical function, pain intensity and unpleasantness of pain at 2- to 12-month intervals. The results indicated that compared with advice and exercise alone there were small, statistically significant improvements in pain intensity and unpleasantness at 2 and 6 weeks for true acupuncture and at all follow-up points for non-penetrating acupuncture. There was no difference in the pain scores (Foster et al., 2007). Supporting evidence on possible decreasing long-term effect can also be found in a recent study by Weiner and colleagues in a randomized and controlled trial, examining the efficacy of periosteal stimulation therapy (PST, osteopuncture) for the treatment of a small sample of 87 community-dwelling older adults with OA-associated chronic knee pain, measured by WOMAC and physical performance (Short Physical Performance Battery or SPPB) using a simple two-arm intervention-control study design. Weiner and colleagues found that pain was reduced significantly more in the PST group than in the control group at post and at 1 month later, but by 2 months, pain levels had regressed to pre-intervention levels (Weiner et al., 2007). A summary of the findings on the effect of acupuncture on QOL in patients with OA can be found in > Table 103-4. Acupuncture in OA treatment is perhaps one of the most popular areas that attracts research attention and has strong cumulative evidence from high-quality research in CAM studies. However, patients with OA have been using a broad array of CAM modalities. A recent observational study by Zhang and colleagues examined a sample of 547 OA patients in Hong Kong on the usage of 11 CAM modalities including chiropractic manipulation, acupuncture, herbal remedies, energy healing, folk remedies, exercise, homeopathy, lifestyle diets, copper bracelets, megavitamin therapy, and magnetic applications, and its association with QOL as measured by SF-36 health scales and an overall Health Utility Index derived from a pre-scored multi-attribute classification system based upon SF-36 health surveys. The study found that the 547 OA patients had used a wide spectrum of alternative therapies and often used a multiplicity of them. After adjusting for socioeconomic variables, years of OA and severity of OA, the use of CAM was not statistically significantly associated with an improvement in the QOL. At the same time, the use of CAM was statistically significantly associated with side effects, including gastric discomfort and gastric ulcer-bleeding (Zhang et al., 2007).
7
Effect of CAM Use on QOL in Treatment of Patients with Other Conditions
CAM has been used extensively in heterogeneous populations worldwide. A recent study suggested that transcendental meditation can be effective in improving the quality of life
Alternative Therapies and Quality of Life
103
. Table 103-4 Acupuncture and QOL in patients with osteoarthritis (OA) Author
QOL instruments
RCT study arms and sample size
Main findings
Scharf et al. (2006)
WOMAC score
Three-arm including acupuncture, sham acupuncture, and conservative therapy. N = 1,007
Improvement in WOMAC in acupuncture and sham acupuncture group. No differences between TCA and sham acupuncture
Berman et al. (2004)
WOMAC SF-36
Three-arm including true acupuncture intervention group, sham acupuncture and education control group. N = 570
Greater improvement in true acupuncture group than the sham acupuncture group in the WOMAC function score, WOMAC pain score, and patient global assessment
Witt et al. (2005)
WOMAC score
Three-arm study design of acupuncture, minimal acupuncture, and a waiting list control. N = 300
No significant difference measured after 52 weeks
Foster et al. (2007)
WOMAC pain subscale, and function, pain intensity and unpleasantness of pain
Three arms of advice and exercise, advice and exercise plus true acupuncture, and advice and exercise plus nonpenetrating acupuncture. N = 352
Small improvements in pain intensity and unpleasantness at two and 6 weeks for true acupuncture and nonpenetrating acupuncture
The studies collectively suggest that there was some effect associated with acupuncture in OA treatment measured by WOMAC, at least in the short time period; however, since there was no statistically significant difference between the acupuncture and sham acupuncture groups, the effect observed from the acupuncture group could very well be the placebo effect
and functional capacity of African American patients with congestive heart failure (CHF) (Jayadevappa et al., 2007) and CHF patients performing Tai Chi exercise had an improvement in symptom scores of heart failure and depression scores measured by the SCL-90-R questionnaire (Barrow et al., 2007). Another recent study has suggested that combining relaxation response and acupuncture treatment can improve the QOL of HIV/AIDS patients (Chang et al., 2007). An additional recent study also indicated that, although there is no effective treatment for irritable bowel syndrome (IBS) and many patients turn to herbal medicine for possible cure, the therapeutic efficacy of a popular ancient herbal Chinese formula in patients with diarrhea-predominant IBS appeared to be ineffective (Leung et al., 2006). A summary of those recent findings can be found in > Table 103-5.
8
Conclusion
Overall, the use of CAM is prolific in population and diverse in modalities. However, the evidence on the efficacy of CAM use on QOL is not convincing. Even for those areas of CAM
1777
Combining relaxation response and acupuncture treatment
Herbal medicine
Chang et al., Patients 2007 with HIV/ AIDS
Leung et al., Patients (2006) with irritable bowel syndrome (IRB)
QOL instruments
Study design and sample size
Symptoms and QOL
Body functioning, spiritual well-being
Symptom scores, and depression scores
Improvements in walking, SF-36 subscales, total score on Minnesota Living with Heart Failure scale, and CES-D
Main findings
IBS symptom scores, short-form SF36
Medical Outcome Study HIV health survey, the Functional Assessment of HIV infection, and the Functional Assessment of Chronic Illness Therapy-Spiritual Well-Being
Randomized doubleblinded placebocontrolled trail N = 119
Two-arm doubleblind randomized controlled trial N = 119
No difference between the intervention and control group
The intervention group showed trends of greater improvements than the control group in emotional, spiritual/peace, physical and mental health
Minnesota Living with Heart Failure, Randomized to Tai Improvement in symptom scores depression scores by SCL-90-R Chi Chuan and of heart failure and depression standard medical care scores without exercise rehabilitation N = 52
Walk, Center for Epidemiologic Depression Randomized to either generic and Scale (CES-D), SF-36, Minnesota TM or health diseaseLiving with Heart Failure scale education N = 23 specific QOL
QOL domain
CAM has been used extensively in heterogeneous populations worldwide. Results on the effect of CAM on QOL varied widely, based upon study population, and CAM modalities
Tai Chi Chuan
Barrow Congestive et al., (2007) heart failure patients
CAM modalities
Transcendental meditation
Targeted treatment
103
Jayadevapp Congestive et al., (2007) heart failure
Author
. Table 103-5 CAM use on QOL in other major conditions
1778 Alternative Therapies and Quality of Life
Alternative Therapies and Quality of Life
103
which have been tested with several relatively large trials, it is inconclusive whether the observed effect is due to true CAM intervention or placebo effect. Often CAM studies are limited in sample size, study design, and control. Patients, researchers, and purchasers should be aware that while there has been an enormous growth of CAM use in recent decades, the efficacy of various CAM modalities on QOL is still not proven.
Summary Points The proportion of the general population in the United States that reports using comple-
mentary and alternative medicine (CAM) has increased from 33.8% in 1990 to 62.1% in 2002. Americans are visiting CAM practitioners more frequently than primary care physicians. Worldwide, the prevalence rates of CAM use in adults ranged from 9 to 70%. There exists a lack of consistency in the body of CAM literature as to precisely what CAM is. The National Center for Complementary and Alternative Medicine (NCCAM), a branch of the National Institutes of Health, has defined complementary and alternative medicine (CAM) as “a group of diverse medical and health care systems, practices, and products that are not presently considered to be part of conventional medicine.” There is also a paucity of evidence on the efficacy of CAM. Historically, the common drawbacks in CAM research include poor study design, small sample size, and inadequate controls. In recent years, high-quality research articles have begun to emerge in leading medical journals. There is a strong interest in the efficacy of CAM in improving the quality of life (QOL) in areas where conventional medicine is inadequate, such as cancer treatment and the treatment of musculoskeletal disease. The evidence on the efficacy of CAM use on QOL is not convincing. Even for those areas of CAM which have been tested with several relatively large trials, it is unclear whether the observed effect is due to true CAM intervention or placebo effect.
References Astin JA. (1998). JAMA. 279: 1548–1553. Banerjee B, Vadiraj HS, Ram A, Rao R, Jayapal M, Gopinath KS, Ramesh BS, Rao N, Kumar A, Raghuram N, Hegde S, Nagendra HR, Prakash Hande M. (2007). Integr Cancer Ther. 6: 242–250. Barnes PM, Powell-Griner E, McFann K, Nahin RL. (2004). In: CDC Adv Data Report: Complementary and Alternative Medicine Use Among Adults: United States, 2002. pp. 1–19. Barrow DE, Bedford A, Ives G, O’Toole L, Channer KS. (2007). Postgrad Med J. 83: 717–721. Barzansky B, Etzel SI. (2003). JAMA. 290: 1190–1196. Berman BM, Lao L, Langenberg P, Lee WL, Gilpin AM, HochbergMC. (2004). AnnInternMed.141: 901–910.
Chang BH, Boehmer U, Zhao Y, Sommers E. (2007). J Altern Complement Med. 13: 807–815. Courneya KS, Segal RJ, Mackey JR, Gelmon K, Reid RD, Friedenreich CM, Ladha AB, Proulx C, Vallance JK, Lane K, Yasui Y, McKenzie DC. (2007). J Clin Oncol. 25: 4396–404. Crawford NW, Cincotta DR, Lim A, Powell CVE. (2006). BMC Complementary and Alternative Medicine. 6: 16. de la Fuente-Fernandez de la R, Ruth TJ, Sossi V, Schultzer M, Calne DB, Stoessl AJ. (2001). Science. 3: 49–57. Demark-Wahnefried W. (2007). J Clin Oncol. 25: 4344–4345.
1779
1780
103
Alternative Therapies and Quality of Life
Dibble SL, Luce J, Cooper BA, Israel J, Cohen M, Nussey B, Rugo H. (2007). Oncol Nurs Forum. 34: 813–820. Diefenbach MA, Hamrick N, Uzzo R, Pollack A, Horwitz E, Greenberg R, Engstrom PF. (2003). J Urol. 170: 166–169. DiGianni LM, Garber JE, Winer EP. (2002). J Clin Oncol. 20: 34S–38S. Downer SM, Cody MM, McCluskey P, Wilson PD, Arnott SJ, Lister TA, Slevin ML. (1994). BMJ. 309: 86–89. Dutta AP, Daftary MN, Egba PA, Kang H. (2003). J Am Pharm Assoc. 43: 81–83. Dy GK, Bekele L, Hanson LJ, Furth A, Mandrekar S, Sloan JA, Adjei AA. (2004). J Clin Oncol. 22: 4810–5. Eisenberg DM, Davis RB, Ettner SL, Appel S, Wilkey S, Van Rompay M, Kessler RC. (1998). JAMA. 280: 1569–1575. Eisenberg DM, Kessler RC, Foster C, Norlock FE, Calkins DR, Delbanco TL. (1993). N Engl J Med. 328: 246–252. Ernst E, Resch KL, Mills S, Hill R, Mitchell A, Willoughby M. (1995). Br J Gen Pract. 45: 506. Ernst E. (2000). Bull World Health Organ. 78(2): 252–257. Ernst E, Schmidt K, Baum M. (2006). Breast J. 12: 526–530. Foster NE, Thomas E, Barlas P, Hill JC, Young J, Mason E, Hay EM. (2007). BMJ. 335: 436. Frank LS, Frank JL, March D, Makari-Judson G, Barham RB, Mertens WC. (2007). Pain Med. 8: 419–424. Gevitz N. (ed.) (1988). Unorthodox Medicine in America. John Hopkins University Press. Institute of Medicine Committee on the Use of Complementary and Alternative Medicine. (2005). Complementary and Alternative Medicine in the United States. The American Public National Academies Press. Jayadevappa R, Johnson JC, Bloom BS, Nidich S, Desai S, Chhatre S, Raziano DB, Schneider R. (2007). Ethn Dis. 17: 72–77. Jonas W. (2002). Ann Amer Acad Politi Soc Sci. 583: 29–43. Kao GD, Devine P. (2000). Cancer. 88: 615–619. Kaptchuk TJ, Eisenberg DM. (2001). Ann Intern Med. 135: 196–204. Kelly JP, Kaufman DW, Kelley K, Rosenberg L, Anderson TE, Mitchell AA. (2005). Arch Intern Med. 165: 281–286. Kessler RC, Davis RB, Foster DF, Van Rompay MI, Walters EE, Wilkey SA, Kaptchuk TJ, Eisenberg DM. (2001). Ann Intern Med. 135: 262–268. Lee MS, Pittler MH, Ernst E. (2007). Support Care Cancer. 15: 597–601.
Leung WK, Wu JC, Liang SM, Chan LS, Chan FK, Xie H, Fung SS, Hui AJ, Wong VW, Che CT, Sung JJ. (2006). Am J Gastroenterol. 101: 1574–1580. Lim P, Sadarangani H, Chan J, H.ng. (2005). Complementary Therapies in Med. 13: 16–24. March LM, Bachmeier CJM. (1997). Ballieres Clinical Rheumatology. 11: 817–834. Moadel AB, Shah C, Wylie-Rosett J, Harris MS, Patel SR, Hall CB, Sparano JA. (2007). J Clin Oncol. 25: 4387–4395. Moerman DE, Jonas WB. (2002). Ann Intern Med. 136: 471–476. Murray JL and Lopez AD. (1996). The Global Burden of Disease. World Health Organization, Geneva, Switzerland. Nahin RL, Pontzer CH, Chesney MA. (2005). Health Aff. 24: 991–993. Nahleh Z, Tabbara IA. (2003). Palliat Support Care. 1: 267–273. National Institute of Cancer. (2005). Cancer Trends Progress Report – 2005 Update. NCCAM. (2002). Get the Facts: What is Complementary and Alternative Medicine., National Institutes of Health – National Center for Complementary and Alternative Medicine. Potter PJ. (2007). J Holist Nurs. 25: 238–248. Richardson J. (2004). Am J Public Health. 94: 1049–1053. Richardson MA, Masse LC, Nanny K, Sanders C. (2004). Support Care Cancer. 12: 797–804. Richardson MA, Sanders T, Palmer JL, Greisinger A, Singletary SE. (2000). J Clin Oncol. 18: 2505–2514. Ruggie M. (2005). Health Aff. 24: 980–990. Scharf HP, Mansmann U, Streitberger K, Witte S, Kramer J, Maier C, Trampisch HJ, Victor N. (2006). Ann Intern Med. 145: 12–20. Verhoef MJ, Balneaves LG, Boon HS, Vroegindewey A. (2005). Integr Cancer Ther. 4: 274–286. Walker G, de Valois B, Davies R, Young T, Maher J. (2007). Complement Ther Clin Pract. 13: 250–257. Warrick PD, Irish JC, Morningstar M, Gilbert R, Brown D, Gullane P. (1999). Arch Otolaryngol Head Neck Surg. 125: 573–579. Weiner DK, Rudy TE, Morone N, Glick R, Kwoh CK. (2007). J Am Geriatr Soc. 55: 1541–7. Wetzel MS, Kaptchuk TJ, Haramati A, Eisenberg DM. (2003). Ann Intern Med. 138: 191–196. Witt C, Brinkhaus B, Jena S, Linde K, Streng A, Wagenpfeil S, Hummelsberger J, Walther HU, Melchart D, Willich SN. (2005). Lancet. 366: 136–143. Wong-Kim E, Merighi JR. (2007). J Health Care Poor Underserved. 18: 118–129. Zhang JX, Woo J, Lau WC, Lee P, Chiu P, Lam D. (2007). Am J Chin Med. 35: 183–93.
104 Leisure-Time Physical Activity and Quality of Life A. Vuillemin 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1782
2 2.1 2.1.1 2.1.2 2.1.3 2.1.4
Observation Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1784 Cross-Sectional Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1784 In Children . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1784 In Adults . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1785 In Elderly . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1786 Summary of Cross-Sectional Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1787
3 3.1 3.2 3.3 3.4 3.5
Longitudinal Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1788 In Children . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1788 In Adults . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1788 In Elderly . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1790 Summary of Longitudinal Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1790 Summary of Observation Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1790
4
Intervention Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1791
5
Follow-Up of Advice Given by a General Practitioner and/or a Specialist of PA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1791
6
The Effects of a Resistance Training Program on Quality of Life . . . . . . . . . . . . . 1792
7
The Effects of an Endurance Exercises Program on Quality of Life . . . . . . . . . . . 1793
8
The Effects of a Specific Exercises Program on Quality of Life . . . . . . . . . . . . . . . . 1793
9
The Effects of a Walking Program on Quality of Life . . . . . . . . . . . . . . . . . . . . . . . . . 1794
10
The Effects of an Exercises Program and a Diet on Quality of Life . . . . . . . . . . . 1794
11
Summary of Intervention Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1795
12
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1795 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1797
#
Springer Science+Business Media LLC 2010 (USA)
1782
104
Leisure-Time Physical Activity and Quality of Life
Abstract: The benefits of > physical activity (PA) on health are well established and there is strong evidence of protective effects of regular PA against major chronic disease. In line with this, increased PA and improved fitness were shown to result in better quality of life (QOL) in the general population. In spite of the use of various measuring instruments of PA and QOL, and the adjustment on a more or less important number of QOL determinants, regular PA appears to improve QOL by enhancing psychological well-being and by improving physical functioning. More precisely, a regular PA is globally associated with a better QOL both in the physical domains with a cross-sectional approach and in the mental domains with a longitudinal approach. Most of the studies were interested in overall level of Physical Activity and only few studies were specifically interested in > leisure-time Physical Activity. However, behavior during leisure remains most easily modifiable compared with PA at work, for example. In addition, intervention studies involved elderly and none was conducted in children or teenagers. Some results remain contradictory but it is possible to assert that an exercises program has beneficial effects on QOL, particularly on its physical dimension. It is interesting to emphasize that the most promising results were obtained from programs facilitating a regular practice equivalent to the current > public health recommendations for physical activity. List of Abbreviations: LTPA, Leisure-Time Physical Activity; PA, Physical Activity; QOL, Quality of Life
1
Introduction
If the benefits of physical activity (PA) on health, measured using objective criteria, are now widely demonstrated, its effects on quality of life (QOL) remain insufficiently investigated in the general population. For most people, QOL decrease with retirement, due to functional capacities decrease and chronic conditions which may limit the performance in activities of daily living. This can explain the biggest interest for elderly populations in the studies investigating relation between level of PA and QOL (Ellingson and Conn, 2000; Rejeski et al., 1996; Rejeski and Mihalko, 2001; Spirduso and Cronin, 2001). A recent systematic review including studies in the general adult population concluded that cross-sectional data showed a consistently positive association between PA and QOL but randomized controlled trials and longitudinal studies are necessary to investigate further this association (Bize et al., 2007). More precisely, the relation between PA and QOL was investigated in intervention studies in which QOL was measured before and after an exercises program. Most of these studies involved subjects with chronic conditions and showed improvement in QOL after intervention. But observational studies are scarce in these populations. Studies conducted in the general population are less numerous. Furthermore, many studies were interested in the effect of PA on one dimension of QOL but few studies used a multidimensional measure of QOL. In addition, overall PA was mostly considered and little is known regarding QOL and its relation with leisure-time PA. > Table 104-1 presents a synthesis of the studies published according to their type (observation/intervention) and population involved (target population/gender). The results of observational studies suggest that PA is positively associated with QOL and that PA is associated both with the physical dimension and the mental dimension (Acree et al., 2006; Brown et al., 2003, 2004; Chen et al., 2005a; Conn et al., 2003; Daskapan et al., 2005;
Leisure-Time Physical Activity and Quality of Life
104
. Table 104-1 Studies published according to their type and population involved Population
Observation studies
Gender Children Boys
Adults
Cross-sectional –
Intervention studies
Longitudinal –
– –
Girls
–
–
Both
Chen et al. (2005a)
Chen et al. (2005b)
–
Men
–
–
Hellenius et al. (1995)
Women Kull (2002)
Wolin et al. (2007)
Lindh-Astrand et al. (2004) Saavedra et al. (2007)
Both
Laforge et al. (1999) Wendel-Vos et al. (2004)
Sorensen et al. (1999)
Brown et al. (2003)
Elley et al. (2003)
Tessier et al. (2007)
Brown et al. (2004)
Lee et al. (2004)
Daskapan et al. (2005) Vuillemin et al. (2005) Shibata et al., (2007) Zahran et al. (2007) Elderly
Men
–
Women Koltyn (2001)
–
Jette et al. (1996)
Lee and Russell (2003) Damush et al. (1999)
Conn et al. (2003) Both
Stewart et al. (2003) – Acree et al. (2006)
Stewart et al. (1997) Fabre et al. (1999) Halbert et al. (2000) King et al. (2000) Li et al. (2001a) Li et al. (2001b) Li et al. (2002) Hopman-Rock and Whesthoff (2002) Fisher and Li (2004) Stiggelbout et al. (2004)
Koltyn, 2001; Kull, 2002; Laforge et al., 1999; Shibata et al., 2007; Stewart et al., 2003; Vuillemin et al., 2005; Zahran et al., 2007). However the cross-sectional approach of these studies doesn’t allow to conclude if PA influence QOL or vice versa. To our knowledge, there are few longitudinal studies in this domain; the results showed that the relation persists in children (Chen et al., 2005b) and adults but essentially for the mental component of QOL in both men (Wendel-Vos et al., 2004) and women (Tessier et al., 2007; Wolin et al., 2007) as well as in elderly (Lee and Russell, 2003).
1783
1784
104
Leisure-Time Physical Activity and Quality of Life
Among intervention studies, few were interested in adults and none in children. The results seem rather contradictory but the reviews available in the literature showed that regular PA is associated with higher QOL (Bize et al., 2007; Ellingson and Conn, 2000; Rejeski et al., 1996; Rejeski and Mihalko, 2001; Spirduso and Cronin, 2001).
2
Observation Studies
2.1
Cross-Sectional Studies (> Table 104-2)
Cross-sectional studies, excepted three studies (Conn et al., 2003; Stewart et al., 2003; Zahran et al., 2007), observed this association in both men and women. One study involved children (Chen et al., 2005a), two involved students (Daskapan et al., 2005; Zahran et al., 2007) and the others involved adults (Brown et al., 2003, 2004; Kull, 2002; Laforge et al., 1999; Shibata et al., 2007; Vuillemin et al., 2005), or elderly (Acree et al., 2006; Conn et al., 2003; Koltyn, 2001; Stewart et al., 2003). Among four studies carried out in elderly, two concerned exclusively women (Conn et al., 2003; Koltyn, 2001).
2.1.1
In Children
Chen et al. (2005a) conducted a cross-sectional study in children aged 12- to 13-year-old (n = 7,887) to examine associations between lifestyle (including PA) and QOL (COOP . Table 104-2 Significant relationship between physical activity and quality of life in cross-sectional studies Quality of life Mental component Cross-sectional studies
Men
Physical component
Women
Men
Women
Children Chen et al. (2005a)
X
X
X
X
Laforge et al. (1999)
X
X
X
X
Kull (2002)
–
X
–
–
Brown et al. (2003, 2004)
X
X
X
X
Daskapan et al. (2005)
X
X
X
X
Vuillemin et al. (2005)
X
X
X
X
Shibata et al. (2007)
X
X
X
X
Zahran et al. (2007)
NS
NS
NS
NS
Koltyn (2001)
–
NS
–
X
Conn et al. (2003)
–
NS
–
NS
Adults
Elderly
Stewart et al. (2003)
NS
NS
NS
NS
Acree et al. (2006)
NS
NS
X
X
X beneficial effect of physical activity on quality of life; NS no statistically significant effect of physical activity on quality of life; – not studied
Leisure-Time Physical Activity and Quality of Life
104
Japanese validated version). PA indicator used in this study was the frequency (very often, often, rarely, almost never) measured using a self-reported physical activity questionnaire for children validated in Japanese (Chen et al., 2003). The results showed an association between the frequency of PA and the eight dimensions of QOL investigated (physical condition, feelings, daily activities, social activities, pains, health change, general health, and global quality of life). Logistic regression showed that children who practiced almost never PA had between 1.30 and 6.35 times more risk of having a poor QOL compared with those who had a practice very often (model adjusted on sex, age, body mass index, social standing and headaches). The most significant results were found for the physical condition dimension with risks equal to 2.14, 4.32 and 6.35 for the categories of frequency, rather often, rarely and almost never, respectively, compared to very often.
2.1.2
In Adults
In a population of students, Daskapan et al. examined the relation between the usual level of PA measured with the questionnaire of Paffenbarger (Paffenbarger et al., 1993) and the level of QOL (Turkish version of the SF-36) (Daskapan et al., 2005). The results showed that the weekly energy expense was positively associated with four dimensions of QOL (physical activity, vitality, social functioning and mental health), after adjustment on body mass index and gender. To clarify this relation, the subjects were divided in two categories according to the level of energy expense suggested in the public health recommendations for PA: high energy expense (1,500 kcal/week) and low energy expense (5 points). Important scores differences were observed between the QOL dimensions according to the PA categories (1.6–9 points at the women and 2.2–9 points in men). Most of these QOL scores differences were clinically significant, particularly in women and when the inactive group was compared to the intense PA group. In summary, the results suggest that the practice of 30 min per day of moderate intensity PA during leisure can be beneficial; a higher intensity being associated with a better QOL, which is in line with Brown et al. (Brown et al., 2003, 2004). Also in line with these results, Shibata et al. showed that in an adult population, meeting the recommended level of PA was associated with better scores on some dimensions of QOL (SF-8) than those who did not (Shibata et al., 2007). The authors emphasized that engaging in PA, even at insufficient levels, had a positive effect on the perception of physical functioning in both genders. These results strengthen the interest of the public health recommendations for PA. In women aged 18–45 years old (n = 660) PA practice was associated with a better QOL (Kull, 2002). This study focused on the mental health dimension of QOL measured with the General Health Questionnaire (GHQ). PA was measured using a single question allowing to classify the subjects in three groups: active person (PA = 3 times per week, 12.7%), moderately active person (PA 1–2 times per week, 34.8%), inactive ( levels of Physical Activity and QOL (physical health, social relationships,
Leisure-Time Physical Activity and Quality of Life
104
and environment) in women living in an independent way (no significant difference for the domain mental health). An analysis of correlation, conducted independently of the type of environment, showed a relation between the global level of QOL and the energy expense (r = 0.45) as well as intense activities (r = 0.58). It was also shown a relation between physical health dimension and total duration of PA (hour/week) (r = 0.46), energy expense (r = 0.47) and intense activities (r = 0.54). No relation was identified between the psychological health, social relationships and environment dimensions of QOL and the level of PA. Conn et al. conducted a study in 198 sedentary women aged 65–97 years. The secondary objective of this study was to examine the association between the adulthood level of PA (until the beginning of the study) and QOL (Conn et al., 2003). The level of PA (endurance exercises) was not measured with a validated instrument. QOL was measured using the SF-36 and the Dartmouth Primary Care Cooperative Information Project (COOP). The correlations between the indicators of PA (number of different PA and total PA duration) and QOL (global score) were very low and not statistically significant. The authors suggested that the results can be partially explained because subjects reported only very few PA during adulthood and the average age at the beginning of PA was about 60–70 years. A study involving 38 men and 44 women aged 55–75 years and not engaged in a moderate or intense PA in a regular way, tried to determine if a more desirable level of aerobic and muscular fitness, of usual PA and body fat was associated with a better QOL (Stewart et al., 2003). The level of PA (expressed in kcal/kg/j) was measured using the Standford 7-day Physical Activity Recall (PAR) questionnaire (Blair et al., 1985) and the QOL using the SF-36. The usual level of PA was only associated with a decrease of the physical pains score (p < 0.01). In one other study, healthy elderly who participated in PA of at least moderate intensity (overall level of PA) for more than 1 h/week had higher QOL in both physical and mental domains of the SF-36 than those who were less active (Acree et al., 2006), particularly physical activity, role-physical, bodily pain, vitality and social functioning.
2.1.4
Summary of Cross-Sectional Studies
To date, few studies were led in a child population, but to our knowledge, the only study available (Chen et al., 2005a) showed that PA is beneficial for QOL with a more marked effect on physical condition dimension. In adults, five (Brown et al., 2003, 2004; Daskapan et al., 2005; Shibata et al., 2007; Vuillemin et al., 2005) of the seven studies presented, centered their interest on the relation between the follow-up of PA recommendations and QOL and found a beneficial effect of PA on QOL in the subjects who reach the level of recommended PA; with however a threshold effect put in evidence by Brown and al., with a weaker QOL in the subjects who declare less than 20 min or more than 90 min of PA per day (Brown et al., 2004). Laforge et al. confirmed the observed tendency by using an original approach based on the stages of change based on PA which allow to better target the actions of > physical activity promotion (Laforge et al., 1999). Among four studies led in elderly (Acree et al., 2006; Conn et al., 2003; Koltyn, 2001; Stewart et al., 2003), two studies showed a beneficial effect of PA on the physical dimension of QOL in both men (Acree et al., 2006) and women (Acree et al., 2006; Koltyn, 2001) while that of Conn et al., also involving women, found no association (Conn et al., 2003). The study of Stewart et al. confirmed this last tendency in both men and women (Stewart et al., 2003). The results of these studies showed that a high level of PA was associated with an improvement of the two main dimensions (physical, mental) of QOL, with a more marked
1787
1788
104
Leisure-Time Physical Activity and Quality of Life
effect on the physical one, in both men and women, except in elderly. However, the crosssectional approach of these studies does not allow us to show how these relations evolve in time.
3
Longitudinal Studies (> Table 104-3)
Five longitudinal studies bring us additional information (Chen et al., 2005b; Lee and Russell, 2003; Tessier et al., 2007; Wendel-Vos et al., 2004; Wolin et al., 2007). These studies concerned essentially an adult population. . Table 104-3 Significant relationship between physical activity and quality of life in longitudinal studies Quality of life Mental component Longitudinal studies
Men
Physical component
Women
Men
Women
Children Chen et al. (2005b)
X
X
X
X
Wendel-Vos et al. (2004)
X
NS
NS
NS
Tessier et al. (2007)
X
X
NS
NS
Wolin et al. (2007)
–
X
–
X
–
X
–
–
Adults
Elderly Lee and Russell (2003)
X beneficial effect of physical activity on quality of life; NS no statistically significant effect of physical activity on quality of life; – not studied
3.1
In Children
Only one study was interested in a population of 9- to 10-year-old children (n = 7,794) (Chen et al., 2005b). The aim was to examine associations between the lifestyle, including PA, and QOL (COOP) during a 3-year period. The indicator used in this study was PA frequency (very often, often, rarely, almost never) (Chen et al., 2003). The results showed that, compared with the children who remained in the category often, those who crossed into the category often to rarely (OR (95%): 2.10 (1.84–2.39)) or who remained in the almost never category (2.21 (1.88–2.59)) have a poor level of QOL, after adjustment on gender, age and body mass index.
3.2
In Adults
While the results of the cross-sectional analyses put in evidence an association between the level of PA and the physical component of QOL whatever gender, the longitudinal analyses
Leisure-Time Physical Activity and Quality of Life
104
showed an improvement in the mental component of QOL during a 5 years follow-up in men aged 20–59 years old (n = 1,871), not significant in women (Wendel-Vos et al., 2004). In this study, PA was measured using the questionnaire from the European Prospective Investigation into Cancer and Nutrition (EPIC) study (47) and QOL was measured using the SF-36 (Dutch version). At inclusion, PA (hour/week) was associated with three dimensions of QOL (physical activity, perceived health, vitality) and the physical component, in both men and women, after adjustment on age and educational level. Significant tendencies were also found for the mental health dimension in men and for limitations due to physical problems and social functioning dimensions in women. After 5 years, the perceived health and vitality dimensions were still significant in both gender and additional associations were found in women (physical activity, limitations due to the physical and physical component). At two time of measure, no significant association was observed with any dimension of QOL using the total indicator of leisure PA. In the longitudinal analysis, the variation of the level of at least moderate intensity leisure PA was associated with a variation of the social functioning dimension in both genders. An increase of 1h/week of leisure PA of at least moderate intensity was associated with a 0.38 increase of the score of this dimension in men and 0.37 points in women (after adjustment on age, educational level and mean values of QOL and PA measures). Furthermore, in men, this variation was also associated with the physical component. The changes of PA level of intensity at least moderate and QOL observed during 5 years of follow-up were weak and we observed a tendency to a QOL decrease in both genders. If we consider the global indicator of leisure PA, only the mental component was associated with the level of PA in men. Different results were found in a longitudinal study over 3 years (Tessier et al., 2007), conducted in 3 891 adults, with a greater improvement of the mental component observed in women. If it is well known that the level of QOL is weaker in women than men, and that this one declines with age, we can also confirm that it can be maintained or its decline slowed down with advancing age in subjects who report a higher level of leisure PA (measured with the MAQ). However, the changes of QOL found during this period of 3 years remain modest. Besides, the subjects (n = 138) who did not declare PA during leisure at two time of measure had a baseline level of QOL (SF-36) weaker at first, and no increase was observed after 3-year. The results of this study suggest that an increase in the level of leisure PA can improve more particularly the mental component of QOL. This relation, more significant in women, is particularly interesting because women tend to be less active than men during their leisure time, what was also found in this study (4 h/week in women and 5 h/week in men). At inclusion, the level of PA (hour/week) was associated with six dimensions of QOL (physical activity, mental health, limitations due to the mental state, social functioning, vitality, perceived health) and the mental component, in both men and women. A significant tendency was also found for the physical component in men. Similar results were obtained after 3 years of follow-up. Longitudinal analysis showed that in men, an increase of 1 h/week of leisure PA was associated with a 0.17 increase in the vitality score and 0.15 in the mental health. In women, an increase of 1 h/week of leisure PA was associated with an increase of the score of social functioning (+0.40), vitality (+0.39), mental health (+0.28) dimensions and the mental component (+0.23). These relations were adjusted on age, educational level, change of body mass index, tobacco consumption, professional status, time watching television, place of residence and levels of PA and QOL measured at inclusion. These two previous studies assessed simultaneously changes in PA and QOL. Wolin et al. investigated the relationship between long-term change in PA and subsequent change in QOL
1789
1790
104
Leisure-Time Physical Activity and Quality of Life
in adults’ women (Wolin et al., 2007). The results showed that women who increased their level of PA between 1986 and 1996 had higher QOL scores (SF-36) in 1996. It is also important to note that women who maintained their PA level during the period did not experienced the increases in QOL observed in women who increased their PA level.
3.3
In Elderly
In women over 70 years, PA practice was associated with a better QOL measured by the mental dimensions of the SF-36 (vitality, social functioning, social functioning, mental health, limitations due to mental state) (Lee and Russell, 2003). Cross-sectional analysis of the data (n = 10,063) showed that more the level of PA was more the scores of QOL were, even after adjustment on the physical component score (SF-36), marital status, BMI and life events during the last 12 months. The longitudinal analysis (n = 6,472) showed a similar tendency although the effects were weaker. The subjects were classified in group groups according to their level of PA measured in 1996 and 1999: sedentary (little or no PA at two time), commitment in a PA practice (little or not in time 1 and weak, average or high at time 2), stopping PA (weak, average or high at time 1 and little or not at time 2), maintenance (weak, average or high at two time). Globally, the scores decreased but the women who stopped PA between the two times of measures have more negative changes in QOL than sedentary women, while a lesser decline was observed in women who maintained their PA or made a commitment in a PA practice, these last ones having the greatest benefit.
3.4
Summary of Longitudinal Studies
A single study led in children was identified (Chen et al., 2005b) and its results confirmed those observed in the cross-sectional study (Chen et al., 2005a): a beneficial effect of PA on QOL during a follow-up over 3 years. In adults, there is one published study (Wendel-Vos et al., 2004) but the cross-sectional results did not show any association between the total level of PA during leisure and QOL, which was found in a recent study (Tessier et al., 2007), but an association with the level of PA of moderate intensity and QOL. The longitudinal analyses led in the study of Wendel-Vos et al. showed an association between the physical component and the level of PA of moderate intensity and an association between the mental component and the total level of PA in men (followed over 5 years) (Wendel-Vos et al., 2004). The association between the total PA and the mental component of QOL was found in the study of Tessier et al. but in a stronger way in women (Tessier et al., 2007; Wolin et al., 2007). A study led in old women (Lee and Russell, 2003), but concerning only the mental dimension of QOL, put in evidence an association between the level of PA and QOL in the cross-sectional analysis, confirmed in the longitudinal analysis (followed over 3 years).
3.5
Summary of Observation Studies
While QOL was measured very frequently with the SF-36, PA was measured using varied questionnaires. In spite of the use of various measuring instruments of PA and QOL, and the
Leisure-Time Physical Activity and Quality of Life
104
adjustment on a number more or less important of determinants of QOL, a regular practice of PA is globally associated with a better QOL rather in the physical domains with a crosssectional approach and in the mental domains with a longitudinal approach. Most of these studies were interested in LTPA (> Table 104-4) (Tessier et al., 2007; Vuillemin et al., 2005; Wendel-Vos et al., 2004; Wolin et al., 2007) which remains most easily modifiable compared with PA at work. Indeed, PA at work may have not the same benefit as PA during leisure but it remains to demonstrate on QOL. . Table 104-4 Significant relationship between leisure-time physical activity and quality of life measured using SF-36 Quality of life Mental component Men Children
Physical component
Women
Men
Women
–
–
–
–
X
X
X
X
Wendel-Vos et al. (2004)
X
NS
NS
NS
Tessier et al. (2007)
X
X
NS
NS
Wolin et al. (2007)
–
X
–
X
Elderly
–
–
–
–
Adults Cross-sectional study Vuillemin et al. (2005) Longitudinal studies
X beneficial effect of physical activity on quality of life; NS no statistically significant effect of physical activity on quality of life; – not studied
4
Intervention Studies (> Table 104-5)
Even if the results of intervention studies are more contradictory, regular PA had benefits on QOL, what was more widely demonstrated in elderly (Ellingson and Conn, 2000; Rejeski et al., 1996, Rajeski and Mihalko, 2001; Spirduso and Cronin, 2001).
5
Follow-Up of Advice Given by a General Practitioner and/or a Specialist of PA
Halbert et al. showed a decline of QOL (SF-36) after 1 year in sedentary subjects aged 60 years and over who followed the recommendations for PA delivered by a specialist of PA (Halbert et al., 2000). More exactly, the scores of QOL decreased simultaneously in the intervention group (n = 149) and in the control group (n = 150) but a significantly more important decrease was found in women in the intervention group for the limitations due to the mental state, limitations due to physical problems, and social functioning dimensions.
1791
1792
104
Leisure-Time Physical Activity and Quality of Life
. Table 104-5 Significant relationship between physical activity and quality of life in intervention studies Quality of life Mental component Intervention studies Children
Men
Physical component
Women
Men
Women
–
–
–
–
NS
–
NS
–
Adults Hellenius et al. (1995) Sorensen et al. (1999)
X
X
NS
NS
Elley et al. (2003)
X
X
X
X
Lee et al. (2004)
X
X
X
X
Lindh-Astrand et al. (2004)
–
–
–
X
Saavedra et al. (2007)
–
X
–
X
Elderly Jette et al. (1996)
X
NS
–
–
Damush et al. (1999)
–
X
–
X
Whitlach and Adema (1996) NS
NS
NS
NS
Stewart et al. (1997)
NS
NS
NS
NS
Fabre et al. (1999)
NS
X
NS
X
Halbert et al. 2000)
NS
NS
NS
NS
King et al. (2000)
X
X
X
X
Li et al. (2001a,b, 2002)
–
–
X
X
Hopman-Rock and Whesthoff (2002)
X
X
X
X
Fisher and Li (2004)
X
X
X
X
Stiggelbout et al. (2004)
NS
NS
NS
NS
X beneficial effect of physical activity on quality of life; NS no statistically significant effect of physical activity on quality of life; – not studied
On the contrary, a QOL improvement after 1 year (SF-36) was observed in sedentary subjects (level of PA lower than the recommendations: 30 minutes of moderate or intense PA at least 5 days a week) aged 40–79 years old who followed the recommendations for PA delivered first by their general practitioner and relieved by a PA professional (Elley et al., 2003). More exactly, QOL scores increased simultaneously in the intervention group (n = 451) and in the control group (n = 427) but a significantly more important increase was found in the intervention group for the perceived health, limitations due to physical problems, vitality and physical pains dimensions (+2.30 to 7.24 points).
6
The Effects of a Resistance Training Program on Quality of Life
Jette et al. showed that a resistance training improves the psychological well-being and the state of perceived health in men living in a independent way after 12–15 weeks, but this result was
Leisure-Time Physical Activity and Quality of Life
104
not found in women (n = 102) (Jette et al., 1996). Damush et al. showed that a training of 8 weeks with two sessions of 45 min a week was not sufficient to improve QOL in a significant way in old women (n = 33) compared with a control group (n = 29) (Damush et al., 1999).
7
The Effects of an Endurance Exercises Program on Quality of Life
PA contributes to social activities by improving the physical and the cognitive function of the subjects, and leads in better one self-esteem by establishing relations with others. This tendency was partially confirmed in a study which showed an improvement in self-esteem of subjects having followed an exercises program of 6 months but no other dimension of the SF-36 was improved (Stewart et al., 1997). A significant improvement of the physical dimension of QOL was observed in elderly following training programs (aerobic or aerobic and mental combined) during 2 months (1 h/ week) (Fabre et al., 1999). The authors did not found significant change in the mental dimension of QOL whatever the training program (aerobic, mental, aerobic and mental).
8
The Effects of a Specific Exercises Program on Quality of Life
King et al. studied two groups of subjects aged 65 and over, reporting no more than 2 PA sessions per week. Each group participated in a specific exercises program (1 h, twice a week during 1 year). A group followed a program combining endurance and strength (Fit & Firm) (n = 50) while the other group followed a stretching and flexibility program (Stretch & Flex) (n = 46). The results showed a more important improvement of the physical health index, the physical pains and emotion QOL scores in the group Stretch & Flex compared with the group Fit & Firm, in whom only an improvement of the energy/fatigue score was found (King et al., 2000). Li and al. were interested in the effects of a Tai Chi program (60 min per session, twice a week, during 6 months) on the perception of physical function measured with the SF-20 (Li et al., 2001a,b; Li et al., 2002). This study, involving 49 subjects aged 65 years and older, showed a major effect of the program on QOL (physical function) and this effect was all the more important as the initial score of physical function was low. In the intervention group (n = 49) the average score of the physical function dimension was 69.63 26.02 at the beginning of the study and 86.10 15.22 after 6 months. A study conducted in menopausal women showed an improvement in QOL (Symptom Check List-90, SCL-90) following an aerobic exercises program of moderate intensity (two supervised 1 h sessions and one free session per week, during 12 weeks (Lindh-Astrand et al., 2004). An intervention study (Stiggelbout et al., 2004) conducted in men and women (71 4.1 years) aged 65–80 years old showed that the program ‘‘Senior More Exercise for (MBvO in Dutch),’’ implying one session (MBvO1, n = 98) or two sessions (MBvO2, n = 53) of 45 min per week during 10 weeks, was not sufficient to improve QOL. QOL was measured using three instruments: Vitality Plus Scale (VPS) (66), TNO Leiden Academic Hospital Adults Quality of Life questionnaire (TAAQOL) and SF-36. PA was measured with a specific questionnaire for elderly (Voorrips et al., 1991). An improvement of QOL was observed in the MBvO2 group for
1793
1794
104
Leisure-Time Physical Activity and Quality of Life
subjects having reported a low level of PA at the beginning of the study. A similar study, in terms of exercises programs and population, showed an improvement of the vitality and perceived health dimensions (SF-36) after 6 weeks of intervention but the level of PA of this population at baseline was lower and the subjects pursued the exercises at home 3–4 times per week (Hopman-Rock and Whesthoff, 2002). Saavedra et al. examined the influence of a medium-impact aquaerobic program (2 weekly sessions of 60 min during 8 months) on QOL level of a group of middle-aged (43.1 9.7 years) healthy women (n = 20). Significant improvements were found for 6 of 8 domains of the SF-36: physical function, role-physical, bodily pain, vitality, social function, and mental health (+12.8 to 32%, p = 0.034–0.001) (Saavedra et al., 2007). Mind-body practices such as yoga are widely popular and practiced during leisure time. Changes in QOL were associated with 3 months of mind-body training as practiced in community-based settings. QOL scores improved in all domains (P < .0001), including a change of +15.5 (21) in the mental health domain (Lee et al., 2004).
9
The Effects of a Walking Program on Quality of Life
Fisher et al. showed the beneficial effect of a supervised walking program (1 h, 3 times per week during 6 months) realized in the neighborhood on QOL of subjects aged 65 and older (n = 582) initially sedentary or inactive (not having participated in PA during the previous 30 days) (Fisher and Li, 2004). QOL was appreciated with SF-12 (physical and mental components) and a satisfaction scale (Satisfaction Life Scale, SWLS) taking into account several covariables (age, sex, educational level, ethnic group, incomes).
10
The Effects of an Exercises Program and a Diet on Quality of Life
The two published studies (Hellenius et al., 1995; Sorensen et al., 1999) did not put in evidence a significant difference of QOL in at risk populations of cardiovascular diseases. However, in the study of Sorensen et al. the mental health dimension of QOL was investigated apart and showed significant results. This study involved 21 women and 198 men from 41 to 50 years old either in an exercises program (n = 48), or in a dietary advice group (n = 52), or in both (n = 65), and compared with a control group (n = 43). The exercises program consisted of a 1 h aerobic physical exercise (two supervised sessions and one session at home per week), during 1 year. QOL was measured using three instruments and mental health, considered in this study as a psychological measure, was appreciated using the General Health Questionnaire (GHQ). If no change of QOL was observed whatever the instrument, a change was observed in mental health measured with the GHQ (total score, skill, difficulty coping) in the group having followed the program. The subjects who participated in the exercises program (>70%, n = 54) improved significantly more their scores measured by the GHQ (total, anxiety, skill, difficulty coping) that the subjects with a weak participation ( Table 104-6) but used different physical activity measure. One of the specific characteristic of PA is its capacity to influence various physiological functions but also various dimensions of QOL simultaneously; but few studies are interested at the same time in the physiological effects of PA and in its effects on QOL. Therefore, we do not know if PA has a direct effect on QOL independent from the improvement of the physiological parameters. However, the study of the relation between PA and QOL is interesting because of the potential action of PA simultaneously on health and well-being through the improvement of QOL. More studies are necessary to determine the minimal and maximal threshold above and beyond which PA can have no effect or negative effects on QOL. Further studies are necessary to bring more precise knowledge of the modalities of PA practice (type, frequency, duration,
1795
1796
104
Leisure-Time Physical Activity and Quality of Life
. Table 104-6 Significant relationship between physical activity and quality of life measured using SF-36 questionnaire Quality of life Mental component Men Children
–
Physical component
Women –
Men –
Women –
Adults Cross-sectional study Laforge et al. (1999)
X
X
X
X
Daskapan et al. (2005)
X
X
X
X
Vuillemin et al. (2005)
X
X
X
X
Wendel-Vos et al. (2004)
X
NS
NS
NS
Tessier et al. (2007)
X
X
NS
NS
Wolin et al. (2007)
–
X
–
X
Longitudinal studies
Intervention studies Sorensen et al. (1999)
X
X
NS
NS
Elley et al. (2003)
X
X
X
X
Lee et al. (2004)
X
X
X
X
Saavedra et al. (2007)
6
X
6
X
Elderly Cross-sectional study Stewart et al. (2003)
NS
NS
NS
NS
Acree et al. (2006)
NS
NS
X
X
–
X
–
–
Longitudinal studies Lee and Russell (2003) Intervention studies Halbert et al. (2000)
NS
NS
NS
NS
Stiggelbout et al. (2004)
NS
NS
NS
NS
X beneficial effect of physical activity on quality of life; NS no statistically significant effect of physical activity on quality of life; – not studied
intensity,) which would allow to clarify the current recommendations for PA and to clarify the public health messages. It implies to develop the use of the instruments of QOL as judgment criteria of the effects of PA and to show the interest of the evaluation of QOL to better understand the effects of PA. It is mostly considered that health behavior determines QOL. However, the reverse is also possible. Globally, it has been shown that the increase of PA level comes along with an improvement of QOL. However, we can also consider that individuals who have a better QOL are more motivated to practice or increase their level of PA. It is reasonable to
Leisure-Time Physical Activity and Quality of Life
104
consider that the relations between QOL and health behavior can be established in both directions. In term of public health, improving one of these two results in a global improvement of health.
Summary Points Physical activity is a modifiable behavior and its regular practice is globally associated with a better quality of life.
Regular physical activity improves quality of life by enhancing psychological well-being and by improving physical functioning.
Incorporating moderate-intensity physical activity in daily life improve quality of life. The most promising results were obtained from programs facilitating a regular practice equivalent to the current public health recommendations for physical activity.
Physical activity contributes to social activities by improving the physical and the cognitive function and leads in better one self-esteem by establishing relations with others.
References Acree LS, Longfors J, Fjeldstad AS, Fjeldstad C, Schank B, Nickel K, Montgomery PS, Gardner AW. (2006). Health Qual Life Outcomes. 4: 37. American College of Sports Medicine Position Stand. (1998). Med Sci Sports Exerc. 30: 975–991. Bize R, Johnson JA, Plotnikoff R. (2007). Prev Med. 45: 401–415. Blair SN, Haskell WL, Ho P, Paffenbarger R. Jr., Vranizan KM, Farquhar JW, Wood PD. (1985). Am J Epidemiol. 122: 794–804. Brown DW, Brown DR, Heath GW, Balluz L, Giles WH, Ford ES, Mokdad A. (2004). Med Sci Sports Exerc. 36: 890–896. Brown DW, Balluz LS, Heath GW, Moriarty DG, Ford ES, Giles WH, Mokdad AH. (2003). Prev Med. 37: 520–528. Chen X, Sekine M, Hamanishi S, Wang H, Gaina A, Yamagami T, Kagamimori S. (2003). J Epidemiol. 13: 278–287. Chen X, Sekine M, Hamanishi S, Wang H, Gaina A, Yamagami T, Kagamimori S. (2005a). Prev Med: . 40: 668–678. Chen X, Sekine M, Hamanishi S, Yamagami T, Kagamimori S. (2005b). Child Care Health Dev. 31: 433–439. Conn VS, Minor MA, Burks KJ. (2003). J Community Health Nurs. 20: 197–208. Damush TM, Damush JG Jr. (1999). Gerontologist. 39: 705–710.
Daskapan A, Tuzun EH, Eker L. (2005). Saudi Med J. 26: 1026–1028. DiPietro L, Caspersen CJ, Ostfeld AM, Nadel ER. (1993). Med Sci Sports Exerc. 25: 628–642. Elley CR, Kerse N, Arroll B, Robinson E. (2003). BMJ. 326: 793. Ellingson T, Conn V. (2000). J Gerontol Nurs. 26: 17–25. Fabre C, Masse´-Biron J, Chamari K, Varray A, Mucci P, Pre´faut C. (1999). Arch Gerontol Geriatr. 28: 9–22. Fisher KJ, Li F. (2004). Ann Behav. Med. 28: 186–194. Halbert JA, Silagy CA, Finucane PM, Withers RT, Hamdorf PA. (2000). Med J Aust. 173: 84–87. Hellenius ML, Dahlof C, Aberg H, Krakau I, De Faire U. (1995). Qual Life Res. 4: 13–20. Hopman-Rock M, Whesthoff M. (2002). J Aging Phys Act. 10: 363–380. Jette AM, Harris BA, Sleeper L, Lachman ME, Heislein D, Giorgetti M, Levenson C. (1996). J Am Geriatr Soc. 44: 644–649. King AC, Pruitt LA, Phillips W, Oka R, Rodenburg A, Haskell WL. (2000). J Gerontol A Biol Sci Med Sci. 55: M74–M83. Koltyn KF. (2001). Womens Health Issues. 11: 471–480. Kull M. (2002). Scand J Med Sci Sports. 12: 241–247. Laforge RG, Rossi JS, Prochaska JO, Velicer WF, Levesque DA, McHorney CA. (1999). Prev Med. 28: 349–360.
1797
1798
104
Leisure-Time Physical Activity and Quality of Life
Lee C, Russell A. (2003). J Psychosom Res. 54: 155–160. Lee SW, Mancuso CA, Charlson ME. (2004). J Gen Intern Med. 19: 760–765. Li F, Fisher KJ, Harmer P, McAuley E. (2002). Am J Prev Med. 23: 92–97. Li F, Harmer P, McAuley E, Duncan TE, Duncan SC, Chaumeton N, Fisher KJ. (2001a). Ann Behav Med. 23: 139–146. Li F, Harmer P, McAuley E, Fisher KJ, Duncan TE, Duncan SC. (2001b). Prev Sci. 2: 229–239. Lindh-Astrand L, Nedstrand E, Wyon Y, Hammar M. (2004). Maturitas. 48: 97–105. Paffenbarger RS Jr., Blair SN, Lee IM, Hyde RT. (1993). Med Sci Sports Exerc. 25: 60–70. Pate RR, Pratt M, Blair SN, Haskell WL, Macera CA, Bouchard C, Buchner D, Ettinger W, Heath GW, King AC. (1995). JAMA. 273: 402–407. Reed GR, Velicer WF, Prochaska JO, Rossi JS, Marcus BH. (1997). Am J Health Promot. 12: 57–66. Rejeski WJ, Brawley LR, Shumaker SA. (1996). Exerc Sport Sci Rev. 24: 71–108. Rejeski WJ, Mihalko SL. (2001). J Gerontol A Biol Sci Med Sci. 56(Spec No 2): 23–35. Saavedra JM, De La CE, Escalante Y, Rodriguez FA. (2007). J Sports Med Phys Fitness. 47: 468–474. Shibata A, Oka K, Nakamura Y, Muraoka J. (2007). Health Qual Life Outcomes. 5: 64. Sorensen M, Anderssen S, Hjerman I, Holme I, Ursin H. (1999). J Sports Sci. 17: 369–377. Spirduso WW, Cronin DL. (2001). Med Sci Sports Exerc. 33: S598–S608.
Stewart AL, Mills KM, Sepsis PG, King AC, McLellan BY, Roitz K, Ritter PL. (1997). Ann Behav Med. 19: 353–361. Stewart KJ, Turner KL, Bacher AC, DeRegis JR, Sung J, Tayback M, Ouyang P. (2003). J Cardiopulm Rehabil. 23: 115–121. Stiggelbout M, Popkema DY, Hopman-Rock M, de Greef M, Van Mechelen W. (2004). J Epidemiol Community Health. 58: 83–88. Tessier S, Vuillemin A, Bertrais S, Boini S, Le Bihan E, Oppert JM, Hercberg S, Guillemin F, Briancon S. (2007). Prev Med. 44: 202–208. Voorrips LE, Ravelli AC, Dongelmans PC, Deurenberg P, van Staveren W. (1991). Med Sci Sports Exerc. 23: 974–979. Vuillemin A, Boini S, Bertrais S, Tessier S, Oppert JM, Hercberg S, Guillemin F, Briancon S. (2005). Prev Med. 41: 562–569. Vuillemin A, Oppert JM, Guillemin F, Essermeant L, Fontvieille AM, Galan P, Kriska AM, Hercberg S. (2000). Med Sci Sports Exerc. 32: 1119–1124. Wendel-Vos GC, Schuit AJ, Tijhuis MA, Kromhout D. (2004). Qual Life Res. 13: 667–677. Whitlatch S, Adema R. (1996). Activities, Adaptation and Aging 20: 75–85. Wolin KY, Glynn RJ, Colditz GA, Lee IM, Kawachi I. (2007). Am J Prev Med. 32: 490–499. Zahran HS, Zack MM, Vernon-Smiley ME, Hertz MF. (2007). J Adolesc Health. 41: 389–397.
105 Spa Therapy and Quality of Life G. Blasche 1
Introduction: An Overview of Spa Therapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1800
2
Therapeutic Pathways of Spa Therapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1802
3 3.1 3.2 3.3 3.4 3.5
Effects on Quality of Life and Well-Being . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1803 Effects of a Change of Location . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1803 Change throughout the Treatment Period . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1803 Short Term (End of Treatment) Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1804 Medium to Long Term Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1806 Cost Efficiency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1807 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1807
#
Springer Science+Business Media LLC 2010 (USA)
1800
105
Spa Therapy and Quality of Life
Abstract: > Spa therapy is a traditional form of treatment aiming at alleviating or preventing chronic disorders such as > rheumatoid arthritis or > osteoarthritis usually associated with an impairment of quality of life. It is based on the use of > natural remedies like thermal, mineral or sea water, mud, or certain climatic factors. These remedies are applied in regular intervals over a time span of 2–4 weeks as baths, drinking cures, packs or as out-door exposure predominantly in an inpatient setting. In addition, other treatments such as massages, exercise and behaviorally oriented methods are applied. It is thought that spa therapy improves quality of life not only by reducing disease burden but also by enhancing physical and psychological health. Disorders commonly treated with spa therapy are low back pain, osteoarthritis, rheumatoid arthritis, > fibromyalgia, > psoriasis and hypertension. An improvement of quality of life is reported for most of these disorders for a time span of up to 9 months. Large though only moderately stable short term improvements occur for mood related variables. The most stable improvements are apparent for physical complaints and functioning. Cost effectiveness has been ascertained for some but not all disorders treated. In conclusion, spa therapy is an effective and highly esteemed treatment improving quality of life predominantly in patients with chronic pain disorders.
1
Introduction: An Overview of Spa Therapy
Spa therapy is a traditional form of treatment aiming at alleviating or preventing chronic disorders usually associated with a moderate to severe impairment of quality of life, such as degenerative or inflammatory arthritis. Spa therapy is based on the use of natural remedies like thermal water, mineral water, sea water, mud, or certain climatic factors. These remedies are applied in regular intervals, e.g., every second day, over a time span of 2–4 weeks as baths, drinking cures, packs or out-door exposure. In addition, other treatments such as massages, exercise and behaviorally oriented methods are applied. Common treatments used in spa therapy are listed in > Table 105-1. Spa therapy, in contrast to > balneotherapy commonly utilizes an inpatient setting. Patients stay at a health resort for the duration of the treatment and receive several treatments per day, allowing for additional leisure time. In many European countries such as Germany, Austria, France and Italy, spa therapy is covered by public health insurance. As spa therapy is considered to be a medical treatment in most European countries, patients receive sick leave for the duration of the treatment. The current policy of health insurance in Austria is to reimburse spa therapy every second year, provided spa therapy is warranted. Historically, spa therapy goes back to the ancient Greek and Roman times. Hippocrates (460–370 BC) introduced bathing together with sweating, walking, and massages as a means to re-establish a proposed balance of fluids (van Tubergen and van der Linden, 2002). Romans took over the Greek tradition and built thermal baths at mineral and thermal springs, often in the vicinity of military posts. Spas were used both for recuperation from disease or injury as well as for recreation. Medical treatment was frequently provided at spas, consisting of applying water to afflicted parts of the body, whole-body water immersion and prescribed drinking of mineral waters. This tradition was continued in middle and new age Europe, with a culmination of spa treatment in the nineteenth and early twentieth centuries, when visiting spas was not only motivated by health care but was also a pastime for the wealthy (van Tubergen and van der Linden, 2002). After the Second World War spa therapy became
Spa Therapy and Quality of Life
105
. Table 105-1 Common treatments used in spa therapy Type of treatment Baths
Treatment Carbonated water baths Sulphur baths Saline or sea water baths Mud baths Other mineral and/or thermal water baths
other natural remedies
Radon bath or inhalation Saline aerosol inhalation Mud packs Climatotherapy
additional treatments
Underwater jet or hose massage Underwater exercise therapy Classical massage Exercise therapy (dry) Physical treatments such as ultrasound Relaxation training Patient education
The table lists common treatments used during spa therapy consisting of ‘‘baths’’, ‘‘other natural remedies’’ and ‘‘additional treatments’’
available to the general public in many European countries as costs were covered by public health insurance. More than any other medical discipline, spa medicine has not only maintained but also elaborated its holistic approach to health and disease to the present day by viewing treatment as a time dependent psycho-physiological process not only curing specific diseases but also enhancing general health. A proponent of this view was Guenther Hildebrandt, a balneologist and physiologist, who understood spa therapy as a means to optimize the > rhythmic functional order (‘‘rhythmische Funktionsordnung’’) of the physiological system (Hildebrandt, 1991). Apart from conducting numerous studies on the effects of spa-therapy on general physiological processes he also co-authored a classic German handbook of spa medicine (Gutenbrunner and Hildebrandt, 1998). Based on these concepts, spa therapy improves quality of life not only by specific disease related effects but also by enhancing physical functioning and psychological well-being. The disorders commonly treated with spa therapy are summarized in > Table 105-2. The majority of these disorders are related to chronic pain of the musculoskeletal system and thus impair quality of life substantially. They are low back pain, a disorder of various etiologies affecting the lower back, osteoarthritis, a degenerative disease of the joints (e.g., hips and knees), rheumatoid arthritis, a chronic systemic disease of the joints marked by inflammatory changes and atrophy, and fibromyalgia, a disorder of unknown etiology characterized by multi-site muscular pain and fatigue. Other disorders treated with spa therapy are psoriasis, an inflammatory skin disease, and essential hypertension.
1801
1802
105
Spa Therapy and Quality of Life
. Table 105-2 Common disorders treated with spa therapy Disorder
Description
Reference
Low back pain
Persistent and often disabling pain located in the lower back
(Pittler et al., 2006)
Osteoarthritis
Degenerative disease of the joints (e.g., (Verhagen et al., 2007) hips and knees)
Rheumatoid arthritis (including ankylosing spondylitis)
A chronic systemic disease of the joints (Hashkes, 2002; van marked by inflammatory changes and Tubergen et al., 2001; atrophy Verhagen et al., 2007)
Fibromyalgia
Disorder of unknown etiology characterized by multi-site muscular pain and fatigue
(Donmez et al., 2005; Zijlstra et al., 2005)
Psoriasis and atopic dermatitis
Inflammatory skin disease
(Even-Paz, 1996; LeauteLabreze et al., 2001; Matz et al., 2003)
Essential hypertension
Persistently increased blood pressure
(Ekmekcioglu et al., 2000; Gutenbrunner et al., 2002)
The table lists disorders commonly treated with spa therapy and provides a brief description of these disorders as well as recent references of studies investigating the effect of spa therapy for people with these disorders
2
Therapeutic Pathways of Spa Therapy
Spa therapy achieves its therapeutic aims via three main pathways (Bender et al., 2005). One pathway is the use of > hydrotherapy. Head out water immersion, i.e., bathing in water, affects an individual both physiologically and psychologically. Physiologically, the buoyancy produced by water leads to a relative decrease in body weight and subsequently a decrease in the tension predominantly of muscles responsible for maintaining the upright position. The increased hydrostatic pressure also brings about changes in the cardiovascular system. It causes venous and lymphatic compression, a subsequent increase of central blood volume, greater venous return and in consequence an increased stroke volume and decreased heart rate. In addition, myocardial efficiency is enhanced (Becker and Cole, 1998). If warm instead of thermo-neural or cool water is used, the increase of peripheral circulation and the decrease of total peripheral resistance lead to a reduction of diastolic blood pressure with parallel increases in heart rate and cardiac output. On the psychological level, the sensation of warm water on the skin and the decrease of body weight lead, partly due to an increased somatosensory awareness, to a decrease of mental activity, pleasant feelings and relaxation (Blasche et al., 2007; Robiner, 1990). The second pathway is balneotherapy, the use of natural remedies with specific physical and/or chemical properties. Carbonated water, for example, increases peripheral perfusion due to the dilatory effects of the carbon dioxide permeating the skin and decreases pain perception (Nishimura et al., 2002). Mud packs and mud baths derive their therapeutic effectiveness through both physical and chemical mechanisms. Peloids, i.e., special kinds of muds used in spa therapy, are said to have an anti-inflammatory effect (Eichler, 1995).
Spa Therapy and Quality of Life
105
In addition, the high viscosity of mud baths permits the use of higher bathing temperatures as heat is transmitted through conduction rather than convection (Beer et al., 2003). It is argued that lasting effects of spa treatment are brought about by the repetitive use of treatments over a time span of several weeks as this leads to adaptive changes in those physiological systems that are stimulated by treatment (Hildebrandt and Gutenbrunner, 1998b). The third pathway is psychological. Spa therapy leads to a decrease of distress and an improvement of mood presumably through factors such as positive bodily experiences, relaxation and a respite from work and other every day demands (Strauss-Blasche et al., 2000a, 2002). According to our own research, spa therapy is usually highly esteemed by patients. In addition, chronobiological factors may come into play through the temporal structure of treatment represented by fixed waking, sleeping and meal times and the variation between treatment and leisure (Agishi and Hildebrandt, 1997). Theoretically, the psychological effects should be more pronounced for resort-based spa therapy than for outpatient balneotherapy, but so far research comparing these two settings has to my knowledge not been conducted. Nevertheless, both the mood enhancing effects of spa treatments as well as the effects of a respite from work have been documented (Blasche et al., 2000b, 2007; Fritz and Sonnentag, 2006). In a so far unpublished study of our group, bathing at 38 Celsius in different types of water is associated with an improvement of well-being and mood as well as a decrease in facial muscular tension and an increase of electrocortical activity in the Alpha band, all indicating physical and mental relaxation.
3
Effects on Quality of Life and Well-Being
3.1
Effects of a Change of Location
Spa therapies are usually associated with a stay at a health resort over a period of 2–4 weeks. Although health resorts are generally arranged to foster recuperation and rest, travelling and adaptation to a new environment is associated with a temporary increase of arousal and stress. An increase of sleep disturbances and nervousness at the beginning of spa treatment has been documented (Hildebrandt and Gutenbrunner, 1998a). These minor disturbances of well-being usually decline within the first few days of therapy. Similar though smaller adaptation effects occur at the end of spa therapy, when people have to travel home and re-engage in their normal life routine (Strauss-Blasche et al., 2004a). These effects are more pronounced for individuals coming home at the beginning of the week than for those coming home close to the weekend, presumably because the weekend provides an opportunity to readjust.
3.2
Change throughout the Treatment Period
Several aspects of well-being and quality of life show a day-to-day change during the course of spa-therapy. Mood, quality of sleep, nervousness and physical complaints have been shown to improve during a three-week resort-based spa-treatment (Hildebrandt and Gutenbrunner, 1998a). > Figure 105-1 illustrates these changes for mood and the restfulness of sleep. The data shown is based on a study with 330 patients (147 females and 183 males,
1803
1804
105
Spa Therapy and Quality of Life
. Figure 105-1 Change of well-being during spa therapy Figure 105-1 Change of mood and the restfulness of sleep over the course of a three week inpatient spa-treatment based on daily diary entries. The figure illustrates mean values and standard errors using previously unpublished data from the study Strauss-Blasche et al. (2004a).
mean age 59.4 10.3 years) with musculoskeletal pain undergoing resort-based spa-therapy in the Austrian spa Bad Gleichenberg. Details on the study are described elsewhere (StraussBlasche et al., 2004a). Due to the reaction of patients to treatment, the increase of well-being is not linear but shows some stagnation in the second week of treatment. These reactive periods are commonly observed in resort-based spa-therapy and have been described in detail (Hildebrandt and Gutenbrunner, 1998a; Hildebrandt and Lammert, 1986). Day-to-day increases of well-being also have been observed for stays at a holiday-resort, though these increases lack the reactive periods of spa-therapy (Strauss-Blasche et al., 2004b).
3.3
Short Term (End of Treatment) Effects
Short term improvements of various aspects of quality of life attributable to spa therapy have been found for a number of disorders such as low back pain (Constant et al., 1998; Guillemin et al., 1994), chronic musculoskeletal pain (Guillemin et al., 2001; Strauss-Blasche et al., 2000a), > gonarthrosis (Wigler et al., 1995; Yilmaz et al., 2004) and breast cancer (StraussBlasche et al., 2005). For example, a significant improvement of physical, mental and general health including self-esteem, anxiety, depression and pain has been reported for low back pain patients receiving a 3 week spa therapy in France (Constant et al., 1998). An example of the extent of pre- to post changes in quality of life is provided in > Table 105-3. Not all aspects of well-being show the same immediate response to resort-based spa therapy.
105
Spa Therapy and Quality of Life
. Table 105-3 Short and medium term effect of spa therapy Variable
Short term effect size
Medium term effect size
Stability (effect size ratio)
Study 1: spa-therapy for musculoskeletal pain (Strauss-Blasche et al., 2000a) Negative mood
1,21
0,54
0,44
Positive mood
0,96
0,52
0,54
Pain intensity
0,73
0,44
0,61
Health satisfaction
0,72
0,41
0,56
Complaints
0,70
0,42
0,60
Study 2: spa therapy for breast cancer (Strauss-Blasche et al., 2005) Emotional Functioning
1,01
0,38
0,38
Global health status
0,62
0,30
0,48
Anxiety
0,50
0,15
0,30
Depression
0,48
0,28
0,58
Pain
0,47
0,42
0,89
Fatigue
0,38
0,38
1,00
Physical functioning
0,21
0,17
0,81
> Table
105-3 illustrates the short and medium term effect of spa therapy on selected variables of quality of life. The effects are quantified as effect sizes (es), es = .2 illustrating a small, es = .5 illustrating a medium and es = .8 illustrating a large effect. Short (at the end of treatment) and medium term (6 weeks after treatment for study 1 and 6 months after treatment for study 2) effects are listed. The numbers provided are based on the two studies indicated
Mood related variables such as positive and negative mood and emotional functioning show greater improvements than variables related to physical functioning and complaints based on their > effect sizes. At the same time, the former aspects of quality of life are less stable and show a greater decline to follow-up than the latter. The large but temporary increase of psychological well-being during spa therapy is presumably a result of resort based life providing not only a respite from every day demands but also the pleasure of bathing and leisure time (Bender et al., 2005; Blasche et al., 2007). Another unique property of spa-therapy is the > homogenization of groups in respect to outcome measures. This phenomenon describes the reduction of group standard deviations in outcome variables from the beginning to the end of treatment, thus leading to greater similarities between patients at the end of spa therapy (Gutenbrunner and Ruppel, 1992; Hildebrandt and Gutenbrunner, 1998a). This effect has been observed for variables such as blood pressure, hand temperature and nervousness (Feyertag et al., 1995; Gutenbrunner and Ruppel, 1992). In a so far unpublished study we found that subjects with high and low levels of psychosocial or occupational stress initially associated with large differences in distress and well-being show a significant convergence in these variables toward the end of treatment, thus signifying homogenization. Put in simple terms, spa therapy at least temporarily seems to ‘‘wash out’’ those differences between people which are related to stress or disease.
1805
1806
105 3.4
Spa Therapy and Quality of Life
Medium to Long Term Effects
Though the amount of available research is limited, spa therapy generally is found to have a medium to long term positive effect on physical and psychological aspects of quality of life for disorders such as low back pain, osteoarthritis, rheumatoid arthritis, fibromyalgia and psoriasis. The duration of these effects is summarized in > Table 105-4. It should, however, be noted that the evaluated spa therapies vary in regard to duration (10–21 days), setting (in- or outpatient) as well as the types of treatment provided. In addition the studies themselves are based on different research designs and are of different methodological quality. These disparities may account for some of the differences found. Spa therapy has a beneficial effect on low back pain in regard to pain intensity, functioning and psychological well-being lasting at least for 3 months (Constant et al., 1998). The effect on functioning was found to last at least for 6 months, the effect on pain at least for 9 months (Constant et al., 1995; Guillemin et al., 1994). Patients with > ankylosing spondylitis were found to improve quality of life, i.e., pain, functioning and overall wellbeing, for a minimum of three
. Table 105-4 Duration of spa-therapy effects on quality of life Duration of spa therapy (days) Low back pain
Mood and psychological well-being
Assessment of effects at Pain Functioning
Reference
21
3 months
yes
yes
yes
(Constant et al., 1998)
21
6 months
–
yes
–
(Constant et al., 1995)
21
9 months
yes
no
–
(Guillemin et al., 1994)
14
3 months
yes
yes
–
(Codish et al., 2005)
14
3 months
–
–
yes
(Tishler et al., 1995)
Osteoarthritis 14
3 months
yes
yes
yes
(Evcik et al., 2007)
6 months (average)
–
yes
–
(Bellometti et al., 2007)
Fibromyalgia 10
3 months
yes
yes
no
(Neumann et al., 2001)
14
6 months
no
yes
–
(Donmez et al., 2005)
19
3 months
yes
yes
yes
6 months
no
no
yes
(Zijlstra et al., 2005)
Rheumatoid arthritis
?
> Table 105-4 summarizes the duration of spa-therapy effects on 3 domains of quality of life for common disorders
treated with spa therapy. The duration of spa therapy is indicated. ‘‘yes’’ refers to a positive effect at the time indicated, ‘‘no’’ to a negative effect ‘‘,-‘‘indicates that the specified effect was not studied
Spa Therapy and Quality of Life
105
months after treatment (Codish et al., 2005; Tishler et al., 1995). Similar findings are reported for knee osteoarthritis, where beneficial effects were found in all three domains of quality of life at 3 months and a positive effect on functioning at 6 months (Bellometti et al., 2007; Evcik et al., 2007). Studies on the effect of spa therapy on fibromyalgia are characterized by partially conflicting findings. A positive effect on pain is found at 3 months but not at 6 months. However, inconsistencies are apparent in regard to the effect on functioning and mood (Donmez et al., 2005; Neumann et al., 2001; Zijlstra et al., 2005). These differences may be due to the type of spa therapy provided, as the study observing a sustained improvement of psychological well-being used exercise and patient education in addition to balneotherapy, whereas the other study did not.
3.5
Cost Efficiency
In a French study on the cost effectiveness of spa therapy for rheumatic disease and musculoskeletal disorders, a decrease in health care utilization during the year after spa therapy was found solely in those patients who never had had spa treatment before. No difference in health care utilization was found in patients who had received spa therapy on at least one other occasion (Allard et al., 1998). In a Dutch study on the cost effectiveness of spa therapy for patients with ankylosing spondylitis, a reduction of health care costs was noted in regard to visits to health care professionals, physiotherapy, and use of medication in the first 40 weeks after treatment (Van Tubergen et al., 2002). Another Dutch study dealing with fibromyalgia failed to find any reductions of cost in the first year after treatment despite a temporary improvement of quality of life (Zijlstra et al., 2007). A study on the effect of spa therapy on primary and secondary osteoarthritis in Italy showed that patients undergoing spa therapy showed a significant reduction in the use of additional medical treatments such as hospital admissions and physical and pharmacological therapies as well as in the absence from work when comparing the year before and the year after treatment (Fioravanti et al., 2003). An Austrian study found a decrease of sick leave when comparing the year before to the year after spa therapy. The mean duration of sick leave before spa therapy was 37.9 days per patient, in the year after spa therapy 33.3 days. The reduction was predominantly related to a decrease of musculoskeletal problems (Meggeneder, 1998). In sum, spa therapy reduces health care utilization and sick leave for some but not all of the evaluated disorders.
Summary Points Spa therapy leads to an improvement of quality of life for disorders such as low back pain, inflammatory arthritis, degenerative arthritis, fibromyalgia and psoriasis.
Generally, effects last for 3–6 months, but for some disorders and some aspects of quality of life durations up to 9 months are observed.
On a short term basis, the largest improvements occur for mood, the most stable long term improvements occur for physical complaints and functioning.
Due to improvement, the similarity of patients in regard to quality of life increases during the course of spa treatment.
Minor short term disturbances of well-being occur at the beginning and end of treatment due to the change of environment associated with staying at a resort.
Spa therapy is highly esteemed by patients.
1807
1808
105
Spa Therapy and Quality of Life
References Agishi Y, Hildebrandt G. (1997). Chronobiologische Gesichtspunkte zur Physikalischen Therapie und Kurortebehandlung,. Wissenschaftliche Schriftenreihe des Institus fu¨r Rehabilitationsmedizin und Balneologie Bad Wildungen, Wissenschaftliche Schriftenreihe des Institus fu¨r Rehabilitationsmedizin und Balneologie Bad Wildungen. Allard P, Deligne J, Van Bockstael V, Duquesnoy B. (1998). Rev Rhum Engl Ed. 65: 173–180. Becker BE, Cole AJ. (1998). Aquatic rehabilitation. In: DeLisa JA, Gans BM (ed) Rehabilitation Medicine: Principles and Practice. 3rd ed. Lippincott-Raven, Philadelphia, pp 997–901. Beer AM, Grozeva A, Sagorchev P, Lukanov J. (2003). Biomed Tech (Berl). 48: 301–305. Bellometti S, Gallotti C, Pacileo G, Rota A, Tenconi MT. (2007). J Prev Med Hyg. 48: 1–4. Bender T, Karagulle Z, Balint GP, Gutenbrunner C, Balint PV, Sukenik S. (2005). Rheumatol Int. 25: 220–224. Blasche G, Melchart H, Leitner D, Marktl W. (2007). Forsch Komplement Med. 14: 274–280. Codish S, Dobrovinsky S, Abu Shakra M, Flusser D, Sukenik S. (2005). Isr Med Assoc J. 7: 443–446. Constant F, Collin JF, Guillemin F, Boulange M. (1995). J Rheumatol. 22: 1315–1320. Constant F, Guillemin F, Collin JF, Boulange M. (1998). Med Care. 36: 1309–1314. Donmez A, Karagulle MZ, Tercan N, Dinler M, Issever H, Karagulle M, Turan M. (2005). Rheumatol Int 26: 168–172. Eichler I. (1995). Pharmakologie der Peloidinhaltsstoffe [Pharmacology of peloids]. In: Marktl W (ed) Theorie and Praxis der Peloidtherapie. Facultas, Vienna, pp 52–64. Ekmekcioglu C, Strauss-Blasche G, Feyertag J, Klammer N, Marktl W. (2000). Altern Ther Health Med. 6: 46–53. Evcik D, Kavuncu V, Yeter A, Yigit I. (2007). Joint Bone Spine 74: 60–65. Even-Paz Z. (1996). Isr J Med Sci 32 Suppl: S11–S15. Feyertag J, Strauss-Blasche G, Boettcher E, Marktl W. (1995). Phys Rehab Kur Med. 5: 15–18. Fioravanti A, Valenti M, Altobelli E, Di Orio F, Nappi G, Crisanti A, Cantarini L, Marcolongo R. (2003). Panminerva Med. 45: 211–217. Fritz C, Sonnentag S. (2006). J Appl Psychol. 91: 936–945. Guillemin F, Constant F, Collin JF, Boulange M. (1994). Br J Rheumatol. 33: 148–151. Guillemin F, Virion JM, Escudier P, De Talance N, Weryha G. (2001). Joint Bone Spine. 68: 499–503.
Gutenbrunner C, Ruppel K. (1992). Phys Rehab Kur Med. 2: 58–64. Gutenbrunner C, Hildebrandt G. (1998). Handbuch der Balneologie and medizinischen Klimatologie. Springer, SpringerGutenbrunner C, Schreiber C, Beck K, Walter N, Ehlebracht-Ko¨nig I, v Pezold E, Gehrke A, Kniesch K, Thon B, Candir F. (2002). Phys Rehab Kur Med 12: 272–283. Hashkes PJ. (2002). Scand J Rheumatol. 31: 172–177. Hildebrandt G, Lammert W. (1986). Z Phys Med Baln Med Klim. 15: 73–80. Hildebrandt G. (1991). J Physiol Pharmacol 42: 5–27. Hildebrandt G, Gutenbrunner C. (1998a). Die Kur Kurverlauf, Kureffekt und Kurerfolg [Spa therapy its course, effects and health improvements]. In: Gutenbrunner C, Hildebrandt G (ed.) Handbuch der Balneologie and medizinischen Klimatologie. Springer, Berlin, pp. 85–186. Hildebrandt G, Gutenbrunner C. (1998b). Therapeutische Physiologie. In: Gutenbrunner C, Hildebrandt G (ed) Handbuch der Balneologie and medizinischen Klimatologie. Springer, Berlin, pp. 5–84. Leaute-Labreze C, Saillour F, Chene G, Cazenave C, Luxey-Bellocq ML, Sanciaume C, Toussaint JF, Taieb A. (2001). Arch Dermatol. 137: 1035–1039. Matz. H, Orion E, Wolf R. (2003). Soziale Sicherheit 9: 549–553. Neumann L, Sukenik S, Bolotin A, Abu-Shakra M, Amir M, Flusser D, Buskila D. (2001). Clin Rheumatol. 20: 15–19. Nishimura N, Sugenoya J, Matsumoto T, Kato M, Sakakibara H, Nishiyama T, Inukai Y, Okagawa T, Ogata A. (2002). Eur J Appl Physiol. 87: 337–342. Pittler MH, Karagulle MZ, Karagulle M, Ernst E. (2006). Rheumatology (Oxford). 45: 880–884. Robiner WN. (1990). J Behav Med. 13: 157–173. Strauss-Blasche G, Ekmekcioglu C, Klammer N, Marktl W. (2000a). Forsch Komplementarmed Klass Naturheilkd. 7: 269–274. Strauss-Blasche G, Ekmekcioglu C, Marktl W. (2000b). Occup Med (Lond). 50: 167–172. Strauss-Blasche G, Ekmekcioglu C, Vacariu G, Melchart H, Fialka-Moser V, Marktl W. (2002). Clin J Pain. 18: 302–309. Strauss-Blasche G, Muhry F, Lehofer M, Moser M, Marktl W. (2004a). J Leisure Res. 36: 293–309. Strauss-Blasche G, Riedmann B, Schobersberger W, Ekmekcioglu C, Riedmann G, Waanders R, Fries D, Mittermayr M, Marktl W, Humpeler E. (2004b). J Travel Med 11: 300–306.
Spa Therapy and Quality of Life Strauss-Blasche G, Gnad E, Ekmekcioglu C, Hladschik B, Marktl W. (2005). Cancer Nurs. 28(5): 390–398. Tishler M, Brostovski Y, Yaron M. (1995). Clin Rheumatol. 14: 21–25. van Tubergen A, Landewe R, van der Heijde D, Hidding A, Wolter N, Asscher M, Falkenbach A, Genth E, The HG, van der Linden S. (2001). Arthritis Rheum. 45: 430–438. Van Tubergen A, Boonen A, Landewe R, Rutten-Van Molken M, Van Der Heijde D, Hidding A, Van Der Linden S. (2002). Arthritis Rheum. 47: 459–467. van Tubergen A, van der Linden S. (2002). Ann Rheum Dis. 61: 273–275.
105
Verhagen AP, Bierma-Zeinstra SM, Boers M, Cardoso JR, Lambeck J, de Bie RA, de Vet HC. (2007). Cochrane Database Syst Rev: CD006864. Wigler I, Elkayam O, Paran D, Yaron M. (1995). Rheumatol Int. 15: 65–68. Yilmaz B, Goktepe AS, Alaca R, Mohur H, Kayar AH. (2004). Joint Bone Spine. 71: 563–566. Zijlstra TR, Braakman-Jansen LM, Taal E, Rasker JJ, van de Laar MA. (2007). Rheumatology (Oxford) 46: 1454–1459. Zijlstra TR, van de Laar MA, Bernelot Moens HJ, Taal E, Zakraoui L, Rasker JJ. (2005). Rheumatology (Oxford) 44: 539–546.
1809
106 Health-Related Quality of Life and Prioritization Strategies in Waiting Lists: Spanish Aspects M. Nu´n˜ez . E. Nu´n˜ez . J. M. Segur 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1812
2 Waiting Lists . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1813 2.1 WL and Surgical Procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1814 3
Health Related Quality of Life Applied to the Study of WL . . . . . . . . . . . . . . . . . . . . . 1817
4
Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1822 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1823
#
Springer Science+Business Media LLC 2010 (USA)
1812
106
HRQL and Prioritization Strategies in Waiting Lists: Spanish Aspects
Abstract: In developed countries, both increases in life expectancy and advances in health technology have resulted in > waiting lists (WL) in the majority of national health systems, where citizens have the right to universal health care and protection. The health and social problems generated by WL have led to measures to minimize their impact. In Spain, maximum guaranteed WL times have been established and supply has been increased by increasing productivity and resources. Most studies on WL have been aimed at establishing priorities based on explicit criteria developed through the participation and consensus of health professionals, patients and the general public. In addition, strategies have been incorporated to improve surgical indications and diagnostic tests. Some studies have exclusively centered on the use of > health-related quality of life (HRQL) measures, which assess the health benefits that directly affect the welfare of the patient. Elective surgical procedures, mainly cataracts and hip and knee arthroplasties are responsible for the most difficult waiting lists in terms of numbers and waiting list time. This chapter presents a view of this complex problem in Spain, based on existing studies and centered on surgical WL and HRQL. List of Abbreviations: AC, Autonomous Communities; ADLs, Activities of Daily Living; CA, > Conjoint Analysis; HRQL, Health-Related Quality of Life; HUI3, > Health Utilities Index Mark 3; NSAIDs, Non Steroidal Anti-Inflammatory Drugs; OA, Osteoarthritis; RPV, Reference Population Values; SF-36, Medical Outcomes Study Short Form-36; SIP, > Sickness Impact Profile; SNHS, Spanish National Health System; TEFR, > Therapeutic Education and Functional Readaptation; THR, Total Hip Replacement; TKR, > Total Knee Replacement; VAS, Visual Analogue Scale; VF-14, > Visual Function Index; WL, Waiting Lists; WLT, Waiting List Time; WOMAC, > Western Ontario and McMaster Universities Osteoarthritis Index
1
Introduction
In developed countries, increased life expectancy has resulted in an aging population which has led to changes in patterns of morbidity and mortality. This, together with the success of new diagnostic and therapeutic health technologies, and the inherent problem of the distribution of scarce resources, have resulted in the appearance of waiting lists (WL) at both the primary and secondary health levels in countries with a national health system providing universal health care (Cerda´ et al., 2002; Editorial, 2002). A WL can be defined as a group of patients who, at a given moment, are waiting for a specific operation, consultation or diagnostic test, whose delay is attributable to the organization and resources available, and thus, in some fashion, represent a way of managing health benefits (Ministerio de Sanidad y Consumo, 2007). WL for surgical procedures generally cause the most controversy as the therapeutic indication is well-established (Espallargues et al., 2003; Nu´n˜ez et al., 2007a). In spite of the differences between health systems, the longest WL are normally for trauma/ orthopedics, ophthalmology and general and digestive surgery and this is reflected in the Spanish health system. Health in Spain is a non contributory benefit which is financed through taxes and is included in the general financing of each of 17 Autonomous Communities (AC) (a process which occurred gradually between 1981 and 2002). It covers approximately 99.5% of the Spanish population. Practically all primary, secondary, and hospital services are free. Drug prescriptions for people aged 65 years), improvements in outcomes due to therapeutic techniques, increased incomes and health expectations, have resulted in vastly increased demand for health care, which is reflected, among other things, by the generation of WL (Cerda´ et al., 2002; Editorial, 2002). Health WL exist in most European countries and affect both primary and specialized health care. Initial actions were aimed at reducing the number of patients on WL. Subsequently, attention turned to reducing waiting list times (WLT) (Espallargues et al., 2000).
1813
1814
106
HRQL and Prioritization Strategies in Waiting Lists: Spanish Aspects
Many European countries began to tackle the problems implied by WL some years ago by guaranteeing waiting times, introducing information systems and, especially, introducing emergency plans. In 1992, Sweden established a 3 month maximum for 12 procedures; in 1993 Ireland introduced a Waiting List Initiative; in 1987 the United Kingdom drew up an emergency plan for WL, as well as times of guarantee from 1991. However, in Spain, it was not until 2003 when the Spanish National Health System (SNHS) introduced a global perspective and established measures for the homogenous treatment of information on WL (> Table 106-1) (Espallargues et al., 2000; Ministerio de Sanidad y Consumo, 2007). Before applying any proposal to organize or study WL, detailed information obtained from explicit, systematic, consensual information systems is necessary, and unfortunately this is not yet always possible (Editorial, 2002).
2.1
WL and Surgical Procedures
WL for elective surgical procedures provoke more debate, partly because the therapeutic indication is well established. Inequalities exist both in the type of surgery and in the expected
. Table 106-1 Indicators for the systematization of surgical waiting lists in the Spanish National Health System N of patients waiting for SP N of patients in structural delaya and rate per thousand inhabitants N of transitorily nonprogrammable patients N of patients waiting after rejection of alternative centre Mean waiting time (days) in patients waiting for SP Mean waiting time of patients in structural delay Mean waiting time of patients after rejection of alternative centre Number of patients on structural waiting according to time N of patients a) 0–30 days, b) 31–90 days c) >90 days N of patients after rejection of alternative center according to time N of patients in sections a) 0–30 days, b) 31–180 days, c) 181–365 days d) >365 days N of entries in records of patients waiting for SP in the period Total N of entries in the period and rate per thousand inhabitants N patients in the period leaving waiting lists Total N of patients leaving in the period Total N of patients leaving by operation in the period Mean wait (days) of operated patients Mean total delay of operated patients Mean delay of patients undergoing programmed operations in the period Prospective mean wait (days) Summary of indicators for the homogenous treatment of information on WL (Date from Spanish Ministry of health). SP surgical procedure; N number a Structural delay: patients who, at a specific moment, are waiting for surgery, and whose wait is attributable to the organization and resources available Transitorily nonprogrammable patients: patients waiting for surgery whose programming is not possible, at a specific moment, due to clinical motives or to a request for postponement due to personal or work reasons
HRQL and Prioritization Strategies in Waiting Lists: Spanish Aspects
106
benefits (e.g., varicose veins). Likewise, there is great geographic variability between regions, including high rates of unnecessary surgery (tonsillectomy, hysterectomy, etc), caused partly by the lack of consensus or clinical protocols for patient selection and disagreement on the optimal time of surgery (eg., knee and hip arthroplasty) (Espallargues et al., 2000; Nun˜ez et al., 2007b). In spite of the differences between health systems, in most countries WL are more frequent in some medical specialties: orthopedics, otolaryngology, ophthalmology, vascular surgery, general surgery and plastic surgery. In Spain, orthopedics, ophthalmology and general and digestive surgery account for 65% of patients on WL (> Figure 106-1) (Ministerio de Sanidad y Consumo, 2007). The different strategies used to solve the problem of WL – emergency plans, guarantees, or prioritization – are of doubtful efficacy when not accompanied by a rigorous analysis of demand. In the context of saturation of supply, prioritization seems to be the most promising alternative of management of the demand because it allows the response to be improved when
. Figure 106-1 Patients on waiting lists in Spain according to medical process (December 2003 and June 2007). The figure shows that waiting lists have been reduced in most specialties between 2003 and 2007. (Data from Spanish Ministry of Health). (Figure composed by authors specifically for this chapter, Nu´n˜ez M et al. 2008)
1815
1816
106
HRQL and Prioritization Strategies in Waiting Lists: Spanish Aspects
the WLT is associated to the benefit obtained, as long as a maximum WLT based on the degree of need is guaranteed (Editorial, 2002). In Spain, the different emergency plans carried out have had uncertain results and, more recently, guaranteed WLT for surgical procedures have been proposed but not yet evaluated, as far as we know. Although systems for the organization, control and reduction of WL have been introduced in Spain, there are few studies on this subject. In recent years, the health authorities have begun to promote studies analyzing methods of patient prioritization, similar to those applied in other countries (Espallargues et al., 2000). Most studies have investigated cataract surgery and hip and knee arthroplasty. Peiro´ et al. (1998) calculated rates per 100 inhabitants of cataract surgeries in eight health regions of Alicante in 1994/1995 in relation to the number of ophthalmologic beds, ophthalmologists and hours of ophthalmology surgery time per 100,000 inhabitants. They found differences between regions and a high correlation between the number of operating hours available and rates of operations, but not with the number of ophthalmologists. Neither was there an association between the number of surgeries and the WL. This suggests that the number of surgeries depends more on the supply than on the demand. Likewise, Norregaard et al. (1998) compared indications for cataract surgery in patients aged 50 years in the United States, Denmark, the province of Manitoba (Canada), and the city of Barcelona (Spain). Information on visual functional impairment (VF-14 score) (Steinberg et al., 1994) and HRQL by the general Sickness Impact Profile (SIP) questionnaire (Bergner et al., 1981) was used. The authors found similar indications for cataract surgery in the United States and Denmark and significantly more restricted indications in Manitoba and Barcelona. Possible explanations included differences in sociodemographic characteristics, access to care, surgeons’ willingness to operate, and patient demand. As stated above, most studies focus on rationalizing lists through models that estimate preferences. Applied to the study of WL, each of the selected criteria is measured by a scale and points assigned to each level of the scale, so that the sum of points situates the patient in a specific position on the list (Pinto et al., 2000). Cerda´ et al. (2002) analyzed the problem of surgical WL from a general perspective within the area of health management and explored quantitative techniques usually used in the management of WL, such as queuing theories, simulation, optimization or multicriteria analysis. Various researchers have used conjoint analysis (CA), which is widely used in market research (Green and Srinivasan, 1978; Luce and Tukey, 1964), to obtain a system of prioritizing for different surgical procedures. Rodriguez-Mı´guez et al. (2004) used a point system based on social preferences for the management of WL for cataract surgery and found that visual incapacity and age appeared to be the most important attributes in the prioritization process; likewise, Rivera et al. (2004), applied CA to prioritize patients on a WL for varicose vein surgery and found that clinical severity, affectation of activities of daily living (ADLs) and the WLT were the most relevant attributes in prioritizing these patients, with clinical severity having the greatest weight; Abad et al. (2006) studied a lineal points system to prioritize patients with benign prostatic hyperplasia on a WL for prostatectomy in the Gallician Health Service applying similar methods. In Catalonia, various studies were carried out to develop prioritization criteria for cataract surgery and knee and hip arthroplasty. A study was carried out in two phases.
HRQL and Prioritization Strategies in Waiting Lists: Spanish Aspects
106
The first, qualitative phase used focal groups combined with nominal groups to identify the most relevant criteria for health professionals, patients, family and the general public in deciding that one patient on a WL should have preference over another. The criteria included disease severity, the probability of recovery, limitations in ADLs, limitations at work, having someone to care for the patient and having someone in charge of the patient (Espallargues et al., 2003). Another study (Espallargues et al., 2004) was carried out in four Barcelona hospitals of different levels to evaluate the validity, viability and clinical utility of the prioritization system (Arnett and Hadorn, 2003; Haddorn and Holmes, 1997; Pinto et al., 2000). Participants included 919 cataract patients and 654 hip and knee arthroplasty patients. The results showed that the instrument was valid although there was some discrepancy between doctors and patients in the assessment of the functional and social situation. Problems were observed in the implementation. In the Basque Country a group of researchers have used the RAM method (expert opinionDelphi panel and available scientific evidence) (Brook et al., 1986) to study explicit criteria for the appropriate indication of surgical procedures for cholecystectomy in patients with nonmalignant diseases (Quintana et al., 2002), cataracts (Ma Quintana et al., 2006; Quintana et al., 2006a) and hip and knee arthroplasty (Escobar et al., 2003; Escobar et al., 2007).
3
Health Related Quality of Life Applied to the Study of WL
Today, it is recognized that HRQL measures can be used as instrument to evaluate the effects of health technology and to establish criteria or limits in order to optimize health products. In this way, the patient becomes a central element of the system. HRQL is regarded as one of the most relevant parameters of the health status (Bowling, 1997). The most-commonly used generic instruments to measure HRQL in any disorder or population is the > Medical Outcomes Study Short Form-36 (SF-36) (Ware and Sherbourne, 1992) The SF-36 has four physical and four mental health dimensions: physical function, social function, physical role limitations, emotional role limitations, general mental health, energy, bodily pain, and general health perception. It allows age-gender standardized comparisons with the general Spanish population (reference population values [RPV]). With respect to the specific disease questionnaire, the > Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) (Bellamy et al., 1988) is recommended for monitoring functional outcomes in the knee and hip joints. The WOMAC has three dimensions: pain, stiffness, and function. Although condition-specific measures are usually more sensitive to change than generic measures, these play an important complementary role by providing information on the overall health status, capturing side effects and the effects of comorbidities. Thus, measurement of patient reported health outcomes have provided an alternate source of valuable information, based, generally on assessments by health professionals. However, studies show that the values, perceptions and preferences of patients for specific results or care processes are not always in accordance with the assessment of health professionals (Nu´n˜ez et al., 2007b). Measures of HRQL are valid, reliable and responsive for the evaluation of health results (Bowling, 1997).The majority of these studies have been carried out in Catalonia and the
1817
1818
106
HRQL and Prioritization Strategies in Waiting Lists: Spanish Aspects
Basque Country and are focused on investigating hip and knee arthroplasty and cataract surgery, as far as we know. Allepuz et al. (2005), developed a linear points system to prioritize patients in WL for hip and knee arthroplasty and cataract surgery based on clinical, functional and social aspects and evaluating the validity of the convergent and discriminating construct of the prioritization system in patients awaiting hip and knee arthroplasty (1,125) and knee and cataract surgery (2,062) from four hospitals in Catalonia, Andalucia, and Aragon included in WL between June 2001 and May 2005. Information was obtained on the priority score and the surgeon’s assessment of priority using a visual analogue scale (VAS) during the medical visit, and on sociodemographic aspects, HRQL questionnaires using the VF-14 (Steinberg et al., 1994), WOMAC and the Health Utilities Index Mark 3 (HUI3)) (Grootendorst et al., 2000) and comorbidities from a telephone interview. The validity of the convergent and discriminating construct was evaluated using the correlation between the priority score, the VAS and HRQL measures. The results found an acceptable validity of questionnaires to establish the priority. The same group in a recent study (Serra-Sutton et al., 2008) in Catalonia carried out 650 telephone interviews with patients on a WL for hip or knee arthroplasty and knee to measure the patient’s HRQL, using the general SF-36 and disease-specific WOMAC questionnaires. In addition, they collected patients’ perception of disease severity. As in other studies, the patients’ HRQL was considerably affected. Patients who perceived their disease as very severe had worse scores in both the SF-36 and WOMAC. Other studies have also found an important relationship between illness representations and HRQL (Lee et al., 2008). Quintana et al. (2000), in the Basque country, after the creation of explicit criteria, mentioned above, studied their validity by field work, and tested the appropriateness of this indication tool for total hip replacement (THR) evaluating the results by HRQL measures (SF-36 and WOMAC). Patients considered appropriate candidates for surgery, based on their composite indication scores, showed more improvement in HRQL (SF-36 and WOMAC) after 3 months in the pain and physical dimensions. In contrast, patients whose indications were considered inappropriate had a moderate improvement in their HRQL. This result partly supported the use of this indication algorithm as a screening tool for assessing the appropriateness of THR surgery in osteoarthritis (OA). Likewise, Quintana et al. (2006b) studied the association between explicit appropriateness criteria for THR and total knee replacement (TKR) and changes in HRQL (SF-36 and WOMAC) before and 6 months after surgery. They found that patients considered appropriate candidates for these procedures had greater improvements than inappropriate candidates in all three WOMAC domains (pain, function and stiffness). In addition, appropriate candidates undergoing THR had greater improvements in the physical function, physical role, bodily pain, and social function (SF-36) than inappropriate candidates. Appropriate candidates for TKR demonstrated greater improvement in the social function domain of the SF-36 after the procedure than those deemed inappropriate. The authors suggest the results show a direct relationship between explicit appropriateness criteria and better HRQL outcomes after THR and TKR surgery and support the use of these criteria for clinical guidelines or evaluation purposes and that the application of these criteria in field work has shown that the current ‘‘first-come, first-served’’ management of WL is not the best method in terms of HRQL. With respect to purely HRQL studies of patients on WL, Nun˜ez et al. (2006) evaluated the effect of therapeutic education and functional readaptation (TEFR) on HRQL in patients with osteoarthritis on a WL for TKR in a randomized controlled trial of 9 months duration. One hundred consecutive outpatients (71 females, mean age 71 years (range 50–86), mean
HRQL and Prioritization Strategies in Waiting Lists: Spanish Aspects
106
disease duration 11.84+/ 10.52 months) were randomized to two groups. The intervention group received TEFR added to conventional (pharmacological) treatment (n = 51). The control group received conventional (pharmacological) treatment only (n = 49). The main outcome variable was self-reported HRQL measured by the WOMAC. Secondary outcomes were general HRQL measured by the SF-36, number of visits to general physicians and their cost. Eighty patients completed the study. Significant improvements in pain and function (WOMAC) and bodily pain and physical function (SF-36), were observed at 9 months (6 months after the intervention) in the TEFR group. Consumption of analgesics increased significantly in the TEFR group compared with controls, whereas the number of non steroidal anti-inflammatory drugs (NSAIDs) diminished in both groups. Pharmacologic therapy included the use of anti-inflammatory medications that have potentially serious long-term side effects. At study entry, it was observed that patients were taking practically no analgesics and there was a predominant use of low doses of NSAIDs. This relative increase in the consumption of analgesics could be due to better compliance as an effect of the educational intervention. The results suggest that a program of TEFR during the period on the WL for TKR may reduce the negative impact of this situation. > Figures 106-2 and > Figures 106-3 show the changes in HRQL (SF-36 and WOMAC scores) of patients on a WL for TKR after the application of a program of TEFR. > Figure 106-2 compares the results with age (65–74 years: mean age of study group 71 years) and sex (female) matched RPV. It shows that the greatest deficiencies were in the
. Figure 106-2 Change in SF-36 dimension scores after a TERF intervention compared with age and gender (female) matched Spanish RPV. RPV: Reference population values. The figure shows that the greatest differences were in the physical function, physical role and bodily pain dimensions and that the general health, vitality and mental health dimension scores were similar to Spanish RPV. TERF: Therapeutic education and functional readaptation. SF-36: Medical Outcomes Study 36-Item Short Form (scores ranging from 0 [worst] to 100 [best]
1819
1820
106
HRQL and Prioritization Strategies in Waiting Lists: Spanish Aspects
. Figure 106-3 Change in WOMAC dimension scores after a TERF intervention. The figure shows that the greatest changes after the TERF program were in the pain and function dimension scores. TERF: Therapeutic education and functional readaptation. WOMAC: Western Ontario and McMaster Universities Osteoarthritis Index (scores ranging from 0 [best] to 100 [worst])
physical function and physical role dimensions and that after the TEFR, the general health, vitality and mental health dimension scores had improved and were similar to RPV. In another, cross-sectional study, the same authors (Nu´n˜ez et al., 2007a) determined: (1) HRQL in patients with severe OA on a WL for TKR and compared it with general Spanish reference population values; (2) the influence of sociodemographic and clinical variables on HRQL dimensions; and (3) the use and cost of resources related to knee OA. HRQL was measured by the WOMAC and SF-36 questionnaires. Sociodemographic and disease characteristics, body mass index, pharmacological treatment and the cost and use of economic resources related to knee OA during the 6-months previous to baseline were recorded. One hundred consecutive outpatients were included. Patients showed worse HRQL measured by SF-36 than Spanish RPV, mainly in the physical function, physical role and bodily pain dimensions. Most of the patients had low educational levels and incomes, a high level of associated comorbidities, mainly cardiovascular and impaired sight; they were overweight and almost half were obese. The eight dimensions of the SF-36 showed worse values than age sex-matched Spanish RPV. The physical function, physical role and bodily pain dimension scores were >40% worse than RPV. Fifty per cent of the patients reported feeling worse or much worse than 1 year before. The factors that had a negative impact (worse) on HRQL (SF-36 and WOMAC) were female gender, disease duration, the number of comorbidities and dissatisfaction with current treatment and feeling worse or much worse than 1 year before. In contrast, factors that had a positive (better) impact on HRQL were older age, higher educational level and higher family income. When the use of resources by these patients was analyzed, the results showed a surprisingly-low use of physician care, with most patients making only one medical visit, generally to a general practitioner The low number of visits to specialists such as rehabilitative physicians is also surprising, as reports have recommended physiotherapy to maintain correct
HRQL and Prioritization Strategies in Waiting Lists: Spanish Aspects
106
muscular tone before interventions as a factor in better postoperative recovery, due to the severe locomotor deficits present in OA. Patients reported taking low doses of NSAIDs, with no case complying with anti-inflammatory or analgesic guidelines. Less than half of the patients had undergone any technical procedure, mainly radiographies. Consequently, their direct medical costs were very low. The main costs incurred by these patients were on nonmedical resources (cost of caregiver time), mainly as a consequence of functional limitation and loss of autonomy. The results seem to suggest indirectly that other aspects of the therapeutic management of this group of patients should be reconsidered. Some results, such as overweight and/or obesity of long evolution, the low number of visits to physicians and other health professionals, or the widespread use of only one pharmacological treatment, suggest little compliance with knee OA management guidelines that emphasize the principal roles of weight loss, exercise, physical therapy, or drugs combined with non-pharmacological strategies to reduce disability and pain in knee OA (Jordan et al., 2003) > Figures 106-4 and > Figures 106-5 show the SF-36 and WOMAC dimension scores for some of the cited studies. The differences observed indicate the diversity in terms of HRQL of patients included on WL. Further studies that will contribute to the establishment of criteria to detect the ideal moment for surgery are needed to address these differences. Although the few studies carried out in Spain on WL are not representative, they seem to support the suggestion by Bernal (2002) that ‘‘it should be remembered that with the exception of the queues produced in the emergency department, the remaining queues produced in the health system are mediated by the decision of a doctor. Therefore an inescapable strategy to
. Figure 106-4 Comparison of the mean WOMAC domain scores in patients on WL for TKR in different studies. The figure shows the differences in WOMAC scores between different Spanish studies of patients on WL. WOMAC: Western Ontario and McMaster Universities Osteoarthritis Index (scores ranging from 0 [best] to 100 [worst])
1821
1822
106
HRQL and Prioritization Strategies in Waiting Lists: Spanish Aspects
. Figure 106-5 Comparison of mean SF-36 dimension scores in patients on WL for TKR in different studies. The figure shows the differences in SF-36 scores between different Spanish studies of patients on WL. SF-36: Medical Outcomes Study 36-Item Short Form (scores ranging from 0 [worst] to 100 [best])
manage WL consists of attenuating the problems derived from uncertainty or ignorance with respect to the diagnosis or prognosis.’’ To this we would add the lack of consensus on protocols for patient selection and the ideal moment to carry out the surgical intervention. The clinical decision to perform a total joint arthroplasty involves weighing the potential risks against the benefits for each patient. HRQL, prosthetic/technical issues, as well as medical, surgical and social factors need to be considered. Despite the encouraging results, anywhere from 15 to 30% of patients receiving total joint arthroplasties report little or no improvement after surgery or are unsatisfied with the results. Although surgeons and referring physicians agree that the primary indications for total joint arthroplasty are pain and dysfunction, the effects of secondary patient characteristics on pain, function and HRQL are undetermined. Determinants of poor pain and functional outcomes have seldom been examined. Because little is known about the determinants for total hip or knee arthroplasty, indications and contraindications for total joint arthroplasties are poorly defined (Jones et al., 2007).
4
Conclusions
In conclusion, the study of WL in Spain began late compared with other developed countries and still has a long way to go. Most studies carried out to date have concentrated on utilitarian methods of assigning priorities. Studies that explicitly consider the HRQL of patients on WL and the implications this may have for health costs and resources are sadly lacking in number. However, determining patients’ HRQL can measure the efficacy and effectiveness of health
HRQL and Prioritization Strategies in Waiting Lists: Spanish Aspects
106
interventions, improve clinical decisions, observe population needs, and determine, in part, the causes and consequences of health differences and ensure optimal use of available resources. Given the sociodemographic changes in Spain, with an increasing number of old people and obese subjects and the fact that care in the family, a traditional feature of Spanish life, is decreasing, leaving an increasing number of people dependent on social and health services, an urgent reappraisal of the makeup of WL and criteria for inclusion on them is overdue. Studies that assess the HRQL of patients before, during and after inclusion on a WL could help to determine guidelines for the management of this important, and increasing, problem.
Summary Points A WL can be defined as a group of patients who, at a given moment, are waiting for a
specific operation, consultation or diagnostic test whose delay is attributable to the organization and resources available. Sociodemographic and morbidity changes, improvements in outcomes due to new diagnostic and therapeutic techniques and increased health expectations, have resulted in vastly increased demand for health care, which is reflected by the generation of WL. WL are a common feature of countries with National Health Systems, even though there are contrasting health models, different types of organization, financing or service provision. WL for surgical procedures provoke more debate, partly because the therapeutic indication is well established. In Spain, orthopedics, ophthalmology and general and digestive surgery account for 65% of patients on WL. With the exception of the queues produced in health centre waiting rooms and in the emergency department, the remaining queues are mediated by medical decisions. WL can be understood as the final stage of a disease process. Therefore, studies that assess the HRQL of patient before, during and after inclusion on a WL could help to determine guidelines for the management of the disease. The use of HRQL measures in patients on WL allow determination of the health benefits that directly affect patients’ welfare. Studies of WL that assess the HRQL of patients can help to determine the suitability of the surgical indication. In patients on WL, determination of the HRQL can provide valuable information that will contribute to the establishment of criteria to detect the ideal moment for surgery or other health intervention. Many more studies from HRQL perspectives are necessary to evaluate the phenomenon of waiting lists in Spain. Programs involving patient education and physical rehabilitation can improve the HRQL of patients on WL.
Acknowledgments We wish to thank F. Segura, D. Buss and R. Ortega for their help.
1823
1824
106
HRQL and Prioritization Strategies in Waiting Lists: Spanish Aspects
References Abad P, Alvarez B, Rodrı´guez E, Rodrı´guez A. (2006). Hacienda Pu´blica espan˜ola/Revista de Economı´a Pu´blica (Instituto de Estudios Fiscales). 179: 113–134. Allepuz A, Espallargues M, Moharra M. (2005). Informatiu AATRM. 37: 6–8. Arnett G, Hadorn DC. (2003). Can J Surg. 46: 290–296. Bellamy N, Buchanan WW, Goldsmith CH, Campbell J, Stitt LW. (1988). J Rheumatol. 15: 1833–1840. Bergner M, Bobbitt RA, Carter WB, Wilson BS. (1981). Med Care. 19: 787–805. Bernal E. (2002). Gac Sanit. 16: 436–439. Bowling A. (1997). 2nd ed. Open University Press, Buckingham, Philadelphia. Brook RH, Chassin MR, Fink A, Solomon DH, Kosecoff J, Park RE. (1986). Int J Technol Assess Health Care. 2: 53–63. Cerda´ E, de Pablos L, Rodrı´guez MV. (2002). INV. 3/02. Editorial. (2002). Gac Sanit. 16: 374-375. Escobar A, Quintana JM, Arostegui I, Azka´rate J, Gu¨enaga JI, Arenaza JC, Garai I. (2003). Int J Technol Assess Health Care. 19: 57–70. Escobar A, Quintana JM, Bilbao A, Iban˜ez B, Arenaza JC, Gutie´rrez L, et al. (2007). J Eval Clin Pract. 13: 429–434. Espallargues M, Gallo P, Joan M, Pons JMV, SampietroColom L. (2000). Age`ncia d’Avaluacio´ de Tecnologia i Recerca Me`diques. Servei Catala` de la Salut. Departament de Sanitat i Seguretat Social. Generalitat de Catalunya. Novembre Espallargues M, Sampietro-Colom L, Castells X, Pinto JL, Rodrı´guez E. (2003). Age`ncia d’Avaluacio´ de Tecnologia i Recerca Me`diques. Catsalut. Department de Sanitat i Seguretat Social. Generalitat de Catalunya Espallargues M, Sampietro-Colom L, Castells X, Comas M, Rodrı´guez E, Pinto JL. (2004). Age`ncia d’Avaluacio´ de Tecnologia i Recerca Me`diques. Catsalut. Department de Sanitat i Seguretat Social. Generalitat de Catalunya Setembre (BR02/2004) Fricker J. (1999). BMJ. 318: 78. Green PV, Srinivasan V. (1978). J Consum Res. 5: 103–122. Grootendorst P, Feeny D, Furlong W. (2000). Med Care. 38: 290–299. Haddorn DC, Holmes AC. (1997). BMJ. 314: 131–134. Jones CA, Beaupre LA, Johnston DWC, Suarez-Almazor ME. (2007). Rheum Dis Clin N Am. 33: 71–86. Jordan KM, Arden NK, Doherty M, Bannwarth B, Bijlsma JW, Dieppe P, et al. (2003). Ann Rheum Dis. 62: 145–155.
Lee BO, Chaboyer W, Wallis M. (2008). J Nurs Scholarsh. 40: 83–90. Luce RD, Tukey JW. (1964). J Math Psychol. 1: 1–27. Ma Quintana J, Escobar A, Bilbao A. (2006). Health Serv Res. 6: 24. Ministerio de Sanidad y Consumo. (2007). http://www. msc.es/estadEstudios/estadisticas/inforRecopilaciones/docs/listaPublicacion07CI.pdf Norregaard JC, Petersen PB, Alonso J, Dunn E, Black C, Andersen TF. (1998). Br J Ophthalmol. 82: 1107–1111. Nun˜ez M, Nun˜ez E, Segur JM, Macule F, Quinto L, Herna´ndez MV, Vilalta C. (2006). Osteoarthritis Cartilage. 14: 279–285. Nu´n˜ez M, Nu´n˜ez E, Segur JM, Macule´ F, Sanchez A, Hernandez MV, Vilalta C. (2007a). Osteoarthritis Cartilage. 15: 258–265. Nu´n˜ez M, Nu´n˜ez E, del Val JL, Ortega R, Segur J, Herna´ndez M, Lozano L, Sastre S, Macule´ F. (2007b). Osteoarthritis Cartilage. 9: 1001–1007. Peiro´ S, Meneu R, Marques JA, Librero J, Ordin˜ana R. (1998). Papeles de Economı´a Espan˜ola. 76: 165–175. Pinto JL, Rodrı´guez E, Castells X, Gracia X, Sa´nchez FI. (2000). Ministerio de Sanidad y Consumo, Secretarı´a General Te´cnica Quintana JM, Aro´stegui I, Azkarate J, Goenaga JI, Elexpe X, Letona J, Arcelay A. (2000). J Clin Epidemiol. 53: 1200–1208. Quintana JM, Cabriada J, de Tejada IL, Varona M, Oribe V, Barrios B, Aro´stegui I, Bilbao A. (2002). Qual. Saf Health Care. 11: 320–326. Quintana JM, Escobar A, Arostegui. (2006a). Health Serv Res. 6: 23. Quintana JM, Escobar A, Arostegui I, Bilbao A, Azkarate J, Goenaga I, Arenaza JC. (2006b). Arch Intern Med. 166: 220–226. Rivera A, Gonza´lez E, Martı´n MA, On˜ate JL, Sa´nchez I. (2004) Cuadernos Econo´micos de I.C.E. 67: 93–106. Rodriguez-Mı´guez E, Herrero C, Pinto-Prades JL. (2004). Soc Sci Med. 59: 585–594. Serra-Sutton V, Allepuz A, Espallargues M. (2008). Informatiu AATRM. 42: 3–5. Steinberg EP, Tielsch JM, Schein OD, Javitt JC, Sharkey P, Cassard SD, Legro MW, Diener-West M, Bass EB, Damiano AM. (1994). Arch Ophthalmol. 112: 630–638. Ware JE, Sherbourne CD. (1992). Med Care. 30: 473–483.
107 Hirsutism and Quality of Life S. Davies 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1826
2
Impact on Well-Being . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1827
3
Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1829
4 4.1 4.2 4.3 4.4
Health Economics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1832 Quality-Adjusted Life Years Generated . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1834 Cost per QALY Gained from Use of Eflornithine Cream . . . . . . . . . . . . . . . . . . . . . . . . . . 1834 Sensitivity Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1834 Willingness-To-Pay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1835
5
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1835 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1836
#
Springer Science+Business Media LLC 2010 (USA)
1826
107
Hirsutism and Quality of Life
Abstract: Hirsutism in women results in significant psychological and social problems. It impacts negatively upon the quality of life of women and is the cause of stress, anxiety and depression, particularly in cultures where hairlessness is viewed as the norm for women. Facially hirsute women perceive themselves to be in poorer health than non-hirsute women, and there are fewer in employment despite having similar educational experiences and levels of attainment. Hirsute women spend much time and effort in removing excess facial: and can feel overwhelmed and frustrated by their efforts. All currently available treatments have limited effectiveness and can have serious side-effects. It is therefore a testament to the levels of distress caused by the problem that women persist with treatment despite the difficulties encountered. There is very limited evidence as to the cost-effectiveness of the various treatments for hirsutism but the evidence that does exist indicates that the use eflornithine cream results in a small gain in Quality Adjusted Life Years (> QALYs) for patients and provides acceptable value for money. When women and their partners were asked to consider what they would be willing to pay for an effective treatment for hirsutism, the average amounts were £40 and £47 per month respectively which is more than is currently spent on alternatives by women with the condition. It is essential that research continues to be undertaken into this distressing problem and that clinicians who treat women with hirsutism deal with them sensitively, acknowledging the major psychological and emotional difficulties these women are facing. List of Abbreviations: BAD, British Association of Dermatologists; > CI, confidence interval; DTB, Drug and Therapeutics Bulletin; EQ5d, EuroQol 5d; FDA, Food and Drugs Administration; GP, general practitioner; > HRQoL, health related quality of life; PCDS, primary care dermatology society; PCOS, polycystic ovary syndrome; QALY, quality adjusted life year; > SD, standard deviation; > VAS, visual analogue scale; > WTP, willingness to pay
1
Introduction
Hirsutism is defined medically as the presence of excessive terminal (coarse) hair that appears in women in a male pattern on the face, back, lower abdomen and inner thighs (Azziz et al., 2000; Deplewski and Rosenfield, 2000). It is estimated that between 5 and 8% of women in the general population are hirsute (Knochenhauer et al., 1998; Rosenfield, 2005) and around 900,000 women in England suffer from excess or unwanted facial hair with only a small percentage of these consulting their General Practitioner (GP) (National Horizon Scanning Centre, 2001). Market research in the USA suggests that more than 22 million women treat unwanted facial hair on at least a weekly basis, but that only 10% of these have ever consulted a physician about this condition (Keegan et al., 2003). Hirsutism is a sign of increased androgen action on hair follicles, either as a result of increased circulating levels of androgens, or increased sensitivity of hair follicles to normal levels of circulating androgens (Hunter and Carek, 2003). The majority of women with androgen levels that are more than twice the upper limit of the normal range have some degree of hirsutism (Reingold and Rosenfield, 1987). However around one half of hirsute women have idiopathic hirsutism which develops without the presence of excess androgen or any underlying medical condition (British Association of Dermatologists (BAD), 2006;
Hirsutism and Quality of Life
107
Reingold and Rosenfield, 1987). Idiopathic hirsutism is often familial and 50% of women with hirsutism have a positive family history of the disorder (Givens and Kurtz, 1992). Familial hirsutism, which typically begins during puberty, is found most commonly in dark-skinned white persons in southern European and South Asian countries in which it is considered to be a normal trait. Hirsutism is uncommon in sub-Saharan and African American blacks and is observed least commonly in East Asians and Native Americans (Goodheart and Uyttendaele, 2008). Women from different ethnic backgrounds (particularly families with Mediterranean or Middle Eastern ancestry) have different patterns of hair growth, in which it can be normal to have some hair on the face, nipples or stomach (BAD, 2006). A major factor in the development of hirsutism is polycystic ovary syndrome (PCOS). Up to one-third of women in the UK have polycystic ovaries (Balen, 1999) and an estimated onethird of these women have polycystic ovary syndrome (Hopkinson et al., 1998) which is usually defined in the UK as polycystic ovaries together with one or more characteristic features such as hirsutism. Hirsutism affects between 5 and 15% of adult women with PCOS (Walling, 2004) and more than 90% of women with hirsutism will have PCOS (Azziz et al., 2004). According to Kitzinger and Wilmot (2002) the three most bothersome symptoms of PCOS commonly reported by affected women are excess hair growth, irregular or absent menstruation, and infertility.
2
Impact on Well-Being
Hirsutism is considered undesirable by women for many reasons. Because excess body hair outside of cultural and social norms can be very distressing (Lipton et al., 2006) most women who seek treatment for hirsutism do so for cosmetic reasons (Hunter and Carek, 2003) and according to Soliman and Wardle (2006) by the time they seek medical advice, many women will have reached a point of desperation. Some physicians recognize hirsutism as any hair growth that is unwanted or embarrassing to women (Ferrante, 1998) and it is maybe this perception that accounts for even small amounts of female hair being seen as undesirable in contemporary Western culture where hairlessness is viewed as the norm for women (Toerien and Wilkinson, 2003). There is some conflicting evidence about the impact of hirsutism on psychosocial distress, or emotional and social functioning. In their study Hahn et al. (2005) found no association between these problems and hirsutism but they did find that hirsutism can lead to feelings of decreased sexual self-worth and sexual satisfaction. This adds to the evidence provided by Coffey and Mason (2003) and Keegan et al. (2003) that hirsutism is possibly linked with problems concerning female identity and sexual self-worth. Also Kitzinger and Wilmot (2002) found that women with PCOS and hirsutism reported that they considered themselves to be freakish, abnormal, and not proper women. A study by Clayton et al. (2005) suggests that the levels of anxiety and depression of hirsute women were higher than those of women attending outpatient departments with newly diagnosed gynecological or breast cancer. According to Azziz (2003) hirsutism has a significant negative impact on psychosocial development. It can have serious psychological consequences and undermines a woman’s confidence and self esteem. Lipton et al. (2006) report that nearly three-quarters of hirsute women have anxiety, one-third have clinical levels of depression and one-third feel uncomfortable in social situations and try to prevent others from coming near them. Hirsute women can experience what is described by Sonino et al. (1993) as
1827
1828
107
Hirsutism and Quality of Life
“social phobia” which was evident when the women met strangers, attended parties and went shopping for example. Soliman and Wardle (2006) report that some hirsute women become reclusive and only venture out after dark, while in young people hirsutism can be a cause of bullying, social isolation, and poor educational performance. These problems of low selfesteem and self consciousness may have an impact on their ability to work because there is evidence to suggest that there are fewer hirsute women in employment than non-hirsute women despite having similar educational experiences and levels of attainment (Davies and Phillips, 2007). See > Table 107-1. The impact of hirsutism on quality of life should also not be underestimated since it has been reported that its affects are similar to those of eczema and psoriasis and exceed those of acne (Finlay and Khan, 1994). A study by Davies and Phillips (2007) considered not only the quality of life of hirsute women in comparison to non-hirsute women, but also the impact on the partners of hirsute women. The study found that a statistically significant difference exists between hirsute and non-hirsute women in relation to health related quality of life. On a scale between 0 (dead) and 100 (perfect health) the mean score for hirsute women was 68.4 while for non-hirsute women it was 87.7. The mean score of the partners of hirsute women was 75.4 indicating that their quality of life was likely to be somewhat affected by the hirsutism of their partner. See > Table 107-2. The perceptions and reactions of women with hirsutism can differ from those of people with whom they have contact or the general public. In a study considering the impact of laser treatment on the quality of life of hirsute women, Conroy et al. (2006) report that patients score their own quality of life significantly higher than the scores attributed to them by doctors
. Table 107-1 Respondent details by group
n Mean age (95% CI)
Women with facial hirsutism
Women without facial hirsutism
46
71
38.3 (35.7, 40.9) 39.8 (36.9, 42.6)
Partners of women with facial hirsutism
Total
22
139
41.8 (37.6, 46.1)
39.6 (37.7, 41.4)
Ethnic origin White
32 (70%)
63 (89%)
15 (68%)
110 (79%)
Other
14 (30%)
8 (11%)
7 (32%)
29 (21%)
Employed
26 (57%)
65 (92%)
19 (86%)
110 (79%)
Housework
13 (28%)
0 (0%)
0 (0%)
13 (9%)
Other
Main activity
7 (15%)
6 (8%)
3 (14%)
16 (12%)
Percentage with degree or professional qualification
39%
47%
46%
44%
Percentage currently taking medication
76%
26%
36%
45%
n number in the group; CI confidence interval
Hirsutism and Quality of Life
107
. Table 107-2 Health status of participants by group – mean score and 95% CI Visual analogue scale (VAS)
Tariff
Women with facial hirsutism
68.4 (62.7, 74.1)
86.0 (80.4, 91.6)
Women without facial hirsutism
87.7 (85.4, 90.0)
96.9 (94.7, 99.1)
Partners of women with facial hirsutism
75.4 (66.5, 84.3)
90.6 (81.5, 99.7)
Overall
79.3 (76.3, 82.3)
92.3 (89.6, 95.0)
CI confidence interval. Visual Analogue Scale (VAS) is expressed as a value between 0 and 100, with higher scores representing better health
. Table 107-3 Effect of excess facial hair on participants’ lives by group – mean score and 95% CI
Effects of facial hair
Women with facial hirsutism
Women without facial Partners of women hirsutism with facial hirsutism
Overall
Bothered by excess facial hair
7.7 (6.9, 8.5)
8.5 (8.1, 8.9)
5.4 (3.9, 6.9)
7.7 (7.2, 8.2)
Bothered by time dealing with excess facial hair
7.4 (6.5, 8.3)
7.2 (6.5, 7.9)
4.9 (3.5, 6.3)
6.9 (6.4, 7.4)
Uncomfortable in exchanges of affection
7.3 (6.3, 8.3)
8.1 (7.5, 8.7)
4.9 (3.2, 6.6)
7.3 (6.7, 7.9)
Uncomfortable when meeting new people
7.6 (6.7, 8.5)
8.5 (8.0, 9.0)
5.0 (3.3, 6.7)
7.7 (7.2, 8.2)
Uncomfortable in public
7.4 (6.4, 8.4)
8.4 (7.9, 8.9)
4.9 (3.2, 6.6)
7.5 (7.0, 8.0)
CI confidence interval. Score of 10 represents an extreme negative effect and 0 has no negative effect
and nurses who were involved in their care. Similarly the Davies and Phillips (2007) study showed that when asked to rate the perceived impact of hirsutism on their lives, non-hirsute women almost always overestimated the impact of the condition, relative to those who actually suffer with the condition. See > Table 107-3. It is possible that some of these problems with quality of life could be partially ameliorated with better information giving. According to Griffing (2007) the treatment of hirsutism should begin with a careful explanation about the cause of the problem and reassurance that the patient is not losing her femininity. However in their recent study Ching et al. (2007) report that patients were not given the best information available and that patients’ perception of inadequate information about the condition is correlated with poor quality of life scores.
3
Treatment
According to the Primary Care Dermatology Society (PCDS) (2005) the symptomatic treatment of hirsutism will be largely the same regardless of the underlying cause. Drugs are usually
1829
1830
107
Hirsutism and Quality of Life
only partially effective on terminal coarse hairs, so other hair removal techniques are also required. Therefore, according to Moghetti and Toscana (2006), management of hirsutism is generally based upon a dual approach: a pharmacological therapy to reduce androgen secretion and/or androgen action, and removal of terminal hair already present by a mechanical method. Mechanical treatments include shaving, plucking and waxing, as well as the use of depilatories (creams that remove hair) and bleaching creams. These simple cosmetic treatments can be carried out safely by women in their own homes and are the cornerstone of care for hirsutism (Rosenfield, 2005). It is estimated that around 80% of women remove or conceal their unwanted facial hair themselves (Keegan et al., 2003). In a study carried out by Lipton et al. (2006) to consider the psychological problems associated with hirsutism and the impact of hair removal on daily life, two thirds of the women said they currently plucked their facial hair while around a quarter reported waxing, trimming or shaving. Most women reported using more than one method and one in four reported using three or more methods. Over half the women reported being overwhelmed by the task of hair removal and over 80% reported getting frustrated at the time they spent on hair removal. Soliman and Wardle (2006) report that some hirsute women spend 2 or 3 hours a day using cosmetic or camouflage methods. This type of hair removal can be costly, not particularly effective and can have unpleasant side-effects. Loo and Lanigan (2002) reported that women spend up to £25 per month on mechanical methods of hair removal, with an average spend of almost £7 per month. However according to the Drug and Therapeutics Bulletin (DTB) (2001) the effects last only for days or weeks. Hunter and Carek (2003) report that although shaving is the easiest and safest method, it is often unacceptable to patients. Bleaching products are generally ineffective for dark hair growth, and skin irritation may occur. Chemical depilatories produce results similar to shaving, but again skin irritation is common. Other types of mechanical methods of hair removal have to be carried out by professionals. These treatments, such as electrolysis and laser treatments, have longer-lasting, potentially permanent, effects. Electrolysis aims to destroy the hair root permanently and is regarded by some researchers as one of the most effective and permanent methods of hair removal, which may be an adjunct to hormonal treatment (Hunter and Carek, 2003). Unfortunately it is relatively expensive and time-consuming. Laser treatments also damage the hair follicles but again the cost of repeated treatments limit their use (Soliman and Wardle, 2006). However in a study by Loo and Lanigan (2002) almost all patients reported that laser epilation was more effective than electrolysis or waxing but over 97% of women who underwent laser treatment found their unwanted hair returned to pre-treatment levels after 6 months. Despite this there was a high level of reported patient satisfaction and over three quarters of the women in the study would have undergone further laser treatment. It is estimated that the cost of these treatments is in excess of $1.8 billion in the US and in the UK these treatments are not usually available on the NHS so patients have to bear the burden of cost of these treatments themselves (DTB, 2001). Turning next to medical or pharmacological treatments, these should be prescribed in conjunction with lifestyle changes (primarily weight loss) and any other mechanical methods of hair removal that are acceptable to the patient. Griffing (2007) states that even though a wide variety of pharmacologic strategies are available, it is important to ensure that the patient understands that current systemic therapy is imperfect. The PCDS (2005) have issued guidelines for the treatment of hirsutism in the UK. > Table 107-4 is adapted from their work. In the USA the suggested treatments for hirsutism are outlined in papers by Hunter and Carek (2003) and Griffing (2007). > Table 107-5 is adapted from that work.
Hirsutism and Quality of Life
107
. Table 107-4 Recommendations for pharmacological treatment of hirsutism in the UK Idiopathic or with PCOS
During menopause
Peri-or post-menopaue
The firstline pharmacological treatment of hirsutism is combined cyproterone acetate/ethinylestradiol, 2 mg/ 35 mg daily for women in whom treatment with a combined oral contraceptive is deemed appropriate
Cyproterone acetate alone can be used in combination with HRT. Combined estradiol/ drospirenone (Angeliq) is a new HRT recently launched in the UK, which has anti-androgenic properties
Eflornithine 11.5% cream is licensed for facial hirsutism in postmenopausal women
Eflornithine 11.5% cream is the firstline treatment option for facial hirsutism in women in whom treatment with combined cyproterone acetate/ethinylestradiol is contraindicated due to the risk of VTE or other risk factors, or has not shown efficacy
Eflornithine 11.5% cream is licensed for use in facial hirsutism and therefore may be preferred over spironolactone in women who cannot take HRT
Spironolactone, although not licensed for the treatment of postmenopausal hirsutism, has been found to be effective and safe and is widely used in this setting. However, in the absence of a license, treatment should be initiated by a specialist
If no beneficial effects are seen with eflornithine 11.5% cream after 4 months, discontinue treatment
If no beneficial effects are seen with eflornithine 11.5% cream after 4 months, discontinue treatment Any therapies not licensed for use specifically in hirsutism should be prescribed only by specialists
Combined cyproterone acetate/ethinylestradiol is not suitable for use in postmenopausal women. However, a combination HRT comprising estradiol/ drospirenone, recently launched in the UK, is likely to provide a suitable alternative
PCOS polycystic ovary syndrome; HRT hormone replacement therapy; VTE venous thromboembolism
It is important to note that none of the drugs listed in > Table 107-5 for the treatment of hirsutism have US Food and Drug Administration (FDA) approval for such use and therefore patients must give informed consent to the use of these drugs after a complete explanation of the potential benefits and risks of a particular treatment and alternative approaches. Also the problems with hair growth tend to return when medication is stopped (Hunter and Carek, 2003; BAD, 2006). The extent of the use and effectiveness of these treatments is variable and has been summarized by Griffing (2007). Oral contraceptives are the most widely used treatment to suppress ovarian androgen production. They are inexpensive and can be used in combination with one of the antiandrogens or other forms of therapy. But they should not be used in women with a history of migraines, known or possible thrombotic disease, or breast or uterine cancer.
1831
1832
107
Hirsutism and Quality of Life
. Table 107-5 Recommendations for pharmacological treatment of hirsutism in the USA Class of drug Oral contraceptives
Drug Ethinylestradiol with norgestimate,desogestrel,norethindrone, ethynodiol diacetate Ethinylestradiol with drospirenone
Antiandrogens
Spironolactone (Aldactone) Flutamide (Eulexin) Finasteride (Proscar)
Glucocorticoids
Dexamethasone Prednisone
Gn-RH agonists
Leuprolide (Lupron)
Antifungal agents
Ketoconazole (Nizoral)
Topical hair growth retardant
Eflornithine (Vaniqa)
Insulin-sensitizing agents
Metformin (Glucophage)
The evidence for the antiandrogens is mixed. Spironolactone blocks androgen receptors and decreases testosterone production. Six months to a year of therapy is usually required before results are noticeable and even then, only approximately one half to three quarters of patients show improvement. Flutamide is a nonsteroidal selective antiandrogen without progestational, estrogenic, corticoid, or antigonadotropin activity. Preliminary data indicate that it is effective as therapy for hirsutism. However, flutamide is expensive and has caused fatal hepatitis. Finasteride is a 5-alpha-reductase inhibitor with efficacy similar to that of spironolactone. This drug should only be used in post menopausal women because it can harm male fetuses exposed to it during the early stages of the pregnancy. Glucocorticoids have been used with variable success in women with adrenal hirsutism and metformin improve insulin resistance and have been shown to be effective in lowering androgen levels and in treating hirsutism. But Harborne et al. (2003) suggest that there is insufficient evidence to warrant metformin for first-line treatment for hirsutism. Eflornithine cream, which takes 2–3 months of regular use to have an effect, works by slowing hair growth. It can be applied after any regular hair removal techniques and is left on the skin to inhibit hair growth. According to a study by Davies and Phillips (2007) which evaluated Eflornithine cream, the most important aspects of a successful treatment for hirsutism as reported by hirsute women are the extent to which hair growth is reduced and the speed with which hair growth is reduced. See > Table 107-6.
4
Health Economics
There is very limited evidence as to the cost effectiveness of the treatments available for hirsutism and what does exist is limited to the use of eflornithine compared with placebo.
Hirsutism and Quality of Life
107
. Table 107-6 Mean importance and SD attached to aspects of treatment Women with facial hirsutism
Women without facial hirsutism
Partners of women with facial hirsutism
Cost
5.0 2.1
4.3 1.6
3.9 2.4
Extent to which hair growth is reduced
6.3 1.5
6.5 0.8
5.3 2.1
Speed with which hair growth is reduced
6.3 1.5
6.3 0.9
5.7 1.8
Convenience
4.7 2.2
5.5 1.4
4.3 2.2
Extent to which treatment is invasive
4.2 2.0
5.6 1.3
4.8 2.1
Risks associated with treatment
3.8 2.2
6.2 2.6
5.1 2.6
Aspects of treatment
SD standard deviation. Score of 1 represents not at all important and 7 represents very important
The only published study was undertaken by Davies and Phillips (2007) and aimed to assess:
The cost per QALY gained from using an effective treatment for hirsutism Patients’ valuation of treatment for excess facial hair using willingness-to-pay (WTP) The study was based in England and four GP practices were asked to identify female patients with facial hirsutism from their patient records. They sent questionnaires to those women, to the partners of those women and to a sample of non-hirsute women. The questionnaires were designed to establish the basic characteristics of each of the groups, their perception of health related quality of life (HRQoL), their perception of hirsutism and the importance they attached to various aspects of treatment options for hirsutism. It also sought to measure their current HRQoL and their valuation of a treatment to reduce the effects of excess facial hair. Facially hirsute women were also asked to indicate how much they currently spent (on average per month) on treatments for the removal of facial hair, over and above what was prescribed by their GP. The current HRQoL of respondents was assessed using the EuroQol 5D (> EQ-5D) questions and a visual analogue scale (VAS) (Brooks, 1996). The EQ-5D is recognized as one of the preferred instruments for measuring the health status of individuals (Brazier et al., 1999). Differences between the responses among hirsute and non-hirsute women were taken to reflect the utility gain that would be derived from an effective treatment, eflornithine cream, for the condition. The respondents’ WTP for treatment was used to value the benefits that women perceive that they would derive from the use of eflornithine cream; the method of contingent valuation was used to measure subjects’ WTP. Respondents were asked whether they would be prepared to pay £2, £5, £10, £15, £20, £25, £50, £100 and £150 per month for an effective treatment. No indication of the costs of treatment was provided and they were assured that they would not be expected to pay for treatment with eflornithine cream. Costs associated with the treatment of facial hirsutism were considered to be only the cost of the cream. It was assumed that no any adverse events would occur and no additional
1833
1834
107
Hirsutism and Quality of Life
consultations with the patients’ GPs would result from the treatment. The cost of eflornithine cream was set at £26.04 per 30-g tube, which is for 2 months’ use, which equates to an annual cost of £156.24. This figure was provided by the pharmaceutical company which manufactures the cream (Shire Pharmaceuticals). QALY figures were calculated over a 10-year perspective and were derived from the differences in EQ-5D scores between hirsute and non-hirsute women. The extent to which any treatment was likely to be effective would also impact on the QALYs gained. Data relating to the effectiveness of eflornithine cream, from its clinical studies program (Hamzavi et al., 2003; Smith et al., 2003) were used to adjust the number of QALYs. A series of sensitivity analyses were undertaken, where acquisition costs were adjusted upwards and effectiveness was reduced, to assess the extent to which additional GP consultations or increased adverse events could impact on the findings. In addition, costs and QALYs were discounted by differing rates to assess the rigor of the findings.
4.1
Quality-Adjusted Life Years Generated
The number of QALYs generated by a completely effective treatment for hirsutism over a 10-year period from the survey was estimated at 1.09 (0.92 discounted at 3.5%) based on the EQ-5D tariff and 1.93 (1.6 discounted at 3.5%) based on the VAS scores. However, evidence from the clinical trials program suggested that the incremental success rate of eflornithine cream over vehicle was 20–30%, and the QALYs gained would also reflect such effectiveness rates. Therefore, an effectiveness rate of 25% has been assumed and the resultant QALYs gained from the use of eflornithine cream are, therefore, between 0.27 and 0.48 (nondiscounted), or between 0.23 and 0.40 (discounted at 3.5%), depending on whether the tariff differential or VAS differential is used.
4.2
Cost per QALY Gained from Use of Eflornithine Cream
Assuming an annual treatment cost of £156 for a woman with facial hirsutism, the mean costs over a 10-year period are £1,560 (£1,300 discounted), and the cost per QALY resulting from the use of eflornithine cream for the treatment of hirsutism is therefore, between £3,250 and £5,778 depending on whether the tariff or EQ-5D differential is used.
4.3
Sensitivity Analysis
A series of sensitivity analyses were undertaken, where costs were adjusted upwards and effectiveness reduced. The results are as follows:
Increasing the cost of eflornithine treatment by 100%, that is, assuming that a 30-g tube lasts only 1 month, results in the cost per QALY increasing to between £6,500 and £11,556.
Increasing the cost of treatment by 100% and reducing QALYs to reflect a treatment
effectiveness of 10% results in the cost per QALY increasing to between £16,166 and £28,624. Increasing the cost of treatment by 200% and reducing QALYs to reflect a treatment effectiveness of 10% results in the cost per QALY increasing to between £32,322 and £57,248.
Hirsutism and Quality of Life
107
Therefore the cost per QALY generated by the use of eflornithine cream, and assuming a relatively low rate of effectiveness, is well within the bounds of threshold levels considered to represent value for money (Stevens et al., 1995; Towse and Pritchard 2002) and this continues to apply following the sensitivity analysis, in virtually all cases, where costs are increased by 100% and QALYs are reduced by 50%.
4.4
Willingness-To-Pay
Women with facial hirsutism indicated that, on average, they spent between £22 and £25 per month on treatments for the removal of facial hair, but they would value an effective treatment for the condition at over £40. However partners of the facially hirsute women were prepared to pay nearly £47 per month for such treatment. The use of a monetary measure to value perceived benefits is controversial but does provide some indication of a woman’s estimate (and that of her partner) of what she would be prepared to sacrifice to receive an effective treatment for hirsutism. The study has demonstrated that women with facial hirsutism, their partners and non-hirsute women would value the existence of an effective treatment for hirsutism more than is currently spent on alternatives by women with the condition.
5
Conclusion
Hirsutism in women results in significant psychological and social problems. It impacts negatively upon the quality of life of women and is the cause of stress, anxiety and depression, particularly in cultures where hairlessness is viewed as the norm for women. Facially hirsute women perceive themselves to be in poorer health than non-hirsute women, and there are fewer in employment despite having similar educational experiences and levels of attainment. Women spend much time and effort in removing excess facial and can feel overwhelmed and frustrated by their efforts. All currently available treatments have limited effectiveness and can have serious side-effects. It is therefore a testament to the levels of distress caused by the problem that women persist with treatment despite the difficulties encountered.
. Table 107-7 Key facts of Hirsutism Hirsutism is defined medically as the presence of excessive terminal (coarse) hair that appears in women on the face, back, lower abdomen and inner thighs Hirsutism affects between 5 and 8% of women in the general population Hirsutism in women results in significant psychological and social problems. It impacts negatively upon the quality of life of women and is the cause of stress, anxiety and depression Women can spend as much as 3 h per day trying to remove facial hair. This can be a costly exercise with some women reported spending up to £25 per month on treatments. As a result women can feel overwhelmed and frustrated by their efforts There are two main treatment options: mechanical methods (such as shaving, bleaching and plucking) and pharmacological methods (such as the oral contraceptive pill or eflornithine cream). All currently available treatments have limited effectiveness and can have serious side-effects
1835
1836
107
Hirsutism and Quality of Life
There is very limited evidence as to the cost-effectiveness of the various treatments for hirsutism and there is certainly a requirement for more research in this area. The evidence that does exist indicates that eflornithine cream results in a small QALY gain for patients and provides acceptable value for money. The study utilizing a willingness to pay approach to elicit respondents’ valuation of a treatment for hirsutism, has found that women with facial hirsutism and their partners would value the existence of an effective a treatment more than is currently spent on alternatives by women with the condition. It is essential that research continues to be undertaken into this distressing problem and that clinicians who treat patients with hirsutism deal with them sensitively, acknowledging the major psychological and emotional difficulties these women are facing.
Summary Points Hirsutism is defined medically as the presence of excessive terminal (coarse) hair that
appears in women on the face, back, lower abdomen and inner thighs. Between 5 and 8% of women in the general population are hirsute. Hirsutism in women results in significant psychological and social problems. It impacts negatively upon the quality of life of women and is the cause of stress, anxiety and depression. It can affect educational performance and the ability to work. Women can spend a significant amount of time and money trying to remove facial hair and they can be left feeling overwhelmed and frustrated by their efforts. There are two main treatment options: mechanical methods (such as shaving, bleaching and plucking) and pharmacological methods (such as the oral contraceptive pill or eflornithine cream). All currently available treatments have limited effectiveness and can have serious side-effects.
References Azziz R. (2003). Obstet Gynecol. 101: 995–1007. Azziz R, Carmina E, Sawaya ME. (2000). Endocr Rev. 21: 347–362. Azziz R, Sanchez LA, Knochenhauer ES, Moran C, Lanzenby J, Stephens KC, Taylor K, Boots LR. (2004). J Clin Endocrinol Metab. 89: 453–462. Balen A. (1999). Lancet. 354: 966–967. Brazier J, Deverill M, Green C. (1999). J Health Serv Res Policy. 4: 174–184. British Association of Dermatologists. (2006). Available at: http://www.bad.org.uk/public/leaflets/hirsutism. asp (Accessed: February 20, 2008). Brooks R. (1996). Health Policy. 37: 53–72. Ching HL, Burke V, Stuckey BG. (2007). Clin Endocrinol. 66(3): 373–379. Clayton WJ, Lipton M, Elford J, Rustin M, Sherr L. (2005). Br J Dermatol. 152: 986–992.
Coffey S, Mason H. (2003). Gynecol Endocrinol. 17: 379–386. Davies S, Phillips CJ. (2007). J Med Econ. 10: 107–118. Deplewski D, Rosenfield RL. (2000). Endocr Rev. 21: 363–392. Drug and Therapeutics Bulletin (DTB). (2001). 39: 1–5. Ferrante J. (1998). Cult Med Psychiatry. 12: 219–238. Finlay AY, Khan GK. (1994). Clin Exp Dermatol. 19: 210–216. Givens JR, Kurtz BR. (1992). Hirsutism, virilization, and androgen excess. In: Hurst JW, Ambrose SS, et al. (eds.) Medicine for the Practicing Physician, 3rd ed. Butterworth-Heinemann, Boston, pp. 568–571. Goodheart HP, Uyttendaele HI. (2008). Available at: http://www.emedicine.com/derm/topic472.htm (Accessed: February 20 2008).
Hirsutism and Quality of Life Griffing GT. (2007). Available at: http://www.emedicine. com/med/topic1017.htm (Accessed: February 20 2008). Hahn S, Janssen OE, Tan S, Pleger K, Mann K, Schedlowski M, Kimmig R, Benson S, Balamitsa E, Elsenbruch S. (2005). Eur J Endocrinol. 153(6): 853–860. Hamzavi I, Tan E, Shapiro J, Lui H. (2003). Lasers Surg Med. 15: 32. Harborne L, Fleming R, Lyall H, Norman J, Sattar N. (2003). Lancet. 361: 1894–1901. Hopkinson ZEC, Sattar N, Fleming R, Greer IA. (1998). Br Med J. 317: 329–332. Hunter MH, Carek PJ. (2003). Am Fam Physician. 67 Available at: http://www.aafp.org/afp/20030615/ 2565.html (Accessed: February 20, 2008). Keegan A, Liao LM, Boyle M. (2003). J Health Psychol. 8:327–345. Kitzinger C, Wilmott J. (2002). Soc Sci Med. 54: 349–361. Knochenhauer ES, Key TJ, Kahsar-Miller M, Waggoner W, Boots LR, Azziz R. (1998). J Clin Endocrinol Metab. 83: 3078–3082. Lipton MG, Sherr L, Elford J, et al. (2006). J Psychosom Res. 61:161–168. Loo WJ, Lanigan SW. (2002). Clin Exp Dermatol. 27: 439–441. Moghetti P, Toscana V. (2006). Best Pract Res Clin Endocrinol Metab. 20: 221–234. National Horizon Scanning Centre: New and Emerging Technology Briefing. (2001). Available at:
107
http://www.pcpoh.bham.ac.uk/publichealth/horizon/ PDF_files/2001reports/Eflornithine.PDF (Accessed February 20 2008). Primary Care Dermatology Society (PCDS). (2005). Guidelines: Medical Management of Hirsutism. Available at: (http://www.evidence-based-medicine. co.uk/guidelines.html) Reingold SB, Rosenfield RL. (1987). Arch Dermatol. 123: 209–212. Rosenfield RL. (2005). N Engl J Med. 353: 2578–2588. Smith SR, Piacquadio D, Beger B. (2003). Abstract from the 61st Annual Meeting of the American Academy of Dermatology. San Francisco, CA. Soliman N, Wardle P. (2006). Available at: http:// student.bmj.com/issues/06/10/education/360.php (Accessed: February 20, 2008). Sonino N, Fava GA, Mani E, et al. (1993). Postgrad Med J. 69: 186–189. Stevens A, Colin-Jones D, Gabbay J. (1995). Health Trends. 27: 37–42. Toerien M, Wilkinson S. (2003). Womens Stud Int Forum. 26: 333–344. Towse A, Pritchard C. (2002). Does NICE have a threshold? An external view. In: Costeffectiveness Thresholds: Economic and Ethical Issues. Office of Health Economics, London, UK. Walling AD. (2004). Am Fam Physician. 69. Available at: (http://www.aafp.org/afp/20040115/tips/17.html) (Accessed: February 20, 2008).
1837
108 Oral Health-Related Quality of Life U. Schu¨tte . M. Walter 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1840
2
Oral Diseases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1841
3
Oral Health in the Past . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1843
4 Broader Understanding of Oral Health . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1844 4.1 Definition of Oral Health . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1844 4.2 Oral Health and General Health . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1845 5
Oral Health and Quality of Life . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1845
6 6.1 6.2 6.3
Oral Health Related Quality of Life . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1846 Measuring Oral Health Related Quality of Life . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1847 Instruments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1849 Data on OHRQoL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1851 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1851
#
Springer Science+Business Media LLC 2010 (USA)
1840
108
Oral Health-Related Quality of Life
Abstract: The increasing life expectancy in developed countries is associated with growth in the incidence of chronic diseases such as obesity, diabetes, and > dental caries. These developments have implications for decrements in quality of life related to general health as well as > oral health. Consequently, medical care and social programs are focusing increasingly upon enhancement of patients’ quality of life. Together with the extension of people’s lifespan, the World Health Organization (WHO) ascertained the enhancement of the quality of life as a central goal of health care systems. With respect to oral health, treatment of oral diseases is extremely costly and therefore a significant economic burden for many industrialized countries. Hence, in order to strengthen dental public health programs and set priorities for investment in oral healthcare-related research and treatments, planners have shown growing interest in quantifying the consequences of oral diseases that affect individuals’ functioning, comfort, and ability to perform everyday activities. The concept of quality of life is multidimensional including health as one dimension. Quality of life is influenced by several personal and social expectations. By definition, measurement of health should cover not only professionals’ views but also patients’ preferences, which reflect their experiences and concerns. In studies of the association between objective measures of dental disease and patient-based ratings of oral health status, objective measures did not accurately indicate patients’ perceptions. Such measures reflect the endpoint of a specific disease process but give no indication of the impact of the disease and its course on functioning and psychosocial well-being. Hence, there is a need for the development of appropriate measures to assess quality of life with regard to oral health and oral diseases. Subjective measures provide an important adjunct to objective clinical measures for determining goals in health care and assessing how well the goals have been met. List of Abbreviations: CAPP, WHO oral health country/area profile program; CPITN, > community periodontal index of treatment need; DIDL, dental impact on daily living; DIP, dental impact profile; > DMF(T)-Index, decayed missing filled teeth index; FDI, Fe´de´ration Dentaire International; GOHAI, geriatric oral health assessment index; HRQoL, health related quality of life; IADR, International Association for Dental Research; OHIP, oral health impact profile; OHIP-ADD Score, oral health impact profile-additive score (of all item responses); OHRQoL, > oral health related quality of life; OH-QOL UK, oral health related quality of life (United Kingdom); QoL, quality of life; SF36, (health survey) short form 36; VAS, visual analogue scale; WHO, World Health Organization
1
Introduction
The increasing life expectancy in developed countries is associated with growth in the incidence of chronic diseases such as obesity, diabetes, and dental caries. These developments have implications for decrements in quality of life related to general health as well as oral health. Both aspects of quality of life may be deteriorating; thus, medical care and social programs are focusing increasingly upon enhancement of patients’ quality of life. Together with the extension of people’s lifespan, the World Health Organization (WHO) ascertained the enhancement of the quality of life as a central goal of health care systems (Reid, 1985). In the achievement of those goals outcome research as a part of > health services research serves an essential role by identifying treatments that produce the best outcomes for patients, evaluating ways in which health care can be organized to optimize benefits for communities, and supporting the
Oral Health-Related Quality of Life
108
development of health care policy at the local and international level through the dissemination of information (Slade et al., 1998). In addition, with respect to oral diseases, the usual methods of treatment are extremely costly and incur a significant economic burden for many industrialized countries. Approximately 5–10% of public health expenditures relate to oral health; oral diseases are the fourth most expensive type of disease to treat in the most developed countries (Peterson et al., 2005). Hence, there has been growing interest in quantifying those consequences of oral disease that affect individuals’ functioning, comfort, and ability to perform everyday activities (Slade and Spencer, 1994). In order to strengthen dental public health programs and set priorities for investment in oral healthcare-related research and treatments, planners increasingly have taken into account patients’ quality of life. The concept of quality of life is multidimensional. It includes health as one of the dimensions. Quality of life is influenced by several personal and social expectations. By definition, measurement of health status should include not only professionals’ views but also patients’ perceptions, which reflect their experiences and concerns. In studies that have assessed the association between objective measures of dental disease and patient-based ratings of oral health status, researchers pointed out that objective measures did not accurately reflect patients’ perceptions. Focusing on the physical structures and processes associated with the mouth, clinical indices are of little use when investigators try to infer the impact of dental status on patients’ lives. Hence, there is a need for the development of appropriate measures to assess quality of life with regard to general health as well as oral health. This chapter wants to give a review about the main measures in terms of > oral health related quality of life (OHRQoL).
2
Oral Diseases
Dentistry is the science concerned with the prevention, diagnosis, and treatment of diseases of the teeth, gums, and related structures of the mouth; it includes all treatments for restoring oral health or restoring a condition that enables an individual to lead a ‘‘socially and economically productive life in a state of complete physical, mental and social well-being’’ (World Health Organization, 2001). Oral diseases comprise a range of conditions, from common diseases such as dental caries and > periodontal diseases to life-threatening illnesses such as oral cancer. Dental caries – meaning the irreversible destruction of the proper substance of the tooth due to metabolic acids produced by the oral flora – is still a major oral health problem in most industrialized countries. However, over the past 20 years a decline of caries has been observed as a result of public health measures and dental care together with changing lifestyles and improved self-care practices. In 1981 the World Health Organization (WHO) and the Fe´de´ration Dentaire International (FDI) formulated the first Global Oral Health Goals to be achieved by the year 2000. For the 12 years of age the global average was committed to be no more than 3 Decayed, Missing and Filled teeth (DMFT-Index, Klein et al., 1938; also see ‘‘Oral health in the past’’). In 2001 the global weighted mean DMFT value for this age group was reported to be 1.74 continuing to decrease (global weighted mean DMFT = 1.61 in 2004) (Peterson et al., 2005; > Figure 108‐1). Nevertheless, between 60 and 90% of all schoolchildren and a vast majority of adults still suffer from tooth decay in developed countries and some Asian and Latin American states (Petersen, 2003). There are still countries with a high rate of caries, for instance, Lebanon (DMFT score of 3.4 in 2000) and Guatemala (DMFT score of 5.2 in 2002; CAPP, 2006). Only 74% of the nations reached the formulated goal, representing 86% of the world population
1841
1842
108
Oral Health-Related Quality of Life
in 2004. Also in most developing African countries, where the situation is not that severe so far, the incidence of dental caries is predicted to increase due to changing living conditions (particularly as a result of growing sugar consumption). Therefore, it must be emphasized that dental caries among children has not been eradicated. It has only been brought under control in some countries (Peterson et al., 2005). Together with periodontal diseases – a group of inflammatory diseases of the gum – dental caries is the most important global oral health burden and among the most frequent chronic diseases in the world (> Table 108‐1).
. Figure 108‐1 Changing levels of caries experience (decayed, missing and filled teeth (DMFT-Index)) among 12-year-olds in developed and developing countries (Peterson et al. (2005) The global burden of oral diseases and risk to oral health (reprinted with permission from april 2008))
. Table 108‐1 Key facts of oral health Key facts of oral health Dental caries and periodontal diseases are the most important global oral health burden Dental caries and periodontal diseases are among the most frequent chronic diseases in the world Dental caries affects 60–90% of school-aged children worldwide There are still countries that suffer from a high rate of caries Dental caries among children has only been brought under control in some countries Traditional methods of treatment are extremely costly and incur a significant economic burden for many industrialized countries, where approximately 5–10% of public health expenditures relate to oral health
Oral Health-Related Quality of Life
108
However, a review of the first Global Oral Health Goals showed that they have been useful and, for many populations, had been achieved or exceeded (Hobdell et al., 2003). They seem to stimulate the awareness of the importance of oral health amongst national and local governments. Hence, the FDI, WHO and IADR (International Association for Dental Research) established a new document that specified realistic goals and standards on oral health to be achieved by the year 2020. Oral health goals provide a helpful orientation guide when trying to raise the level of health.
3
Oral Health in the Past
For most of its history, dentistry was limited to clinical issues in its conception and understanding of oral health. This perspective arose from the fact that researchers themselves assumed that the consequences of dental and oral conditions primarily were either nonexistent or negligible. Indeed, oral diseases were accepted as widespread in the population, frequently asymptomatic, and not giving rise to observed changes in behavior compared with more serious diseases. Apart from pain and life-threatening cancers, they were considered as not having any impact on social life and only linked with cosmetic issues (Davis, 1976). For example, some investigators have claimed that work loss due to dental disorders is negligible when compared with that caused by major chronic or disabling disorders such as cancer (Reisine and Miller, 1986). Accordingly, disease-based measurements were generated in order to document the oral health status of a population. While they assess the presence and severity of pathological conditions, these clinical indices neglect the impact of such conditions on functioning and social well-being. They only reflect the end-point of a specific disease process. For instance, one clinical index focuses on the number of decayed, missing and filled teeth per person (DMF(T)-Index; Klein et al., 1938) as measure of the caries experience of individuals, groups or a population. Every tooth with caries history is assigned to one of the following categories (> Table 108‐2). . Table 108‐2 DMFT-index (Klein et al., 1938) D
Decayed
M
Missing
F
Filled
Tooth with a cavity (decay) Tooth is missing (because of caries) Tooth with a filling (has had a cavity)
Although the DMF(T)-Index has been used universally since its introduction and has received remarkably little challenge, its limitations need to be recognized. Measuring the lifetime caries experience of an individual, the DMF is invalid when teeth have been lost for reasons other than caries. For example, in higher ages teeth can be lost for periodontal reasons and among teenagers for orthodontic reasons (Burt and Ekklund, 2005). Hence, in these age groups results have to be analyzed more carefully. Other indices, like periodontal indices are based on pocket depth, gum inflammation, and unmet treatment needs (e.g., CPITN; Ainamo et al., 1982). According to this index the mouth is divided into sextants, each sextant having special index teeth. The index teeth are examined and assigned to a certain score. The highest score per sextant is recorded (> Table 108‐3).
1843
1844
108
Oral Health-Related Quality of Life
. Table 108‐3 CPITN-index (Ainamo et al., 1982) Code
Status
Sign for
0
Healthy gingival and periodontal conditions
Healthy conditions
1
Gingival bleeding on probing
Gingivitis
2
Calculus or iatrogenic marginal imperfections and gingival bleeding Gingivitis
3
Depth of the gingival pocket 4–5 mm
Periodontitis
4
Depth of the gingival pocket 6 mm and above
Severe Periodontitis
4
Broader Understanding of Oral Health
In the middle of the last century, views shifted regarding the perception of health. Newer definitions of health demand a broader understanding than previously was popular. One of the first definitions that implied the multidimensional nature of health was the 1946 WHO definition of health, which stated that, ‘‘Health is a state of complete physical, mental and social well-being and not merely the absence of disease or infirmity.’’ This definition was strongly reaffirmed during the international conference on primary health care in Alma Ata in 1978 (WHO, 1978). As a consequence of this broader comprehension, the traditional medical model of health needed to be expanded by incorporating social aspects of health.
4.1
Definition of Oral Health
In dentistry, in the early 1980s, there also was a paradigm shift towards accepting the fact that dental and oral diseases do impose a significant burden on the individual and the community (Cushing et al., 1986; Locker and Grushka, 1987; Reisine, 1988). Studies revealed the impact of dental pain and functioning on everyday life: for example, a cross-sectional survey in London (UK) showed that one in six 8-year-olds had experienced toothache that caused them to cry (Shepherd et al., 1999). In another study of a sample of Thai primary school children, nearly three-fourths of the children were restricted in eating; among about one-fourth of those children, such restrictions had mainly clinical causes such as toothache and sensitive teeth (Gherunpong et al., 2004). Thus, severe caries detracts from children’s quality of life; they experience pain, discomfort, acute infections, eating and sleeping disruption. Moreover, toothache leads to school absence. In addition, it is evident that oral diseases restrict activities, not only in school but also at work and at home, causing millions of lost school and work hours worldwide each year (Reisine, 1988). Within 74% of a Thai population sample aged 35–44-years, daily performance was affected by oral state. Emotional stability was one of the performances highly affected (46.5%), whereas toothache was the major causal oral condition (32.7%; Adulyanon et al., 1996). To summarize, oral health influences how people grow, enjoy life, look, speak, chew, taste food and socialize, as well as their feelings of social well-being (Locker, 1997). As a consequence of this awareness of the impact of dental diseases, oral health has been more broadly defined as
Oral Health-Related Quality of Life "
108
the ability to chew and eat the full range of foods native to the diet, to speak clearly, to have a socially acceptable smile and dentofacial profile, to be comfortable and free from pain, and to have fresh breath (Sheiham and Spencer, 1997).
Therefore, oral health refers to more than just having healthy teeth.
4.2
Oral Health and General Health
Additionally, over the years growing evidence suggests that oral health is integral to general health and that the relationship between the two dimensions is interactive. Severe periodontal disease, for example, increases the risk of a cardiovascular disease. In turn, the condition of the gums can be influenced by an individual’s general health status; for example, severe periodontal disease is considered the sixth most frequent complication of diabetes. Due to use of medications – most notably in older adults – reduced salivary flow rates may occur; such reduced salivary flow is related to dental caries via restriction of the protective role of the saliva against caries. This protective role consists of simple dilution, buffering plaque acids, and being a source of minerals and immunological plaque inhibitory factors (Flink, 2007). Furthermore, the role of bacteremia as a sequel of oral disease and its treatment is well known in the etiology of bacterial endocarditis. Moreover, it has became apparent that within the context of the wider socio-environmental milieu, oral and other chronic diseases share common determinants (e.g., unhealthy diet, smoking, harmful alcohol abuse). Therefore, more and more emphasis has been placed on the > Common risk/health factor approach, which means promoting health by controlling a small number of risk factors that have a major impact on a large number of diseases. This approach leads to lower costs as well as greater efficiency and effectiveness than disease-specific approaches (Sheiham and Watt, 2000). All things considered, the impact of oral diseases is not limited to the oral cavity. Oral health affects people’s lives physically (producing a significant impact on people’s general health) and on a psychosocial level. As a result, poor oral health often diminishes our quality of life.
5
Oral Health and Quality of Life
Quality of life (QoL) has been defined by the World Health Organization as ‘‘an individual’s perception of their position in life in the context of the culture and value systems in which they live and in relation to their goals, expectations, standards and concerns. It is a broad ranging concept incorporating in a complex way the person’s physical health, psychological state, level of independence, social relationships, personal beliefs and their relationship to salient features of the environment’’ (WHO, 1997). According to this definition, the concept of QoL is multidimensional and brings together the wide-ranging aspects of health status. The QoL concept created a demand for new health status measures in contrast to clinical measures of disease status, and a demand for the involvement of patient-based assessments. Knowledge of the patient’s attitude towards his or her own state of (oral) health is of great value. It is the motivation to seek dental care and influences patient’s satisfaction with respect to the treatment received. Furthermore, patient-based assessments offer investigators insight into patients’ own views of their (dental) care needs. Involving the individual patient’s perspective can lead to greater levels of cooperation, higher degrees of patient satisfaction,
1845
1846
108
Oral Health-Related Quality of Life
and more effective use of available resources. Hence, in recent decades the person’s own perception of health (patient centeredness) has become more important when trying to raise the level of health. All in all, improvement of QOL can be regarded as one of the main goals of medical interventions.
6
Oral Health Related Quality of Life
In former times, aspects of the patient’s awareness of his or her health status and quality of life were integrated into the practice of medical care and were not considered as something independent. During treatment, patients’ subjective awareness of health status was looked at rather broadly by asking the patient about his or her current health state (e.g., ‘‘Are you comfortable today?’’). Studies have revealed, however, that in dentistry objective diagnostic findings often differ from the patients’ subjective views and that objective diseases state not a good predictor for care seeking (Heydecke et al., 2003a). Several studies, conducted within the frame of health needs assessment as one part of health services research, detected in many cases a great discrepancy between the professional’s diagnostic findings and the patient’s awareness regarding their oral health status (Giddon et al., 1976; Heydecke et al., 2003a; Reisine and Bailit, 1980; Smith and Sheiham, 1980; Walter et al., 2001). Looking, for instance, at the need of prosthetic treatment, Smith and Sheiham found a wide discrepancy between the normative and perceived needs of an elderly population. Only 42% of those who were clinically assessed as needing treatment felt that they required it and only 19% had actually tried to obtain it. Many of the elderly mentioned a number of barriers to obtaining dental care; such barriers included the cost of treatment, fear of the dentist, immobility and the feeling that they should not ‘‘bother’’ the dentist. Similarly, Walter et al. found a remarkable difference in the need for prosthetic treatment within a representative sample of the over 14-year-old population in the German federal state of Saxony. Walter et al. reported that 580 out of 714 participants (81%) had objective prosthetic needs compared with 13% who clearly indicated subjective demand (> Figure 108‐2). Furthermore, Heydecke at al. pointed out that no clinical variables were significantly correlated with patient satisfaction before or after treating edentulous patients with mandibular implant or conventional prostheses. They concluded that clinicians’ assessments of the quality of denture-supporting tissues are poor predictors of patient satisfaction with the kind of prosthesis supplied. Moreover, other investigations demonstrated that despite the fact that most professionals believe that restorations after tooth loss in the posterior region are necessary, patients of different age groups (even in countries with highly developed oral health care systems) do not share this opinion. Patients often accept these open spaces in the > premolar and > molar regions and want to leave them untreated. This observation supports the idea that clinically judged need for treatments cannot always be translated into demands for care. In addition, the increasing life expectancy forces health professionals more and more to find out therapies which preserve or even enhance the patient’s quality of life effectively in order to keep the patients well and fit over a long period of their life (reduction of morbidity). The interest in gaining more accurate knowledge regarding patients’ awareness of health calls for information about the functioning of the person as a whole, including subjectively perceived symptoms, and not merely about the oral cavity.
Oral Health-Related Quality of Life
108
. Figure 108‐2 Normative and subjective need and their combinations (Walter et al., 2001)
6.1
Measuring Oral Health Related Quality of Life
Due to this interest, the measurement of Oral Health-related Quality of Life (OHRQoL) as a subset of Health-related Quality of Life (HRQoL) became accepted as an explicit criterion of evaluation being an indicator of patient’s well-being and the quality of health care services (> Figure 108‐3). The general Health Related Quality of Life (HRQoL) notion emerged in the late 1960s and was based on the expanded definition of health as the key focus. When trying to quantify the consequences of oral diseases, it has been recognized that objective measures of disease were not entirely suitable to capture the new concept of health, particularly the aspects of mental and social well-being. Despite the same clinical state, there can be differences in patients’ perceptions of health status. In South Australia, for example, a comparison between HIV-infected patients and general patients, all receiving publicly funded care, showed that pain, functional limitations, and social disability occurred significantly more often in HIV-infected patients (Coates et al., 1996). This result indicates the limitation of clinical somatic components, which are not able to represent a person’s overall oral health status. A search for more appropriate concepts and indicators to identify the multifactorial impacts on quality of life has begun. The OHRQoL, which cannot be observed directly, needs to be visualized by means of suitable indicators. Based on self-ratings it therefore comprises the (1) patients’ perceptions of their illnesses/health, (2) the functional, psychological, social and economic implications, (3) the resulting limitations to their daily activities, and (4) the patients’ assessment of the success of the therapy. At first, existing generic measures and conceptual models of health and HRQoL, (for example, the SF36) were used as a basis in order to measure OHRQoL.
1847
1848
108
Oral Health-Related Quality of Life
. Figure 108‐3 (Oral) Health as one domain of Quality of life (John and Micheelis, 2003) (John and Micheelis (2003) Mundgesundheitsbezogene Lebensqualita¨t in der Bevo¨lkerung: Grundlagen und Ergebnisse des Oral Health Impact Profile (OHIP) au seiner repra¨sentativen Stichprobe in Deutschland (reprinted with permission from april 2008))
But studies pointed out that these instruments were not suitable because they were not sensitive to all oral health problems (e.g., effects of tooth loss and edentulism; see also Heydecke et al., 2003b). As a consequence, a number of new models and measurement instruments specific to OHRQoL have been developed. In 1995, Gift and Atchison developed a multidimensional conceptual model of OHRQoL based on the structure of the HRQoL model proposed by Patrick and Erickson (Gift and Atchison, 1995; Patrick and Erickson, 1993). Similar to that, Locker earlier developed a model for oral health in 1988 using the framework of the World Health Organization (WHO) International Classification of Impairments, Disabilities and Handicaps. Describing the consequences of disease, he delineated the dimensions that have to be involved when generating appropriate (oral) health measures (Locker, 1988; > Figure 108‐4). According to his conceptual framework, disease can lead to impairment, defined as any anatomical loss or abnormality. Impairment in turn may lead to functional limitations like such as problems in uttering certain sounds after the loss of a tooth. Besides functional restrictions, pain and discomfort – either physical or psychological – could appear as consequences of impairment. Both functional limitations and discomfort may give rise to physical, psychological or social disability. Locker characterized disability as any limitation in or lack of ability to perform activities of daily living. In case of the missing tooth, this disability could be the negative experience of being misunderstood during conversation due to mispronunciations. Disability then may lead to handicap as the final consequence. Disability is more likely to occur when both discomfort and functional limitation exist, and handicap is more probable if all three have occurred. However, as Locker pointed out, the linking arrows in > Figure 108‐4 should not be looked upon as something necessary. In general, his model of oral health was often used to identify conceptual domains in the hierarchy of social impact (Slade and Spencer, 1994). In general, patient based measures of oral health status are significant in dental education, dental research, and clinical practice and therefore can be used for
Screening and monitoring for psychosocial problems in individual patient care Clinical trials
Oral Health-Related Quality of Life
108
. Figure 108‐4 The conceptual model of (oral) health. Locker D (1988) Measuring Oral Health: A Conceptual Framework (reprinted with permission from april 2008)
Population surveys of perceived health problems Medical audits Cost-utility and cost-effectiveness analyses Outcome measures in health services or evaluation research.
6.2
Instruments
Fundamentally, there are three categories of OHRQoL measurement, as indicated by Slade: (1) social indicators, (2) global self-ratings of OHRQoL, and (3) multiple-item questionnaires on OHRQoL (Slade, 2002). Social indicators look at the effect of oral conditions at the community level, and are therefore meaningful to policy-makers. In large population surveys, for instance, social indicators are used to express the burden of oral diseases due to work loss, days of restricted activities, and school absence. The second category, global self-ratings of OHRQoL, is also known as single-item ratings. Upon being asked a general question about one’s oral health, the person answers a global question such as ‘‘How do you rate your oral health today?’’ by using either a categorical or visual analogue scale (VAS) format. The most widely used method to assess OHRQoL is the multiple-item questionnaire. Based on the concept of (oral) health, researchers have developed a number of instruments to measure oral-specific health status on a broader basis (refer to > Table 108‐4 for examples). The available questionnaires vary widely in terms of the number of questions/items, format of questions, and response options. As an example, the Dental Impact Profile/25, which looks at appearance, eating, speech, confidence, happiness, social life, and relationships, has 25 questions and three response categories (good effect, bad effect, and no effect). However, within the Sociodental Scale/14, the patient can choose only between yes or no when answering the 14 questions related to chewing, talking, smiling, laughing, pain, and appearance (Cushing et al., 1986; Strauss and Hunt, 1993). The best documented and most popular instrument for measuring OHRQoL is the Oral Health Impact Profile (OHIP; Slade and Spencer, 1994). Forty-nine unique statements that describe the impact of oral disorders were initially derived from 535 statements obtained in
1849
1850
108
Oral Health-Related Quality of Life
. Table 108‐4 Examples of currently available oral specific health status measures Authors
Name of measure
Cushing et al. (1986)
Social impacts of dental diseases
Atchison and Dolan (1990)
Geriatric oral health assessment index (GOHAI)
Strauss and Hunt (1993)
Dental impact profile (DIP)
Slade and Spencer (1994)
Oral health impact profile (OHIP)
Locker and Miller (1994)
Subjective oral health status indicators
Leao and Sheiham (1996)
Dental impact on daily living (DIDL)
Adulyanon and Sheiham (1997)
Oral impacts on daily performances
McGrath and Bedi (2001)
OH-QOL UK
Source: Allen (2003) Health and quality of life outcomes
interviews with 64 dental patients. All of the statements were transferred into questions and assigned to one of the conceptually formulated subscale specified by Locker: 1. 2. 3. 4. 5. 6. 7.
Functional limitation, Physical pain, Psychological discomfort, Physical disability, Psychological disability, Social disability, and Handicap.
Subjects are asked how frequently they have experienced the particular impacts. Responses are graded on a Likert-type scale, offering the person five different response options (never = 0, hardly ever = 1, occasionally = 2, fairly often = 3, very often = 4). The additive score of all item responses (OHIP-ADD score) ranges from 0 to 196 and rises with the level of impairment. A major advantage of this measure is that the statements were derived from a representative patient group, and were not conceived by dental research workers. The OHIP questionnaire is a reliable and valid measuring instrument and has well documented psychometric properties. It has been used extensively in population based studies and clinical research, and in general populations as well as patients with specific oral disorders such as orofacial pain (Murray et al., 1996). Moreover, the questionnaire is available in different languages including English, French, Spanish, German, and Chinese. In recent years Hungarian, Spanish, Japanese and Arabic versions have been generated (Al-Jundi et al., 2007; Lopez and Baelum, 2006; Szentpetery et al., 2006; Yamazaki et al., 2007). The development of language specific questionnaires is important, because the perception of quality of life has a subjective component and therefore can vary from one culture to another. Also short versions of this instrument with 20 and 14 items have been published (Allen and Locker, 2002; Awad et al., 2000; Slade, 1997). The number of questionnaires to assess oral health continues to grow, mainly because of the demand of more specific measures such as those for dental anxiety (McNeil and Rainwater, 1998), and conditions such as dentofacial deformity (Cunningham et al., 2000). Moreover, some questionnaires can also target the assessment of specific population groups such as children (Jokovic et al., 2002). Hence, measurements can be divided into generic and specific ones.
Oral Health-Related Quality of Life
6.3
108
Data on OHRQoL
Whereas demographic variables like gender, age and education are rated different in their influence on OHRQoL, (removable) denture status seems to be a generally accepted strong predictor of impaired quality of life in terms of oral health. This was found, for instance, in a large national population-based survey conducted in Germany (John et al., 2003). Using the OHIP-G as the measuring instrument, John et al. pointed out that people without dentures had less score points than people with removable denture. Highest score points were found in subjects with complete denture highlighting the value of natural dentition in terms of quality of life. These findings are in accordance with results by Walter et al. (2007). Examining determinants of oral health-related quality of life in a cross-cultural German-Canadian sample, they detected removable denture wearing as a significant explanatory variable. In addition, impaired OHRQoL was more frequently in persons with treatment need in oral surgery, endodontics and prosthetic dentistry. Looking at the oral health and treatment need among older adults, Locker and Slade ascertained that the impact of oral conditions is not only influenced by dental status but also by dental visiting patterns (Locker and Slade, 1993). Using the OHIP in their evaluation, they concentrated on the subjects responding ‘‘fairly often’’ and ‘‘very often’’ to one or more items in each subscale. Generally two-fifth (43.5%) of the participants were frequently bothered by functional consequences of oral disorders and slightly more than one-fifth (21.2%) were frequently bothered by oral pain. Slightly less than one-fifth (17.3%) experienced some form of psychological discomfort because of poor oral health. These findings indicate that oral conditions have a significant influence on daily living in this population. Analyses additionally showed that edentulous people were more likely to report problems than dentate. Only referring to those who retain natural teeth, patients who regularly visit the dentist showed less problems in all subscales than those who see the dentist only in case of pain and other trouble. Most demographic factors as age and education were found to have lesser significance compared to factors like denture status (John et al., 2004; Walter et al., 2007). Only gender seems to have an influence on OHRQOL: in several studies women had significant higher OHIP-ADD scores than men (e.g., Walter et al., 2007). Generally, it is worth mentioning, that objective measures of dental disease (such as the presence of dental caries or periodontal disease) and patient based measures are not conflicting but complement one another in collecting clinical findings, diagnostic data, in formulating the aim of the therapy and in assessing the therapeutic success. Clinical indicators remain an essential component of oral assessment, particularly among children and young adults, for whom elimination of disease is often both possible and beneficial (Corson et al., 1999). However, in relation to attempts to raise the level of health of groups of people, the OHRQoL is an essential component of the assessment of outcomes of oral health care, as it captures the results of public health programs. Within the context of oral health, some authors consider the improvement of Oral Health-Related Quality of Life (OHRQoL) to be the most important contribution of dentistry.
Summary Points Oral health affects people’s life physically and on a psychosocial level, has a significant impact on peoples’ general health and often diminishes our quality of life. Therefore, oral health refers to more than just having healthy teeth.
1851
1852
108
Oral Health-Related Quality of Life
As a consequence of the awareness of the impact of dental diseases, oral health has been
more broadly defined as ‘‘the ability to chew and eat the full range of foods native to the diet, to speak clearly, to have a socially acceptable smile and dentofacial profile, to be comfortable and free from pain, and to have fresh breath.’’ Trying to quantify the consequences of disease, it has been recognized that objective measures of disease were not entirely suitable to capture the new concept of health, particularly the aspects of mental and social well-being. Indicating the pathophysiology of oral diseases, they fail to take account of the subjective perspective. A patient based assessment of (oral) health status is essential to the measurement of health. Social indicators, global self-ratings of OHRQoL and multiple-item questionnaires on OHRQoL are categories of OHRQoL measurement. The best documented and most popular instrument for measuring OHRQoL is the Oral Health Impact Profile (OHIP; Slade and Spencer, 1994).
References Adulyanon S, Vourapukjaru J, Sheiham A. (1996). Community Dent Oral Epidemiol. 24(6): 385–389. Adulyanon S, Sheiham A. (1997). In Slade G., editor. Measuring Oral Health and Quality of Life. Chapel Hill: University of North Carolina: Dental Ecology. Ainamo J, Barmes D, Beagrie G, Cutress T, Martin J, Sardo-Infirri J. (1982). Int Dent J. 32(3): 281–291. Al-Jundi MA, Szentpetery A, John MT. (2007). Int Dent J. 57(2): 84–92. Allen PF. (2003). Health Qual Life Outcomes. 1: 1–8. Allen JF, Locker D. (2002). Int J Prosthodont. 15: 446–450. Atchison KA, Dolan TA. (1990). J Dent Educ. 54(11): 680–687. Awad MA, Locker D, Korner-Bitensky N, Feine JS. (2000). J Dent Res. 79: 1659–1663. Burt AB, Ekklund SA. (2005). In: Burt AB, Ekklund SA (eds.) Dentistry, Dental Practice and the Community. WB Saunders, St. Louis Missouri, pp. 194–202. CAPP. (2006). WHO Oral Health Country/Area Profile Programme. Available at: http://www.whocollab.od.mah.se/index.html (accessed on 20 April 2008). Coates E, Slade GD, Goss AN, Gorkic E. (1996). Aust Dent J. 41: 33–36. Corson MA, Boyd T, Kind P, Allen PF, Steele JG. (1999). Br Dent J. 187(9): 481–484. Cunningham SJ, Garratt AM, Hunt NP. (2000). Community Dent Oral Epidemiol. 28(3): 195–201. Cushing AM, Sheiham A, Maizels J. (1986). Community Dent Health. 3(1): 3–17. Davis P. (1976). Sci Med. 10: 329–335. Flink H. (2007). Studies on the prevalence of reduced salivary flow rate in relation to general health and
dental caries, and effect of iron supplementation. Swed Dent J Suppl. (192): 3–50. Gherunpong S, Tsakos G, Sheiham A. (2004). Health Qual Life Outcomes. 2: 57. Giddon DB, Mosier M, Colton T, Bulman JS. (1976). Public Health Rep. 91(6): 508–513. Gift HC, Atchison KA. (1995). Med Care. 33(11): NS57–NS77. Heydecke G, Klemetti E, Awad MA, Lund JP, Feine JS. (2003a). Int J Prosthodont. 16: 307–312. Heydecke G, Locker D, Awad MA, Lund JP, Feine JS. (2003b). Community Dent Oral Epidemiol. 31: 161–168. Hobdell M, Petersen PE, Clarkson J, Johnson N. (2003). Int Dent J. 53(5): 285–288. John MT, Koepsell TD, Hujoel P, Miglioretti DL, LeResche L, Micheelis W. (2004). Community Dent Oral Epidemiol. 32(2): 125–132. John MT, LeResche L, Koepsell TD, Hujoel P, Miglioretti DL, Micheelis W. (2003). Eur J Oral Sci. 111(6): 483–491. John M, Micheelis W. (2003). Mundgesundheitsbezogene Lebensqualita¨t in der Bevo¨lkerung: Grundlagen und Ergebnisse des Oral Health Impact Profile (OHIP) au seiner repra¨sentativen Stichprobe in Deutschland. IDZ-Information 1: 1–28. Jokovic A, Locker D, Stephens M, Kenny D, Tompson B, Guyatt G. (2002). J Dent Res. 81(7): 459–463. Klein H, Palmer CE, Knutson JW. (1938). Public Health Rep. 53: 751. Leao A, Sheiham A. (1996). Community Dent Health. 13 (1): 22–6. Locker D, Grushka M. (1987). J Dent Res. 66(9): 1414–1417. Locker D. (1988). Community Dental Health. 5: 3–18.
Oral Health-Related Quality of Life Locker D, Slade G. (1993). J Can Dent Assoc. 59(10): 830–833; 837–838; 844. Locker D, Miller Y. (1994). J Public Health Dent. 54 (3): 167–176. Locker D. (1997). In: Slade GD (ed.) Measuring Oral Health and Quality of Life. University of North Carolina, Dental Ecology, Chapell Hill, pp. 11–23. Lopez R, Baelum V. (2006). BMC Oral Health. 6: 11. McGrath C, Bedi R. (2001). Community Dent Health. 18 (3): 138–143. McNeil DW, Rainwater AJ. III. (1998). J Behav Med. 21 (4): 389–410. Murray H, Locker D, Mock D, Tenenbaum HC. (1996). J Orofac Pain. 10: 316–323. Patrick DL, Erickson P. (1993). Health Status and Health Policy. Quality of Life in Health Care Evaluation and Resource Allocation. Oxford University Press, New York. Petersen PE. (2003). The World Oral Health Report 2003. (WHO/NMH/NPH/ORH/03.2). Peterson PE, Bourgeois D, Ogawa H, Estupinan S, Ndiaye C. (2005). Bull World Health Organ. 83: 661–669. Reid JJA. (1985). Fam Pract. 2(2): 57–60. Reisine ST, Bailit HL. (1980). Soc Sci Med. 14A(6): 597–605. Reisine ST, Miller J. (1986). Soc Sci Med. 22: 1309–1314. Reisine ST. (1988). Ann Rev Public Health. 9: 1–19. Sheiham A, Watt RG. (2000). Community Dent Oral Epidemiol. 28: 399–406. Sheiham A, Spencer J. (1997). In: Pine CM (ed.) Community Oral Health. Reed Educational & Professional Publishing, Oxford, pp. 39–54. Shepherd MA, Nadanovsky P, Sheiham A. (1999). Br Dent J. 187: 38–41.
108
Slade GD. (2002). In Inglehart MR, Bagramian RA (eds.) Oral Health-Related Quality of Life. Quintessence Publishing, Illinois. Slade GD, Strauss RP, Atchison KA, Kressin NR, Locker D, Reisine ST. (1998). Community Dent Health. 15(1): 3–7. Slade GD. (1997). Community Dent Oral Epidemiol. 25(4): 284–290. Slade GD, Spencer AJ. (1994). Community Dent Health. 11: 3–11. Smith JM, Sheiham A. (1980). Community Dent Oral Epidemiol. 8(7):360–364. Strauss RP, Hunt RJ. (1993). J Am Dent Assoc. 124(1): 105–110. Szentpetery A, Szabo G, Marada G, et al. (2006). Eur J Oral Sci. 114(3): 197–203. Walter MH, Woronuk JI, Tan H, Lenz U, Koch R, Boening KW, Pinchbeck YJ. (2007). J Public Health. 15: 43–50. Walter MH, Wolf BH, Rieger C, Boening KW. (2001). J Oral Rehabil. 28: 708–716. World Health Organization. (2001). International Classification of Functioning, Disability and Health. World Health Organization, Geneva. World Health Organization. (1978). Primary Health Care. Report of the International Conference on Primary Health Care, Alma-Ata, U.S.S.R, 1978. World Health Organization, Geneva (http://www. who.int/hpr/NPH/docs/declaration_almaata.pdf). World Health Organization. (1997). WHOQOL Measuring Quality of Life. WHO, Geneva (MSA/MNH/ PSF/97.4). Yamazaki M, Inukai M, Baba K, et al. (2007). J Oral Rehabil. 34 (3): 159–168.
1853
109 Quality of Life Issues in Chronic Fatigue Syndrome P. G. McKay . C. R. Martin 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1856
2
Chronic Fatigue Syndrome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1856
3
Health-Related Quality of Life (HRQL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1858
4
HRQL in CFS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1859
5
Tools Used for Assessing HRQL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1860
6
The Quality of Life Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1860
7
The SF 36 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1860
8
Euro-QOL 5D (EQ-5D) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1861
9
Sickness Impact Profile . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1862
10 Eur-QOL 5D (EQ-5D) in CFS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1862 11 SF-36 in CFS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1863 12 The Sickness Impact Profile (SIP) in CFS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1863 13 The Quality of Life Index in CFS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1864 14 HRQL Assessment Tool for CFS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1864 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1865
#
Springer Science+Business Media LLC 2010 (USA)
1856
109
Quality of Life Issues in Chronic Fatigue Syndrome
Abstract: > Quality of life has become an area of important clinical concern within the context of the > Chronic Fatigue Syndrome. The association of deficits in quality of life in this group has been thrown into sharp relief by the more recent conceptualization of this distressing presentation as likely of being of biological, rather than psychiatric, origin. Chronic fatigue syndrome is an enigmatic presentation, the etiology of which remains speculative, this arising to the vacuum of evidence regarding causation being filled by opinion rather than fact, often to the detriment to the patient group as a whole. This chapter will explore the salient issues regarding assessment of quality of life in patients diagnosed with chronic fatigue syndrome. List of Abbreviations: CFS, chronic fatigue syndrome; CHD, coronary heart disease; EQ-5D, EuroQOL Five Dimensions; HRQL, > health-related quality of life; ME, > Myalgic Encephalomyelitis; QLI, Quality of Life Index; SF-36, short-form 36; SIP, sickness impact profile
1
Introduction
In 1958 the World Health Organization defined health as, not only being free of disease or infirmity, but being in a state of complete physical, mental and social well being (WHO 1958). Although some years have passed, Health Related Quality of Life (HRQL) is an issue, which only in recent years has been given more consideration in health care practice and is an area which is important for research (Taylor, 1995; Testa and Simonson, 1996). However in Chronic Fatigue Syndrome (CFS) there is little evidence of specific research carried out into HRQL or specific tools developed for measuring HRQL in CFS. This is a cause for concern as it may be speculated based on non-HRQL research that the symptoms of CFS impact greatly on their quality of life (Anderson and Ferrans, 1997; Rakib et al., 2005; Taylor, 2004; Tuck and Human, 1998).
2
Chronic Fatigue Syndrome
Chronic Fatigue Syndrome/Myalgic Encephalomyelitis (CFS/ME) was a term introduced in the 1950s to describe an illness characterized by muscle pain, neurological, and psychological symptoms (MacIntyre, 1998). CFS has an unknown etiology, although there is evidence to suggest that a major significant life event or an acute illness may precede the onset of this syndrome (Friedberg and Jason, 2001; Werbach, 2000; NICE, 2007; The Scottish Short Life Working Group, 2002). There is evidence to suggest that a significant amount of patients who have CFS present with an infection or virus, which was preceded by some degree of stress (MacIntyre, 1998; Mehendale et al., 2002; NICE, 2007). The major conjectured etiological factors are shown in > Figure 109-1. CFS presents with flu like symptoms, fatigue, painful lymph nodes and sore throat, to name but a few (NICE, 2007). The main psychiatric and medical differential diagnostic categories are shown in > Figure 109-2 and > Figure 109-3. respectively. The diagnosis of CFS to date is based on the Centre for Disease Control (1994) criteria as outlined in > Table 109-1. In August 2007 the > National Institute for Health and Clinical Excellence published a set of guidelines for the diagnosis and management of CFS/ME in adults and children. NICE estimate that at any one time in General Practice 40 patients out of 10,000 are likely to have CFS/ME, but the prevalence at this time is unknown (NICE, 2007). These guidelines have only
Quality of Life Issues in Chronic Fatigue Syndrome
109
. Figure 109-1 Possible contributory factors to the presentation of chronic fatigue syndrome shows the possible etiological factors in the development of the chronic fatigue syndrome. Each category is dynamic rather than definitive and there is now some consensus that an interaction between these discrete factors may precipitate the onset of chronic fatigue syndrome. Chronic fatigue syndrome remains a poorly understood clinical presentation in terms of etiological factors and the interaction among them
. Figure 109-2 Main psychiatric differential diagnoses in chronic fatigue syndrome. The main psychiatric differential diagnoses in determining a primary diagnosis of chonic fatigue syndrome. These need not be exclusionary as comorbid anxiety and depression is common in the presentation of chronic fatigue syndrome. Possible etiological factors in the development of the chronic fatigue syndrome
1857
1858
109
Quality of Life Issues in Chronic Fatigue Syndrome
. Figure 109-3 Main medical differential diagnostic groups in chronic fatigue syndrome show the main medical differential diagnostic categories in determining a primary diagnosis of chonic fatigue syndrome. The list is extensive and limited for the purposes of brevity in the figure. These diagnostic categories are exclusionary. The observation that many medical differential diagnostic categories have diverse symptom profiles has been used as a powerful argument for a physiological etiological mechanism to this perplexing presentation
been published in England and Wales and at this stage the > Scottish Intercollegiate Guidelines Network (SIGN) have not yet published guidelines on CFS/ME and this may suggest that there is a significantly higher number of patients suffering from CFS/ME in Britain at this time.
3
Health-Related Quality of Life (HRQL)
HRQL is becoming increasingly important in identifying the concerns of patients, clinicians and researchers in the management of all aspects of patient care, taking a more holistic approach towards a wide range of social and personal concepts (Gyatt et al., 1993). Measuring HRQL is very subjective when looking at the individual with a chronic illness which impact on their psychological, social and economic factors of every day life, because individuals measure their quality of life based on their beliefs, values and circumstances (Taylor, 1995). HRQL aims to identify significant information on disability and dysfunction associated with different chronic health problems, disease and chronic illness (Gyatt et al., 1993; Hopkins, 1992; Taylor, 1995). It has been identified by numerous studies carried out relating to HRQL that patients with the same illness or dysfunction can give a variety of different answers, which may be attributed to the varying severity of the same disease (Anderson and Ferrans, 1997; Gyatt et al., 1993; Vetter, 2007). Carrying out HRQL research gives policy makers and influential institutions, the ability to gain an in-depth understanding of different client groups, which may influence the redistribution of resources into areas where individual’s quality of life has been severely affected (Smee 1992).
Quality of Life Issues in Chronic Fatigue Syndrome
109
. Table 109-1 Current US case definition of chronic Fatigue Syndrome (Fukuda et al., 1994) 1. Medically unexplained chronic fatigue, experience for at least 6 months, which is of new or definite onset, that is not substantially alleviated by rest, that is not the result of ongoing exertion, and that results in substantial reduction in occupational, educational, social and personal activities. Anxiety disorders, somatoform disorders, and nonpsychotic or nonmelancholic depressions are not exclusionary The following conditions, if present, exclude diagnosis of CFS: past or current major depression with melancholic or psychotic features, delusional disorders, bipolar disorders, schizophrenia, anorexia nervosa, bulimia, or alcoholic or substance abuse within 2 years before the onset of CFS or any time afterward 2. Concurrent occurrence of four or more of the following symptoms, which must be persistent or recurrent during six or more months of the illness and do not predate the fatigue Self reported persistent or recurrent impairment in short-term memory or concentration severe enough to 0cause substantial reductions in previous levels of occupational, educational, social, or personal activities Sore throat Tender cervical lymph or axillary lymph nodes Muscle pain Multiple joint pain without joint redness or swelling Headaches of a new type, pattern or severity Unrefreshing sleep Postexertional malaise lasting more than 24 h It shows the established diagnostic criteria for chronic fatigue syndrome. It is clear from the variety of symptoms that accompany the presentation that there is likely to be a large degree of variability in terms of symptom clusters between individual patients. Chronic fatigue syndrome is an enigmatic presentation with many perspectives on the etiological aspects of the disorder. This may also explain the inclusion of such diverse symptoms in the diagnostic criteria
A number of instruments to measure HRQL have been developed to assess varying degrees of severity within the different domains related to HRQL, which will identify areas for intervention (Public Health Reports, 1994). Areas that impact on the patient’s quality of life which are normally reviewed are the Activities of Daily Living, an umbrella term for patients being able to carry out the simple tasks of every day life, such as washing, dressing and eating to more complex tasks as social, personal and financial activities (Asbring, 2001; Taylor, 1995). The Public Health Report (1994) identifies and measures domains of HRQL as physical functioning, social well being, social – role functioning and health perceptions of individuals with deteriorated health.
4
HRQL in CFS
In CFS there is much research into the cause, but to date there is no practical or precise treatment which is beneficial to the patient. Thus many CFS patients find themselves trying to cope with the illness by symptom management and their continual drive and hope to find the exact treatment and cure for a debilitating condition which impacts greatly on their quality of
1859
1860
109
Quality of Life Issues in Chronic Fatigue Syndrome
life (Rakib et al., 2005). HRQL in CFS is an important element of the syndrome, and is an area of CFS that has been neglected (Shepherd, 1998; Taylor, 2004). Diagnosis and achieving acceptance that CFS is a true illness with severe debilitating physical and physiological symptoms is difficult. Thus, many sufferers find themselves without support or help from services that in other illness are afforded to patients showing similar symptoms. HRQL in CFS has been described as the patient having to make major changes to their life. Many CFS patients describe the impact of their illness as having two lives, the life before and a life after CFS (Schoofs et al., 2004; Tuck and Human, 1998). This feeling has been identified in studies, which did not look into quality of life with CFS, but rather the “lived experience” of CFS. CFS patients, who were once able to carry out a simple task, now require a degree of assistance because of this debilitating syndrome, which greatly impacts on their HRQL. Employment and economic factors are greatly imposed on, as many individuals with CFS are only able to work part time or not at all. This impacts on their social contacts and standard of living resulting in an increasing financial dependence on third parties.
5
Tools Used for Assessing HRQL
There are no specific tools developed for the assessment of HRQL in patients with CFS instead it is generic assessment tools that have been used in the limited number of studies carried out. Generic HRQL assessment tools measure issues across a wide variety of illnesses. As HRQL is very subjective the general consensus is that when measuring HRQL in CFS more research needs to be undertaken, as limitations within research have been identified. The limitations are mainly due to recruitment of participants, where it is the most debilitated CFS suffers that take part in the studies as these individuals attend clinics. This may therefore not give a true representation of HRQL in CFS patients in general practice (Anderson and Ferrans, 1997; Hardt et al., 2001).
6
The Quality of Life Index
The Quality of Life Index (QLI) (Ferrans, 1990) comprises 34 items, measuring a patient’s satisfaction and importance of the items categorized on the questionnaires. The topics that the 34 items measures are categorized under four main headings, namely: Health and Functioning, Social and Economic, Psychological/Spiritual, and Family. The two items, namely Satisfaction and Importance are measured on a likert type scale, measuring from, Very Important to Very Unimportant and in turn Very Dissatisfied to Very Satisfied (Ferrans, 1990; Ferrans and Powers, 1985, 1992). Validity and Reliability of the QLI has been well established in previous studies (Ferrans, 1990; Ferrans and Powers, 1985, 1992). This scale has been used quite commonly in CFS and HRQL.
7
The SF 36
This is a generic assessment tool, which can be used to measure quality of life in the general population for people over the age of 14. The SF36 questionnaire (Ware et al., 1993) asks 36 questions within eight domains of health namely: Pain (2 questions), how individuals
Quality of Life Issues in Chronic Fatigue Syndrome
109
perceive their general health (6 question), vitality (4 questions), physical functioning (10 questions), social functioning (2 questions), role limitations through physical problems (4 questions), role limitations because of emotional problems (3 questions) and mental health (5 questions) (Brazier et al., 1992). This tool aims to identify positive and negative aspects of health, which are important to patients and provides very detailed information of HRQL (Garratt et al., 1993). A score of greater than 60 is considered best health, in contrast scoring less than 35 is considered poorest health (Schoofs et al., 2004). SF-36 takes around 15 min to complete, and the reliability and validity of this tool has been established through rigorous testing (Brazier et al., 1992; Garratt et al., 1993; Ware 1993). The established measurement model of the SF-36 is shown in > Figure 109-4.
8
Euro-QOL 5D (EQ-5D)
The EQ-5D is a tool which uses a questionnaire covering five domains, namely Mobility, Self Care, Usual Activities, Pain and lastly Anxiety and Depression (EuroQol Group, 1990). Each of the elements uses a visual analogue scale to measure 243 of the possible health states, where the participant can rate their own health between 0 and 100. Each rating is then given a numerical value which represents a patient has either no problems, moderate problems or severe problems within the five domains, and been validated by measuring the results from the
. Figure 109-4 Subscales and higher order domains of the SF-36 quality of life assessment tool shows the subscales of the SF-36 and their relationship to two higher order physical health and mental health domains. These higher order domains are often found to be significantly correlated in clinical groups. The figure reveals one common measurement model of the SF-36, however a number of others have been proposed
1861
1862
109
Quality of Life Issues in Chronic Fatigue Syndrome
. Figure 109-5 Subscales domains of the EuroQOL EQ-5D quality of life assessment tool shows the main scale dimensions of EQ-5D. Though each scale measures a conceptually distinct domain, each scale contributes to the overall measurement of quality of life. The EQ-5D is particularly useful in that domain distinctiveness allows comparative domain-specific deficits to be revealed relatively clearly
general public (Kind et al., 1998). The relationship of the EQ-5D dimensions to overall quality of life is shown in > Figure 109-5.
9
Sickness Impact Profile
The sickness impact profile (SIP) is a 136 item questionnaire on behavior and quality of health. It measures 12 parameters namely: Sleep and Rest, Emotional Behavior, Home Management, Body Care, Work, Communication, Mobility, Ambulation, Social Interaction, Behavior, Recreation and Eating (Bergner et al., 1981). Completing this questionnaire is time consuming, but results in a through assessment of an individuals well being.
10
Eur-QOL 5D (EQ-5D) in CFS
This is not the most popular tool to use for measuring HRQL in CFS, but its use may be beneficial for patients who have severe CFS as it can be completed in a limited amount of time compared to other HRQL assessment tools (Myers and Wilks, 1998). This tool may be useful in measuring patient’s quality of life if they were going to be surveyed daily over a period of time. This is because the questionnaire collects data relating to how the patient is feeling on the day of data collection, rather than how they have been feeling, over a period of time. In addition care must be taken when using this assessment tool, as data collected may be subjective as symptoms in CFS can vary on a daily basis. CFS patients are not always able to
Quality of Life Issues in Chronic Fatigue Syndrome
109
carry out activities or tasks that they carried out the previous day because of symptoms being unpredictable (Myers and Wilks, 1998; Tuck and Human, 1998; Taylor, 2004) identified that when using this tool patients results indicated that there were limited issues with severe mobility problems or self care, where there is alternative evidence to suggest that is an area in CFS patients that can be severely impaired (Anderson and Ferrans, 1997; Myers and Wilks, 1998). This tool is very subjective in CFS in relation to anxiety and depression, where there is evidence that patients with CFS suffer from a high degree of anxiety and depression compared to other groups of chronic illness (Rakib et al., 2005). This is a cause for concern, as many CFS patients do not see themselves as anxious and depressed. Instead they give evidence of having these symptoms due to lack of understanding about their condition and lack of positive health care treatment, thus they are angry and frustrated with their condition (Tuck and Human, 1998; Schoofs, 2004). Treatment for anxiety and depression in CFS has proved to be ineffective in some cases (Rakib et al., 2005). The EQ-5D simply asks the question are they anxious or depressed which results in the out come score of being very high, where it may be in fact how they are feeling on the day of the assessment and not in fact a true picture of how they are feeling overall (Anderson and Ferrans, 1997; Ax et al., 2001; Myers and Wilks, 1998). This tool would not appear to be the most appropriate for measuring HRQL in CFS, as it indicates a high level of anxiety, depression and a higher level of disability in CFS.
11
SF-36 in CFS
The SF-36 is the most commonly used tool in HRQL in patients with CFS. To complete the SF-36 questionnaire takes a long time, and may cause CFS suffers to become tired (Myers and Wilks, 1999). This tool allows HRQL to be assessed over a period of 4 weeks, which may give a better indication of the HRQL in a patient with CFS compared to other assessment tools, which only assess how the patient is feeling on that day. This is significant as CFS sufferer’s condition is variable on a day-to-day basis. However there is evidence to suggest that the SF-36 questionnaire is limited in use due to the range of scores in physical and emotional factors, which may lead to inaccurate data due to the SF-36 not being able to distinguish between the varying severity of the condition within the different degrees of CFS (Myers and Wilks, 1999; Buchwald et al., 1996). The SF-36 is highly sensitive to anxiety and depression, thus CFS suffers may display high levels of this, but the questionnaire does not take into account what the cause of this is (Buchwald et al., 1996; Komaroff et al., 1996).
12
The Sickness Impact Profile (SIP) in CFS
Completing the Sickness Impact Profile can take a long period of time, which may not be ideal for patients with CFS due in some cases to their concentration, being impaired thus filling out this questionnaire may lead to an exacerbation of their symptoms (Schoofs et al., 2004). However the Sickness Impact Profile gives an in depth look at the 12 main areas of daily living. Research carried out by Schweitzer et al. (1995) found that CFS severely impacts on all areas of quality of life, the results found that only people with terminal cancer or stroke had a worse quality of life than CFS sufferer. The main areas affected were social, every day tasks and work. The Sickness Impact Profile did not fixate on the mental health aspect of CFS, which may give a more informative impression of HRQL in CFS patients than compared to other HRQL
1863
1864
109
Quality of Life Issues in Chronic Fatigue Syndrome
assessment tools which give a greater impression that CFS patients have a high degree of anxiety and depression which may lead clinicians to treat these symptoms, rather than the underlying cause of CFS. (Schoofs et al., 2004). Anxiety and depression do not always determine HRQL in individuals. However when the SIP was used to survey CFS suffers in the U.K, USA and Germany the findings were found consistent, which illustrates validity and reliability of the SIP as a HRQL assessment tool (Hardt et al., 2001).
13
The Quality of Life Index in CFS
The Quality of Life Index is very accommodating to individuals with CFS as it takes into account that different dimensions of quality of life are important to different people, and how CFS impacts on them is different depending on the severity of CFS that a patient has (Anderson and Ferrans, 1997). The study carried out by Anderson and Ferrans (1997) identified that in all the domains, the Quality of Life Index showed that there was a significantly lower score in CFS patients compared to other chronic illnesses, which showed a significant reduction in HRQL. However the CFS scores in depression were significantly high which is an area, which needs to be given consideration. Anderson and Ferrans (1997) agrees that the route cause of depression has not been established regarding whether the patient was predisposed to depression before they were diagnosed with CFS or whether or not they are depressed as a result of the CFS.
14
HRQL Assessment Tool for CFS
The HRQL assessment tools are generic tools which do not focus on the specific characteristics of CFS. Although these tools do give some understanding of HRQL in CFS they are not with out their limitations. As illustrated the assessment tools do not give a clear understanding of anxiety and depression within CFS, where in some HRQL research tools the extent of anxiety and depression is exaggerated. Ax et al. (2001) highlights that CFS is not a psychiatric disease, instead sufferers find a way to adapt and cope with the impact CFS has on their lives, and is often managed by avoidance of situations, which in turn can contribute to worry (Nater et al., 2006; Van Damme et al., 2006). Therefore it may be pertinent to suggest that a disease specific model is devised for CFS, or when carrying out this kind of research participants are also assessed with psycho diagnostic tests such as the Hamilton Rating Scale for Depression. This was carried out by Vizi et al. (2007) when they looked at depression and the quality of life in living related renal transplantation. This study showed that although there were some mild depressive symptoms, especially when the recipient rejected the kidney, quality of life still had been improved, by helping improve another individual’s quality of life. Fatigue is another area which impacts on CFS suffers and although not the only symptom, is important in gaining a better understanding into the illness. It may therefore be suggested that patients be asked to complete such tools as the Pittsburg Sleep Quality Index to assess how lack of sleep is impacting on quality of life. A study by Theadom et al. (2007) looked at sleep and coping in quality of life in Fibromyalgia. They identified that Fibromyalgia and CFS have similar presenting symptoms (Geisser et al., 2007; Mehendale et al., 2002). Wagner-Raphael et al. (1999) found that quality of life in CFS was diminished with increasing fatigue.
Quality of Life Issues in Chronic Fatigue Syndrome
109
Consideration should be given to adapting or creating a HRQL assessment tool for future use as the tools used are flawed and not related to CFS. Consideration should be given to the individual when adapting an assessment tool as generic HRQL assessment tools completion time may be cumbersome to CFS suffers. Patients with CFS feel that their quality of life is affected by their variability of symptoms, cognitive dysfunction causing embarrassment in work and social situations and diminished ability to undertake social relationships and activities which result in a negative response from third parties (Anderson and Ferrans, 1997; Tuck and Human 1998). Therefore in future studies it is important that consideration is given to the uniqueness of this syndrome in order that an accurate assessment of HRQL can be undertaken to enable CFS suffers to receive the support and understanding they require to help them cope with this devastating illness.
Summary Points CFS is a unique condition of unknown etiology that impacts greatly on HRQL. HRQL is becoming of increasing interest to clinicians and researchers, but to date is an area in CFS that has been neglected.
HRQL measures aspects of an individual’s life pertaining to their activities of daily living and impact on social and economical issues.
HRQL is very subjective. Without a cure for CFS it is important that an understanding of CFS and how it impacts
on HRQL is established to help with the severe debilitating symptoms this can cause for CFS patients. To date there are no disease specific instruments for CFS and HRQL. HRQL is measured using generic instruments, which do give an understanding of HRQL in CFS, but they have limitations, which can impact on the results. The HRQL instruments that are available are time consuming and for patients with CFS this can exacerbate their symptoms, which can be affected by poor cognitive function. Consideration should be given to adapting or establishing a disease specific tool for CFS suffers to provide researchers with more pertinent data.
References Anderson JS, Ferrans CE. (1997). J Nerv Ment Dis. 185(6): 539–367. Asbring P. (2001). Journal of Advanced Nursing. 34 (3): 312–319. Ax S, Gregg VH, Jones D. (2001). Clinical Psychology Review. 21: 161–182. Bergner M, Bobbitt RA, Carter WB, Gilson BS. (1981). Med Care. 19: 757–805. Brazier JE, Harper R, Jones NMB, O’Calhain A, Thomas KJ, Usherwood T, Westlake L. (1992). Br Med J. 305: 160–164.
Buchwald D, Pearlman T, Umali J, Schmaling K, Katen W. (1996). Am J Med. 101(4): 364–370. EurQol Group. (1990). Health Policy. 16: 199–208. Ferrans CE. (1990). Oncol Nurs Forum. 17: 15–21. Ferrans CE, Powers J. (1992). Res Nurs Health. 15: 29–38. Ferrans CE, Powers MJ. (1985). Adv Nurs Sci. 18: 15–24. Friedberg F, Jason LA. (2001). J Clin Psychol. 57: 433–455. Fukuda K, Straus SE, Hicke I, Sharpe MC, Dobbins JG, Komaroff A. (1994). Ann Intern Med. 21: 953–959.
1865
1866
109
Quality of Life Issues in Chronic Fatigue Syndrome
Garratt AM, Ruta DA, Abdalla MI, Buckingham JK, Russell IT. (1993). Br Med J. 306: 1440–1444. Gyatt GH, Feeny DH, Patrick DL. (1993). Ann Intern Med. 118: 622–629. Hardt J, Buchwald D, Wilks D, Sharpe M, Nix WA, Egle VT. (2001). J Psychosom Res. 51: 431–434. Hopkins A. (1992). In: Hopkins A (ed.) Measures of the Quality of Life and the Uses to Which Such Measures May be Put. Royal Collage of Physicians, London. p. 1–12. Kind P, Dolan D, Guidex C, Williams A. (1998). Br Med J. 316: 736–741. Komaroff AL, Fagioli LR, Doolittle TH, Gandek B, Gleit MA, Guerrieror T, Korrush RJ, Ware NC, Ware JE, Bates DW. (1996). Am J Med. 101: 281–290. MacIntyre A. (1998). ME Chronic Fatigue Syndrome, A Practical Guide, Thorsons Health. England. Mehendale A, Goldman MP. (2002). Pain Practice. 2(1): 35–46. Myers and Wilkes (1998). Quality of Life Research. 8: 9–16. Nater UM, Wagner D, Solomon L, Jones JF, Unger ER, Dimitris A, Papani C, Reeves WC, Heim C. (2006). Journal of Psychosomatic research. 60(6): 567–573. National Institute for Health and Clinical Excellence. (2007). Chronic Fatigue Syndrome/ Myalgic Encephalomyelitis (or Encephalopathy): Diagnosis and Management of CFS/ME in Adult and Children. NICE, London. Rakib A, White PD, Pinching AJ, Hedge B, Newbery N, Fakhoury WK, Priebe S. (2005). Qual Life Res. 14: 11–19. Schoofs N, Bambini D, Ronning P, Bielak E, Woehi J. (2004). Orthop Nurs. 23: 364–374. The Scottish Short Life Working Group. (2002). Chronic Fatigue Syndrome/Myalgic Encephalomyelitis: Outline for Development of Services for CFS/ ME in Scotland. Report of the Scottish Short Like Working Group, Scotland, December 2002.
Schweitzer R, Kelly B, Foran A, Terry B, Whiting J. (1995). Soc Sci Med. 41: 1367–1372. Shepherd C. (1998). Living with ME the Chronic/ Postviral Fatigue Syndrome. Vermilion, London. Smee C. (1992). In: Hopkins, (ed.) Measures of the Quality of Life and the Uses to Which Measures May be Put. Royal Collage of Physicians, London. Taylor RR. (2004). Am J Occup Ther. 58: 35–43. Taylor SE. (1995). Health Psychology, 3rd ed. McGrawHill international Editions, London. Testa MA, Simonson DC. (1996). N Engl J Med. 334: 835–846. Theadom A, Cropley M, Humphery KL. (2007). J Psychosom Res. 62: 145–151. Tuck I, Human M. (1998). J Psycho Nurs Ment Health Serv. 36: 15–20. Van Damme S, Crombez G, Van Houdenhove B, Mariman AN, Michielsen W. (2006). Journal of Psychosomatic Research. 61(5): 595–599. Vetter TR. (2007). Anaesth Analg. 104: 703–718. Vizi A, Signorelli MS, Veroux M, Giammanes G, Maugen S, Nicoletti A, Veroux P. (2007). Transplant Proc l. 39: 1791–1793. Wagner-Raphael LJ, Jason LA, Ferrari JR. (1999). Journal of Occupational health Psychology. 4 (1): 63–71. Ware JE, Brook RH, Davies AR, Lohr KN. (1981). Am J Public Health. 71: 620–725. Ware JE. (1993). British Medical Journal. 306 (6980): 1429–1430. Ware JE, Snow KK, Kosinski MK, et al. (1993). SF-36 Health Survey Manual and Interpretation Guide. The Health Institute, Boston, MA, 1993. Werbach JE. (2000) Nutritional Strategies for Treating Chronic Fatigue Syndrome. Alternative Medicine Review. Vol. 5 (2) pp93–95. World Health Organization. The First Ten Years of the World Health Organization. WHO, Geneva.
110 Hemoglobin Fluctuations and Correlation with Quality of Life and Fatigue G. Caocci . R. Baccoli . G. La Nasa 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1868 2 Anemia and Hb Fluctuations in Myelodysplastic Syndromes and Multiple Myeloma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1870 3 Observation of Hematologic Patients with Hb Fluctuations . . . . . . . . . . . . . . . . . . . . . . . 1871 4 Mathematical Models for the Assessment of Hemoglobin Fluctuations . . . . . . . . . . . 1871 5 Assessment of Quality of Life and Fatigue in Hematologic Patients . . . . . . . . . . . . . . 1873 6 Mean Hemoglobin and ‘‘Variaglobin’’ Values Among MDS and MM Patients . . . 1874 7 Should Hemoglobin Fluctuations be Considered as a Parameter Affecting QOL Among Hematologic Patients? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1875 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1878
#
Springer Science+Business Media LLC 2010 (USA)
1868
110
Hemoglobin Fluctuations and Correlation with Quality of Life and Fatigue
Abstract: A reduction of > Hemoglobin (Hb) levels is common in patients with hematologic malignancies such as > Multiple Myeloma (MM) and > Myelodysplastic syndromes (MDS) and is a frequent complication of myelosuppressive chemotherapy. > Anemia is often associated with a severe impact on physical well-being, quality of life (QOL) and fatigue. Treatment of anemia requires red blood cell (RBC) transfusions, but in a high proportion of patients it is possible to achieve sustained levels of Hb with > recombinant human erythropoietin (rHuEPO and Darbepoetin alfa). Although Hb levels are routinely used to assess the severity of anemia, this numeric gradient does not clearly correlate with the clinical symptoms of anemia, which may be different among patients with the same Hb levels. Hematologic patients under poly-transfusion or treated with erythropoietin-stimulating agents may experience recurrent Hb fluctuations. Repeated RBC transfusions are associated with highly variable and unstable Hb concentrations and the fluctuating Hb may negatively influence homeostasis. Dynamic Hb fluctuations may increase fatigue and ulteriorly aggravate QOL due to recurrent ‘‘down-up-down’’ Hb excursions. These considerations prompted a group of researchers to create a mathematical model capable of transforming measurements of Hb fluctuations into a numerical parameter (> Variaglobin Index) and to correlate this parameter with quality of life and fatigue. Among transfusion-dependent MDS patients and transfusionfree MDS and MM patients (MDS patients treated with rHuEPO and MM patients treated with derbepoietin alfa), a lower amplitude of the Variaglobin Index was found to be significantly correlated with a better quality of life and less fatigue. No significant differences were observed for the amplitude of Hb fluctuations when comparing patients treated with Darbepoetin alfa or rHuEPO. This innovative mathematical model makes it easy to monitor anemia in oncohematologic patients and to adjust therapy accordingly. List of Abbreviations: EORTC, European Organization for Research and Treatment Of Cancer; FAB, French-American-British; Hb, hemoglobin; IPSS, > International Prognostic Scoring System; MDS, myelodysplastic syndromes; MM, multiple myeloma; QLQ-C30, quality of life questionnaire with 30 items; QOL, quality of life; RBC, red blood cell; rHuEPO, recombinant human erythropoietin
1
Introduction
Good quality of life (QOL) is defined as a state of physical and psychosocial well-being in which the individual is able to perform everyday activities and is satisfied with daily function (Andersson et al., 1993; Sardell and Trierweiler, 1997). QOL is a subjective, multidimensional concept incorporating physical, cognitive, emotional and social functioning. Fatigue is a subjective lack of physical and mental energy that is perceived by the individual to interfere with habitual and desired activities (Groopman and Itri, 2000; Richardson, 1995). Among the variety of factors affecting QOL and fatigue in cancer patients, low Hemoglobin (Hb) levels can play a crucial role (Bokemeyer et al., 2007; Khayat, 2000). Patients with hematologic malignancies are frequently diagnosed with anemia and it has been shown that this concomitant disorder can have a major negative impact on physical well-being, QOL and fatigue. However, the multi-factorial origin of cancer makes it difficult to assess the precise impact of anemia on QOL and fatigue in these patients. Treatment of anemia requires red blood cell (RBC) transfusions, but in a high proportion of patients it is possible to achieve sustained levels of Hb with recombinant human erythropoietin (rHuEPO) (Ludwig et al., 1993).
Hemoglobin Fluctuations and Correlation with Quality of Life and Fatigue
110
In patients suffering from hematologic malignancies, anemia can be graded as mild (Hb levels ranging from 10 to 12 g/dl), moderate (Hb levels of 8–10 g/dl), or severe (Hb levels below 8 g/dl). Although Hb levels are routinely used to assess the severity of anemia, this numeric gradient does not clearly correlate with the clinical symptoms of anemia, which may be different among patients with the same Hb levels. Therefore, it can be postulated that other hemodynamic factors may concur in determining the clinical symptoms of anemia. So far, only a few clinical nephrologists have applied a different approach to Hb levels by considering them not only from a static viewpoint (i.e., to measure in a precise time a precise concentration of Hb in 100 ml of blood) but also by describing their dynamic fluctuations over time. Hb ‘‘cycling’’ is a commonly occurring phenomenon in hemodialysis patients receiving erythropoietin-stimulating agents (Fishbane and Berns, 2005). A patient is considered to experience Hb cycling when Hb levels present an amplitude of >1.5 g/dl for more than eight consecutive weeks. In the studies performed by Walker and Pussel (2007), the within-patient variability in Hb levels was compared among hemodialysis patients receiving intravenous epoetin alfa or darbepoetin alfa. Hb fluctuations were calculated as the mean within-patient variance. The authors postulated that the adverse effects of Hb fluctuations were possibly the result of tissue ischemia and the consequential compensatory activity of the heart. Whether Hb variability should be attributed to random fluctuations, epiphenomenon or phenomenon is still a matter of debate (Berns and Fishbane, 2005; Collins et al., 2005). It can be hypothesized that hematologic patients under poly-transfusion or treated with erythropoietin-stimulating agents may also experience a sort of recurrent Hb ‘‘cycling’’ and that these recurrent ‘‘down-up-down’’ Hb excursions might increase fatigue and contemporarily lower QOL. Although the immediate effect of blood transfusion is crucial, the resulting increase in Hb levels is transient with return to baseline within a relatively short time. (> Table 110-1) Repeated RBC transfusions are associated with highly variable and unstable Hb concentrations and the fluctuating hematocrit may trigger distinct physiologic compensa¨ sterborg, 2002). tory mechanisms, such as an increase in cardiac output (O
. Table 110-1 Key points Anemia
Reduction of blood Hb levels: it can be graded as mild (Hb levels ranging from 10 to 12 g/dl), moderate (Hb levels of 8–10 g/dl), or severe (Hb levels below 8 g/dl)
Hemoglobin
The respiratory pigment in red blood corpuscles. It is composed of an iron-containing substance called ‘‘heme,’’ combined with globin
Hemoglobin fluctuation
Recurrent ‘‘down-up-down’’ hemoglobin excursions experienced by patients under poly-transfusion or treated with erythropoietin-stimulating agents
Variaglobin
Mathematical function created to assess the amplitude of Hb fluctuations
Quality of life
State of physical and psychosocial being
Fatigue
Perceived lack of physical and mental energy
Myelodysplastic syndrome
Heterogeneous hematologic disorders characterized by peripheral blood cytopenia often requiring transfusion support
Multiple myeloma
Hematologic malignancy characterized by a poor prognosis; life-threatening complications are infections, renal failure, anemia and bone disease
1869
1870
110
Hemoglobin Fluctuations and Correlation with Quality of Life and Fatigue
This switched the emphasis to a search for a numeric parameter capable of expressing the amplitude of > hemoglobin fluctuations and their correlation with QOL and fatigue. Baring in mind the previously mentioned reports of the nephrologists who used a variance model to study Hb fluctuations over time, a more complex mathematical model was set up to evaluate the amplitude of Hb fluctuations during a given time. Besides a numeric value, the method also offers a graphical representation of the phenomenon. This mathematical platform was successfully applied in a previous study of MDS patients (Caocci et al., 2007).
2
Anemia and Hb Fluctuations in Myelodysplastic Syndromes and Multiple Myeloma
Anemia is a common finding in patients with hematologic malignancies and is often a consequence of chemotherapy – a paradigmatic example is Multiple Myeloma (MM) – or chronic Myelodysplastic syndromes (MDS). MM is characterized by a poor prognosis: infections and renal failure are the major lifethreatening complications, while anemia and bone disease are the principal causes of the poor quality of life of patients (Kyle et al., 2003; Wisloff et al., 1996). Anemia is present in two thirds of patients at diagnosis of MM and reflects the course of the disease since it worsens during resistant or progressive disease, but ameliorates when the disease is controlled by treatment (Barosi et al., 2004). rHuEpo has been used in MM to increase Hb concentration, reduce RBC transfusion requirement and improve QOL. Darbepoetin alfa has also been used in MM patients and was found to have a beneficial effect on hematologic parameters and QOL (Hedenus et al., 2003). MDS are heterogeneous hematologic disorders characterized by ineffective hematopoiesis and peripheral blood cytopenia with a variable risk of progression to acute myelogenous leukemia. Most MDS patients present with symptoms of anemia and transfusion support is often needed. Chronic anemia is the most important clinical feature. In MDS patients the QOL is reduced and fatigue is considered to be the most influential symptom (Casadevall et al., 2004; Hellstro¨m-Lindberg et al., 2005; Servaes et al., 2002). Fatigue is not an isolated symptom but rather involves lethargy, decreased mental alertness, physical weakness and poor concentration. Alongside the physical symptoms associated with MDS, other symptoms such as fatigue, uncertainty, lack of understanding the disease process, fear of conversion to acute leukemia and lack of communication with physicians, may have an influence on QOL. Hb levels have been shown to have a significant influence on fatigue and QOL, particularly in MDS patients receiving RBC transfusion support or erythropoietin treatment. Several authors have reported a significant correlation between Hb values and QOL but it has been suggested that QOL may be influenced by factors other than Hb levels (Jansen et al., 2003; Oliva et al., 2005). Anemia has been found to be a negative prognostic factor for the outcome of treatment and survival of patients with various types of malignant diseases. Dependency on transfusions has been shown to affect overall survival, probably because it is associated with a more aggressive pathology and not all transfusion-dependent patients receive adequate iron-chelation therapy (Alessandrino et al., 2002; Cazzola and Malcovati, 2005). MDS patients are currently treated according to the International Prognostic Scoring System (IPSS) score (Greenberg et al., 2007). However, all patients should be appropriately informed about the risks and benefits of the proposed treatment schedule and be given the
Hemoglobin Fluctuations and Correlation with Quality of Life and Fatigue
110
possibility to participate in the clinical decision-making process. Physician-patient communication should not only include the chances of cure and the risk of treatment-related mortality, but also the likely trajectory of QOL since the latter parameter is capable of affecting any single choice (Sekeres et al., 2004). In fact, the majority of MDS patients only receive supportive therapy considering that their advanced age makes them unsuitable candidates for aggressive chemotherapy or innovative therapeutic schemes. The issue of QOL is fundamental for these patients and treatment of the anemia with transfusions or erythropoietin support is mandatory.
3
Observation of Hematologic Patients with Hb Fluctuations
A cohort of 42 hematologic patients were analyzed: 32 MDS patients (20 males, 12 females) with a mean age of 73 years (range 64–83) and 10 MM patients (six males, four females) with a mean age of 71 (range 37–80). According to the > French-American-British (FAB) classification, 20 of the MDS patients (64%) were affected by refractory anemia, and 12 (36%) by refractory anemia with less than 10% excess blasts. The majority of these patients (90%) had low or intermediate-1 risk IPSS scores (Bennett, 2000; Greenberg et al., 2007). Hemochromocytometric analysis was performed in each MM or MDS patient every 4–6 days for 1 month. The observation period started 1 month before the administration of the QOL questionnaire. Twenty of the 32 MDS patients were transfused each time their Hb levels fell below 8 g/dl. The mean number of RBC units transfused during the 1-month period was five, administered in 2–3 times. 30% of transfused patients were non-responders to previous treatment with rHuEPO. Twelve of the 32 MDS patients (37%) were treated with rHuEPO (10,000 IU s.c. three or five times a week) and were transfusion-free. Ten MM patients were treated with darbepoetin alfa (150 mg once weekly) and were transfusion-free. Two subgroups of 20 transfusion-dependent and 22 transfusion-free patients were stratified. Other two subgroups of 12 rHuEPO and ten darbepoetin alfa patients were stratified and considered for statistical analysis.
4
Mathematical Models for the Assessment of Hemoglobin Fluctuations
Among the reports in the literature, Walker and Pussel, (2007) describe a statistical model designed to record five Hb observations for each patient and calculated the logarithm of each within-patient variance on a squared scale. Moreover, a mixed-effects analysis of variance model was fitted to the log-transformed within-patient variances and weighting was based on the number of observations minus one for each patient. Treatment (epoetin alfa or darbepoetin alfa) was considered as a fixed factor and patient number was considered as a repeated factor. Another research team observed the hemodynamic Hb fluctuations in both MDS patients treated with red blood cell (RBC) transfusions or rHuEPO and MM patients treated with darbepoetin alfa and, consequently, set up a mathematical model to evaluate the amplitude of the Hb fluctuations. The obtained variable, or Variaglobin Index, was then correlated with QOL and fatigue (Caocci et al., 2007). Matlab software (The mathworks, Inc.; version 6) was used to implement the mathematical method. For each patient, 6 Hb values (expressed in g/dl) registered during a 1-month period (e.g., 6.8; 8.0; 7.4; 8.8; 7.3; 8.8) at more or less regular intervals (e.g., at days þ6; þ12; þ17; þ23; þ27) were considered.
1871
1872
110
Hemoglobin Fluctuations and Correlation with Quality of Life and Fatigue
In a subgroup of 22 transfusion-free patients, the Hb temporal function was reconstructed by cubic spline interpolation of the Hb values measured at different times. In a subgroup of 20 transfusion-dependent patients, the Hb values immediately after transfusions were not available and so the Hb temporal function was reconstructed using both measured and synthetic data. The sequential mathematical steps are described in a previous report (Caocci et al., 2007). Output management software has the ability to return: 1. A numeric value for the Variaglobin Index (higher values indicate a higher amplitude of Hb fluctuations) 2. A numeric value corresponding to the mean Hb observed during a given time interval 3. A 3D-Graphical plot of Hb fluctuations compared to a plane representing the mean Hb value in each patient (> Figures 110-1 and > 110-2) This mathematical model, despite its apparent complexity, is extremely functional and userfriendly. The hematologist only needs to fill in a simple mask with six Hb values and five time intervals; the software automatically elaborates the inserted data and returns the above-said values in a sub-mask for end plotting of the Hb fluctuation graph.
. Figure 110-1 3D-graphical plot showing the amplitude of Hb fluctuations in a transfused patient compared to a plane representing the mean Hb values. Tx transfusions; Hb hemoglobin; the black points on the graph correspond to the six Hb values registered at regular intervals during a given time. The software calculated the mean value of Hb (8.7 g/dl) and the Variaglobin Index (17.8)
Hemoglobin Fluctuations and Correlation with Quality of Life and Fatigue
110
. Figure 110-2 Calculation of the Variaglobin Index in a transfusion-free patient treated with derbopoietin. Tx transfusions; Hb hemoglobin; the black points on the graph correspond to the six Hb values registered at regular intervals during a given time. The software calculated the mean value of Hb (9.3 g/dl) and the Variaglobin Index (2.1)
5
Assessment of Quality of Life and Fatigue in Hematologic Patients
The QLQ-C30 questionnaire is a multidimensional 30-item questionnaire for the assessment of self reported QOL created by the European organization for research and treatment of cancer (EORTC) quality of life study group (Aaronson et al., 1993). This questionnaire is composed of scales evaluating physical, emotional, cognitive, role and social functions besides the global quality of life. Three symptom scales evaluate nausea and vomiting, pain and fatigue, while six single items assess dyspnea, diarrhea, constipation, appetite loss, sleep disturbances and financial difficulties. The response categories are either dichotomous such as ‘‘yes’’ or ‘‘no’’, or with four categories: ‘‘not at all’’, ‘‘a little,’’ ‘‘quite a bit,’’ ‘‘very much’’ or distributed on a modified visual analogue scale from 1 to 7. All scores are linearly transformed to a 0–100 scale. The previously mentioned cohort of patients were asked to judge the QOL during a 1-month observation period before the administration of the questionnaire. On the EORTC QLQ-C30 questionnaire, higher scores for function scales and the global quality of life indicate better functioning: results were summarized as very poor (0–20), poor (21–40), intermediate (41–60), good (61–80) and very good (81–100). Higher scores on the symptom scales reflect problems: results were summarized as none to slight (0–29), moderate
1873
1874
110
Hemoglobin Fluctuations and Correlation with Quality of Life and Fatigue
(30–69) and severe (70–100). Student’s t-test was used for comparisons of global quality of life and fatigue in the two groups of transfusion-dependent and transfusion-free patients and in the two groups of rHuEPO and darbepoetin alfa treated patients. The correlation between Hb levels, Variaglobin Index and QOL or fatigue was analyzed using Spearman’s correlation coefficient(r). The threshold for significance was p = 0.01 (2-tailed).
6
Mean Hemoglobin and ‘‘Variaglobin’’ Values Among MDS and MM Patients
All 42 patients completed and returned the QOL questionnaire (> Table 110-2). The mean QOL of the patients was rated as intermediate (46.2). Nevertheless, the transfusion-free patients perceived a significantly better QOL (score 58.7) compared to the transfused patients (poor QOL: score 32.9; p < 0.001). The mean score for fatigue was moderate (45.9) but in transfused patients it was worse (65.0) than in transfusion-free patients (score 27.8; p < 0.001). The mean Hb value observed in the total cohort of patients was 9.4 g/dl (range 7.8–12.3) and the mean Variaglobin Index was 10 (range 2.1–20.1); the mean Hb value in the transfusion-dependent patients was 8.3 g/dl with a Variaglobin Index of 13.5 and in the transfusionfree patients 10.4 g/dl with a Variaglobin Index of 6.7. Spearman’s coefficient did not reveal significant association (p < 0.01) between the mean Hb levels observed in each patient and QOL (r = 0.34; p = 0.03) but was significant when comparing Hb levels and fatigue (r = 0.47; p = 0.02). A significant association was found between Variaglobin Index and QOL, with lower levels of Variaglobin resulting in a better QOL (r = 0.68; p < 0.001) (> Figure 110-3). A significant association was also observed between Variaglobin Index and fatigue, (r = 0.78; p < 0.001): lower levels of Variaglobin were associated with less fatigue (> Figure 110-4).
. Table 110-2 Patient characteristics and clinical results Number of patients
Mean Age
MDS/ MM
Tx/EPO/ Darbe
Mean Hb
Mean Variaglobin Index
Mean QOL
20
73
MDS
Tx
8.34
13.51
32.9
65
12
73
MDS
EPO
10.15
6.09
67.4
31.9
10
65
MM
Darbe
10.8
7.52
50
23.3
EPO + Darbe
10.4
6.7
58.7
27.8
46.2
45.9
Total = 42
71
9.41
10
Mean Fa
MDS myelodysplastic patients; MM multiple myeloma patients; Tx transfusion-dependent patients; EPO rHuEPO treated and transfusion-free patients; Darbe darbepoetin alfa treated and transfusion-free patients; QOL Global quality of life score; Fa Fatigue score. Higher scores on the global quality of life scale indicate better functioning: very poor (0–20), poor (21–40), intermediate (41–60), good (61–80) and very good (81–100). Higher scores on the Fatigue scale reflect problems: none to slight (0–29), moderate (30–69) and severe (70–100). Higher scores on the Variaglobin Index reflect higher amplitude of Hb fluctuations
Hemoglobin Fluctuations and Correlation with Quality of Life and Fatigue
110
. Figure 110-3 Correlation between the Variaglobin Index and QOL in MDS and MM patients. MDS myelodysplastic patient; MM multiple myeloma patients; QOL quality of life; (r = 0.68; p < 0.001) Higher scores on the global quality of life scale indicate better functioning: very poor (020), poor (2140), intermediate (4160), good (6180) and very good (81100). Higher scores on the Variaglobin Index reflect higher amplitude of Hb fluctuations
When comparing rHuEPO with darbepoetin alfa treated patients, the mean Hb value observed in the two cohorts of patients was 10.2 (range 8.3–12.3) and 10.8 (range 8.7–12.0), respectively; the mean Variaglobin Index was 6.1 (range 4.4–9.8) and 7.52 (range 2.1–12.8), respectively; no statistically significant differences were observed between the two groups for mean Hb values or Hb fluctuations (> Figure 110-5).
7
Should Hemoglobin Fluctuations be Considered as a Parameter Affecting QOL Among Hematologic Patients?
Numerous reports have described the unique problems that affect QOL and fatigue in cancer patients. Anemia has been shown to have a significant influence on fatigue and QOL, particularly in MDS patients receiving RBC transfusion support or rHuEPO treatment (Casadevall et al., 2004; Hellstro¨m-Lindberg et al., 2005; Jansen et al., 2003; Servaes et al., 2002) or in MM patients treated with darbepoetin alfa (Kyle et al., 2003; Wisloff et al., 1996). Yet, many other hemodynamic factors may be involved in the clinical symptoms of chronic
1875
1876
110
Hemoglobin Fluctuations and Correlation with Quality of Life and Fatigue
. Figure 110-4 Correlation between the Variaglobin Index and Fatigue in MDS and MM patients. MDS myelodysplastic patients; MM multiple myeloma patients; (r = 0.78; p < 0.001) Higher scores on the Fatigue scale reflect problems: none to slight (029), moderate (3069) and severe (70100). Higher scores on the Variaglobin Index reflect higher amplitude of Hb fluctuations
anemia. Experience shows that MDS and MM patients with recurrent and ample excursions of Hb levels, particularly transfused patients, often complain of fatigue and poor QOL. The amplitude of Hb fluctuations in these patients can be measured to investigate its influence on the aforesaid parameters. In fact, the Variaglobin Index is strongly associated with QOL and fatigue. In particular, low values for the Variaglobin Index (minimal amplitude of Hb fluctuations during a 1-month time period) correspond to reduced fatigue and improvement in the global QOL. Moreover, since QOL assessment is a multidimensional tool, which also includes psychosocial issues, it can be argued that transfused patients may suffer from repeated visits to the hospital or outpatient department (Thomas, 1998). Interestingly, Hb fluctuations in patients also seem to determine other effects. In a recent study, Oliva et al. (2005) suggested that unstable Hb levels and chronic anemia in elderly MDS patients may lead to cardiac remodeling (gradual cardiac enlargement and left ventricular hypertrophy) and a consequential decline in QOL. Hb ‘‘cycling’’ or fluctuation is also a phenomenon occurring in hemodialysis patients receiving erythropoietin-stimulating agents and its adverse effects may be the result of tissue ischemia and consequential cardiac compensation (Fishbane and Berns, 2005).
Hemoglobin Fluctuations and Correlation with Quality of Life and Fatigue
110
. Figure 110-5 Hb mean values and Hb fluctuations in rHuEPO and Darbepoetin treated patients. VI Variaglobin Index; Hb hemoglobin; NS not significant; rHuEPO white bar; Darbepoetin alfa black bar
Walker and Pussell (2007) found that Hb variability was greater among patients treated with rHuEPO than in those treated with darbepoetin alfa. They postulated that this difference may be due to differences in chemical structure, receptor binding affinity or serum half-life. Darbepoetin alfa is a glycosylated form of epoetin alfa and has a threefold longer serum halflife; moreover, epoetin alfa is more potent than darbepoetin alfa in supporting the growth of erythroid burst-forming units (Jamal et al., 2006). These findings were not confirmed in the observed groups of MDS and MM patients: Hb fluctuations in the group of rHuEPO treated MDS patients were not statistically different to those found in the group of darbepoetin alfa treated MM patients. The QOL of MDS and MM patients can be improved by lowering the amplitude of Hb fluctuations. One approach could be to transfuse patients more frequently but with less RBC units. This condition can be well-satisfied when activating a home care program: patient compliance to transfusion of RBC units is likely to be better at home than in hospital. Treatment with rHuEPO is another alternative and, as indicated in international guidelines, should be recommended in patients with a high probability of positive response (Rizzo et al., 2002). In the setting of chronic illness it is important to improve and preserve the individual’s QOL and to reduce fatigue. The study of hemodynamic parameters other than Hb levels in MDS or MM patients may be able to meet this purpose. In this context, it is necessary to underline the importance of evaluating the dynamic fluctuations of Hb concentrations over time. The mathematical model described here is a simple tool for the numerical evaluation of Hb fluctuations and may represent a small but important step towards reducing the clinical symptoms of > cancer-related anemia. The method makes it easy to monitor anemia in oncohematological patients and to adjust therapy so that the stress factors associated with hemodynamic homeostasis can be reduced.
1877
1878
110
Hemoglobin Fluctuations and Correlation with Quality of Life and Fatigue
Summary Points Anemia is frequently diagnosed in patients with hematologic malignancies such as MDS or
MM, and is associated with a severe impact on physical well-being, QOL and fatigue. RBC transfusions or treatment with recombinant human erythropoietin are required. Although Hb levels are routinely used to assess the severity of anemia, this numeric gradient of concentration does not clearly correlate with the clinical symptoms of anemia, which may be different among patients with the same Hb levels. Repeated RBC transfusions are associated with highly variable and unstable Hb concentrations and the fluctuating Hb may negatively influence homeostasis. Dynamic Hb fluctuations may increase fatigue and ulteriorly aggravate QOL due to recurrent ‘‘downup-down’’ Hb excursions. An innovative mathematical model is capable of transforming the amplitude of hemoglobin fluctuations into a numerical parameter known as the Variaglobin index. Low values for the amplitude of Hb fluctuations over time correspond to reduced fatigue and improvement in global QOL. To reduce the amplitude of Hb fluctuations, a rational approach could be to transfuse patients more frequently but with less RBC units. This condition can be well-satisfied when activating a home care program. According to international guidelines, treatment with rHuEPO or darbepoetin alfa is a suitable alternative.
References Aaronson NK, Ahmedzai S, Bergman B, Bullinger M, Cull A, Duez NJ. (1993). J Natl Cancer Inst. 85: 365–376. Alessandrino EP, Amadori S, Barosi G, Cazzola M, Grossi A, Liberato LN, Locatelli F, Marchetti M, Morra E, Rebulla P, Visani G, Tura S. (2002). Haematologica. 87: 1286–1306. Andersson RT, Aaronson NK, Willen D. (1993). Qual Life 2: 369–395. Barosi G, Boccadoro M, Cavo M, Corradini P, Marchetti M, Massaia M, Merlini G, Tosi P, Tura S. (2004). Haematologica. 89: 717–741. Bennett JM. (2000). Int J Hematol. 72: 131–133. Berns JS, Fishbane S. (2005). Semin Dial. 19: 257–259. Bokemeyer C, Aapro MS, Courdi A, Foubert J, Link H, Osterborg A, Repetto L, Soubeyran P. (2007). Eur J Cancer. 43: 258–270. Caocci G, Baccoli R, Ledda A, Littera R, La Nasa G. (2007). Leuk Res. 31: 249–252. Casadevall N, Durieux P, Dubois S, Hemery F, Lepage E, Quarre` MC, Damaj G, Giraudier S, Guerci A, Laurent G, Dombret H, Chomienne C, Ribrag V, Stamatoullas A, Marie JP, Vekhoff A, Maloisel F, Navarro R, Dreyfus F, Fenaux P. (2004). Blood. 104: 321–327.
Cazzola M, Malcovati. (2005). N Engl J Med. 352: 536–538. Collins AJ, Brenner RM, Ofman JJ, Chi EM, StuccioWhite N, Krishnan M, Solid C, Ofsthun NJ, Lazarus M. (2005). Am J Kidney Dis. 46: 481–488. Fishbane S, Berns JS. (2005). Kidney Int. 68: 1337–1343. Greenberg P, Cox C, LeBeau MM, Fenaux P, Morel P, Sanz G, Sanz M, Vallespi T, Hamblin T, Oscier D, Ohyashiki K, Toyama K, Aul C, Mufti G, Bennett J. (2007). Blood. 89: 2079–2088. Groopman JE, Itri JE. (2000). J Natl Cancer Inst. 91: 1616–1634. Hedenus M, Adriansson M, San Miguel J, Kramer MH, Schipperus MR, Juvonen E, Taylor K, Belch A, Alte´s A, Martinella G, Watson D, Matcham J, Rossi G, Littlewood TJ. (2003). Br J Haematol. 122: 394–403. Hellstro¨m-Lindberg E, Gulbrandsen N, Lindberg G, Alghren T, Dahl IM, Dyberal I, Grimfors G, HesseSundin E, Hjorton M, Kanter-Lewenshon L, Linder ¨ berg G, PorwitO, Luthman M, Lo¨fvenberg E, O MacDonald A, Radlund A, Samuelsson J, Tangen JM, Winquist I, Wisloff F. (2005). Br J Haematol. 29: 1217–1219. Jamal NM, Kryzanski W, Cheung W, Lau CY, Messner HA. (2006). J Int Med Res. 34: 42–51.
Hemoglobin Fluctuations and Correlation with Quality of Life and Fatigue Jansen AJG, Essink-Bot ML, Beckers EAM, Hop WCJ, Schippeurus MR, Van Rhenen DJ. (2003). Br J Haematol. 121: 270–274. Khayat D. (2000). Semin Oncol. 27 (Suppl 4): 9–11. Kyle RA, Gertz MA, Witzig TE, Lust JA, Lacy MQ, Dispenzieri A, Fonseca R, Rajkumar SV, Offord JR, Larson DR, Plevak ME, Therneau TM, Greipp PR. (2003). Mayo Clin Proc. 78: 21–33. Ludwig H, Fritz E, Leitgeb C. (1993). Ann Oncol. 4: 161–167. Oliva EN, Dimitrov BD, Benedetto F, D’Angelo A, Nobile F. (2005). Leuk Res. 29: 1217–1219. ¨ sterborg A. (2002). In: Nowrousian (ed.) Recombinant O Human Erythropoietin (rhEPO) in Clinical Oncology – Scientific and Clinical Aspects of Anemia in Cancer – Part 2. Springer-Verlag, Wien New York, pp. 224. Richardson A. (1995). Eur J Cancer Care. 4: 20–32.
110
Rizzo JD, Lichtin AE, Woolf SH, Seidenfeld J, Bennet CL, Cella D, Djulbegovic B, Goode MJ, Jakubowski AA, Lee SJ, Miller CB, Rarick MU, Regan DH, Browman GP, Gordon MS. (2002). J Clin Oncol. 20: 4083–4107. Sardell AN, Trierweiler SJ. (1997). Cancer. 72: 3355–3365. Sekeres MA, Stone RM, Zahrieh D, Neuberg D, Morrison V, De Angelo DJ, Galinsky I, Lee SJ. (2004). Leukemia. 18: 809–816. Servaes P, Verhagen C, Bleijenberg G. (2002). Eur J Cancer. 38: 27–43. Thomas ML. (1998). Leuk Res. 22:s41–s47. Walker R, Pussel BA. (2007) Nephrology 12: 448–451. Wisloff F, Eika S, Hippe E, Hjorth M, Holmberg E, Kaasa S, Palva I, Westin J. (1996). Br J Haematol. 92: 604–613.
1879
111 Anemia and Quality of Life: Association with Diagnosis and Treatment of Anemias D. R. Thomas 1
Frequency of Anemia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1883
2
Association of Anemia with Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1883
3
Impact of Treating Anemia on Quality of Life Measures . . . . . . . . . . . . . . . . . . . . . . . . . 1886
4
Iron Deficiency Anemia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1887
5
Post-Operative Anemia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1887
6
Anemia and Chronic Kidney Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1887
7
Anemia and Cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1889
8
Chemotherapy Induced Anemia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1889
9
Anemia and Hematological Malignancies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1889
10 Anemia and Cardiovascular Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1890 11 Anemia and AIDS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1890 12 Anemia and Rheumatoid Arthritis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1890 13 Difficulties with Measuring Quality of Life and Anemia . . . . . . . . . . . . . . . . . . . . . . . . . 1890 14 Summary and Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1891 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1891
#
Springer Science+Business Media LLC 2010 (USA)
1882
111
Anemia and Quality of Life: Association with Diagnosis and Treatment of Anemias
Abstract: > Anemia appears to have a direct effect on health-related quality of life across the spectrum of many disease conditions. The presence of anemia has been linked to frailty, functional impairment, mobility impairment, and falls in older persons. The chief reasons for this rise in both incidence and prevalence of anemia with aging are gender-related changes in sex hormones and the accumulation of comorbid conditions. Anemia also appears to be an independent predictor of mortality, whether or not there is an associated underlying disease. List of Abbreviations: AIDS, acquired immunodeficiency syndrome; KDQ, kidney disease questionnaire; KS, Karnofsky scale; LASA, linear analog scale assessment; MOS-HIV, medical outcomes study in human immunodeficiency virus; QOL, quality of life; SF-36, short form-36 questionnaire; SIP, sickness impact profile Because of the association of anemia with other chronic conditions, it has been difficult to separate the effect of anemia on quality of life measures from the effect due to the underlying comorbid condition. In anemia which is responsive to treatment, including > iron deficiency, > anemia of chronic kidney disease, anemia of malignancy, chemotherapy induced anemia, AIDS related anemia, and anemia associated with rheumatoid arthritis, improvement in health-related quality of life has been demonstrated. In these conditions, treatment with iron and/or erythropoietin therapy increases hemoglobin concentration, improves quality of life, and may decrease mortality. In other common anemia’s, including > anemia of chronic disease and > unexplained anemia, where correction of the anemia may be more recalcitrant, little data on health-related quality of life is available. The measurement of health-related quality of life is fraught with methodological difficulties and comparison across trials is often impossible. But the preponderance of evidence suggests that there is a direct and independent effect of hemoglobin concentration on the symptom scores used to assess quality of life in a number of disease conditions. The presence of anemia has been linked to frailty, functional impairment, mobility impairment, and falls in older persons (Dharmarajan et al., 2006; Di Fazio et al., 2006; Kamenetz et al., 1998; Steinberg 2006). In an older community-based population, the prevalence of anemia was 24%. Anemia was threefold more prevalent in African Americans subjects and was strongly associated with poorer physical and cognitive function, and predicted a decrease in both over a 4-year period. Persons with anemia had a 1.7 times higher risk for mortality at 8 years, with no difference by sex or race (Denny et al., 2006). Women with a hemoglobin concentration between 130–140 g/L have better mobility and lower mortality compared to those with a hemoglobin concentration of less than 120 g/L (> Table 111‐1) (Chaves et al., 2002). Prolonged anemia results in left ventricular hypertrophy (Levin et al., 1996), and is strongly associated with an increased risk of subsequent myocardial infarction (Wu et al., 2001). Among older persons with congestive heart failure, the presence . Table 111‐1 Definition of anemia Gender
Hemoglobin level (grams per liter)
Men
less than 130
Women
less than 120
World Health Organization definitions of anemia
Anemia and Quality of Life: Association with Diagnosis and Treatment of Anemias
111
of anemia increases the mortality risk by 34% (Ezekowitz et al., 2003), and cardiovascular mortality has been shown to increase by 1.6% for every 1% decrease in hematocrit (McClellan et al., 2002). Vascular dementia, but not Alzheimer’s dementia, has been associated with anemia (Milward et al., 1999). Among persons with chronic obstructive pulmonary disease, measures of quality of life are significantly lower in individuals with anemia compared to those without anemia (Krishnan et al., 2006). The strong association with disease outcome suggests a high impact on health-related quality of life (Thomas, 2007).
1
Frequency of Anemia
The cause of anemia varies with the population studied. The most common cause of anemia in population surveys is iron deficiency. Iron deficiency anemia occurs in 3% of children aged 1–2 years, 2% of adolescent girls, 1% of adolescent boys and men, and 5% of women of childbearing age. In persons older than 50 years, 7% have iron deficiency anemia (Looker et al., 1997). Other studies demonstrate a frequency of iron deficiency anemia in older adults ranging from 8 to 15 percent (Joosten et al., 1992). The prevalence of all cause anemia increases with each decade of life over the age of 70, despite the fact that hemoglobin levels do not change with age in healthy persons. In the Established Population data for adults, 9% of persons aged 71–74 years were anemic. The frequency of anemia increases differentially with age, reaching 41% for men and 21% for women by age 90 years (Salive et al., 1992). A similar trend was reported in the third National Health and Nutrition Examination Survey, where the prevalence of anemia increased from 11% of males aged 70–79 years to 22% of males aged 80–89 years (Hsu et al., 2002). A very high prevalence of anemia occurs in older adults in long-term care settings, ranging from 32 to 48% (Artz et al., 2004; Chaves et al., 2002). Despite this high prevalence of anemia, particularly in older persons, anemia remains under diagnosed and under treated (Thomas, 2004). The causes of anemia in a older, community-based population include anemia of chronic disease (28%), iron deficiency anemia, (17%), vitamin B12 and/or folate deficiency (10%), anemia of chronic kidney disease (8%), and unexplained anemia (no evident cause for anemia, 37%) (Ferrucci et al., 2007). The most common causes of anemia in an institutionalized long-term care population were iron deficiency anemia (14%), anemia of chronic disease (8%), chronic renal insufficiency (6%), presumed bone marrow failure/> myelodysplasia (5%), hypothyroidism (2%), and a > hemoglobinopathy (2%). No cause for the anemia was found in 27% of subjects (> Table 111‐2, > Figure 111‐1) (Artz et al., 2004).
2
Association of Anemia with Disease
The chief reasons for this rise in both incidence and prevalence of anemia with aging are gender-related changes in sex hormones and the accumulation of comorbid conditions. Hypogonadism in older males is commonly associated with approximately a 10 g/L fall in hemoglobin concentration (Weber et al., 1991). Furthermore, men who have functional hypogonadism from pituitary adenomas exhibit anemia (Ellegala et al., 2003), and men with prostate cancer who are undergoing therapy with total androgen blockade frequently are anemic (Bogdanos et al., 2003).
1883
1884
111
Anemia and Quality of Life: Association with Diagnosis and Treatment of Anemias
. Table 111‐2 Common causes of anemia in differing populations Population Community population
Cause of anemia Anemia of chronic disease
28
Iron deficiency anemia
17
Vitamin B12 and/or folate deficiency
10
Anemia of chronic kidney disease Institutionalized population
Frequency (percent)
8
Unexplained anemia
37
Iron deficiency anemia
14
Anemia of chronic disease
8
Chronic renal insufficiency
6
Presumed bone marrow failure/ myelodysplasia
5
Hypothyroidism
2
Hemoglobinopathy
2
Unexplained anemia
27
Frequency of anemia (percentage) in population surveys of community-dwelling and institutionalized residents of long-term care facilities
. Figure 111‐1 Prevalence of anemia
Anemia and Quality of Life: Association with Diagnosis and Treatment of Anemias
111
In older persons, anemia of chronic disease accounts for 35–40% of cases (Joosten et al., 1992). The underlying disease conditions linked to anemia of chronic disease include cancer (excluding chemotherapy), congestive heart failure, hepatitis C, diabetes, rheumatoid arthritis, osteoarthritis, hypertension, stroke, asthma, inflammatory disorders, and recent surgery (Thomas and Cepeda, 2006). Multiple etiologies for anemia can occur in the same individual, requiring careful diagnostic evaluation. In particular, anemia due to chronic disease and chronic kidney disease may frequently co-exist in the same person. Anemia of chronic disease is mediated by pro-inflammatory cytokines. A polypeptide, hepcidin, is produced in the liver in response to lipopolysaccharide and interleukin-6 and is inhibited by tumor necrosis factor alpha (Nemeth et al., 2004). The expression of hepcidin results in decreased duodenal absorption of iron and blocking of iron release from macrophages, thus producing anemia in the presence of adequate iron stores. Other proinflammatory cytokines, including interleukin-1 and tumor necrosis factor alpha, have been shown to directly inhibit erythropoietin expression in vitro (Jelkmann, 1998). Much higher amounts of erythropoietin are required to restore the formation of erythroid colony-forming units in the presence of high concentrations of interferon-gamma or tumor necrosis factor alpha (Means and Krantz, 1991). The severity of the anemia of chronic disease is directly related to the severity of the underlying chronic disease and the amount of circulating cytokines. Anemia is also strongly associated with chronic kidney disease, due to decreased production of erythropoietin by the kidney. Forty-two percent of persons with chronic kidney disease are anemic, with 53% of subjects having a hematocrit below 30 (Nissenson et al., 2001). The development of anemia begins at relatively high levels of renal function (creatinine clearance of 60 ml/min in men and 40 ml/min in women) (Hsu et al., 2001). For each 10 ml/min/1.73 m2 decrease in estimated glomerular filtration rate, the hematocrit declines by 3%, and for every 1 mg/dl increase in serum creatinine, the hematocrit decreases by 1.2% (Kazmi et al., 2001). Compared to subjects with a creatinine clearance of greater than 80 mL/min, the hemoglobin hemoglobin concentration decreases by 100 g/L in women and 140 g/L in men with a creatinine clearance of 20–30 mL/minute (Ania et al., 1997). Persons with chronic kidney disease often have coexisting iron deficiency anemia. Iron deficiency anemia, anemia of chronic disease, and anemia of chronic kidney disease account for almost half of all anemia. Notably, unexplained anemia is the single largest category, accounting for a third to nearly half of all cases of anemia. The etiology of unexplained anemia is complex and remains much of a mystery. After adjusting for age, sex, and hemoglobin concentration, erythropoietin levels are lower in subjects with unexplained anemia compared to those with iron deficiency anemia, suggesting that erythropoietin may be an etiological factor in unexplained anemia (Artz et al., 2004). Although iron deficiency anemia is associated with a high compensatory erythropoietin level, the anemia of chronic disease shows a bimodal erythropoietin response (low in some persons and high in others) (Ferrucci et al., 2007). Unexplained anemia, along with vitamin B12 and/or folate deficiency, tends to be mild in severity and shows little compensatory erythropoietin response. Persons with anemia of chronic disease have higher levels of interleukin-6 and C-reactive protein (but not tumor necrosis factor alpha) compared to non-anemic controls. Persons with unexplained anemia have significantly lower C-reactive protein than nonanemic controls. In addition, persons with unexplained anemia have lower levels of interleukin-6, tumor necrosis factor alpha and C-reactive protein compared to other types of anemia, suggesting that unexplained anemia is not related to inflammation (Ferrucci et al., 2007). Interleukin-6 levels are markedly elevated in persons with anemia of chronic disease, but there is no difference in
1885
1886
111
Anemia and Quality of Life: Association with Diagnosis and Treatment of Anemias
interleukin-6 levels between unexplained anemia and iron deficiency anemia (Artz et al., 2004). These studies suggest that the chief association with unexplained anemia is a low erythropoietin level. However, the low levels of pro-inflammatory markers and low lymphocyte counts suggest that this reduced erythropoietin response in unexplained anemia is not attributable to inflammation. The nutritional anemia’s, including deficiency in iron, vitamin B12, or folate, and anemia due to blood loss and drug side effects also occur in persons with chronic disease. Other causes of anemia include acute blood loss (7%), myelodysplasia syndrome (about 5%) and vitamin B12 deficiency (3–5%). Large numbers of older persons with anemia remain undiagnosed despite evaluation (Joosten et al., 1992). This epidemiological data indicates that anemia occurs most frequently as a comorbid condition associated with other diseases often those associated with increased proinflammatory cytokines. Anemia appears to be an independent predictor of morbidity and mortality, even when adjusted for conditions known to produce lower hemoglobin concentrations, such as malignancy, peptic ulcer disease, and infections (Izaks et al., 1999).
3
Impact of Treating Anemia on Quality of Life Measures
Because of the association of anemia with other chronic conditions, it has been difficult to separate the effect of anemia on quality of life measures from the effect due to the underlying comorbid condition (> Table 111-3). For example, the Short Form-36 (SF-36) quality of life
. Table 111-3 Association of anemia with quality of life outcome Type of anemia
Quality of life outcome
Iron deficiency anemia
Improved
Post-operative anemia
No difference
Anemia of chronic kidney disease
Improved
Cancer-related anemia
Improved
Chemotherapy induced anemia
Improved
Anemia associated with cardiovascular disease
Improved
Anemia associated with acquired immunodeficiency syndrome
Improved
Anemia associated with rheumatoid arthritis
Improved
The association of anemia by classification with quality of life outcomes in published trials
score for persons with chronic kidney disease are much lower than population norms whether or not anemia is present (Perlman et al., 2005). Thus, the quality of life measures may reflect more on the impact of the chronic condition rather than the anemia per se. This confounding effect on quality of life can only be controlled for when the anemia can be corrected by treatment. Nutritional anemia’s, anemia of chronic kidney disease, and cancer-associated anemia are most amenable to treatment. Other anemia, including anemia of chronic disease and unexplained anemia, currently can be improved only by addressing the underlying condition.
Anemia and Quality of Life: Association with Diagnosis and Treatment of Anemias
111
A major advance in the treatment of anemia occurred with the cloning of the human gene for erythropoietin in 1983, and the demonstration of the efficacy of recombinant human erythropoietin treatment in dialysis patients in 1986. Previously, patients with erythropoietinassociated anemia required blood transfusions when they became symptomatic, despite concerns about transfusion safety. Currently, erythropoietin-associated anemia can be titrated to near-normal hemoglobin concentrations. The ability to correct anemia in previously recalcitrant conditions has allowed assessment of quality of life measures independent of the presence of the chronic condition. The results for the impact of the correction of anemia on health-related quality of life by cause of the anemia is summarized below.
4
Iron Deficiency Anemia
Treatment of iron deficiency anemia has been demonstrated to improve quality of life measures. In 44 women age 18–50 years, the Piper Fatigue Scale scores were higher and the SF-36 physical health, mental health and vitality scores were lower in iron deficient women compared to non-iron deficient women. Improvement is these scores followed correction of the anemia (Patterson et al., 2001). In 92 premenopausal women with iron deficiency anemia, the baseline SF-36 vitality and general health scores were significantly lower than national norms. After iron replacement therapy, all domain scores were comparable to or greater than the national norms. The greatest improvement in physical function, vitality, and general health perception occurred in women with a hemoglobin level less than 90 g/L at baseline (Ando et al., 2006). Treatment with iron in non-anemic women 18–55 years of age whose primary condition was unexplained fatigue resulted in an improved perceived level of fatigue compared to nonanemic, placebo treated women (Verdon et al., 2003).
5
Post-Operative Anemia
In thirty patients undergoing unilateral hip arthroplasty, hemoglobin and a quality of life questionnaire was measured pre-operatively and at 7, 28 and 56 days post-operatively. The lowest pre-operative hemoglobin was 110 g/L. Twelve patients required blood transfusions for a post-operative hemoglobin less than 90 g/L. Approximately two-thirds of the post-operative hemoglobin deficit was corrected by day 28. Quality of life scores did not show any relationship with hemoglobin in the post-operative period (Wallis et al., 2005). The findings in this study may be limited by the finding that the mean hemoglobin concentration declined to only 106 g/L on post-operative day 7 and the fact that some patients may have remained limited in physical activity due to the surgery.
6
Anemia and Chronic Kidney Disease
Compared to the United States adult population, persons with chronic kidney disease have lower Short Form-36 (SF-36) quality of life scores, and persons on chronic hemodialysis have
1887
1888
111
Anemia and Quality of Life: Association with Diagnosis and Treatment of Anemias
even lower scores. The individual components of the SF-36 quality of life score correlates directly with the hemoglobin concentration in all measures except pain (Perlman et al., 2005). In predialysis persons with chronic kidney disease, treatment with recombinant human erythropoietin for 48 weeks resulted in a significant increase in energy and physical function compared to a significant decrease in physical function in the control group. The Sickness Impact Profile and four scales from the Medical Outcomes Study Short Form were used in this study for assessment of quality of life (Roth et al., 1994). Treatment with recombinant human erythropoietin for 12 weeks in predialysis subjects resulted in a significant improvement in quality of life score measured by a questionnaire in persons whose anemia was corrected (Kleinman et al., 1989). Another trial reported a significant improvement in quality of life and work capacity in subjects receiving recombinant human erythropoietin (Ganguli et al., 1997). A very small trial of 11 subjects treated with recombinant human erythropoietin perceived an increased sense of well-being, felt more energetic and were more able to perform their work (Lim et al., 1989). Increases in a sense of well-being, energy level and work capacity has been demonstrated in another trial among predialysis subjects (Teehan, 1991), and shown an increase in measured exercise capacity using a bicycle ergonometer in a similar study (Clyne and Jogestrand, 1992). Quality of life, using the Linear Analog Scale Assessment (LASA) and the Kidney Disease Questionnaire (KDQ) subscales, has been evaluated in persons with chronic kidney disease who were not on hemodialysis and who were treated with erythropoietin. An increase in hemoglobin concentration was a statistically significant predictor of improvement in quality of life scores. Based on a two unit change in hemoglobin concentration, the greatest incremental improvement in quality of life score per unit of hemoglobin concentration increase occurred at a hemoglobin concentration of 110–120 g/L (Lefebvre et al., 2006). In a prospective study of stable hemodialysis patients (excluding diabetic patients and those with severe comorbidity) treated with recombinant human erythropoietin, quality of life was assessed using the Karnofsky scale (KS) and the Sickness Impact Profile (SIP) questionnaire. The mean KS scores increased significantly and the mean global score of SIP decreased from baseline to 6 months, while no significant changes were observed in the control group. The effect was independent of age and positively related to the degree of improvement in the hematocrit level (Moreno et al., 1996). All studies in patients with chronic kidney disease which reported an assessment of quality of life have shown a statistically significant difference favoring treatment with recombinant human erythropoietin (Cody et al., 2007). The treatment of anemia in persons on peritoneal or hematolysis, the most common indication for recombinant human erythropoietin, clearly increases the concentration of hemoglobin and decreases the need for blood transfusions, and is arguably the standard of care (Tangalos et al., 2004). However, the major benefit in patients with chronic kidney disease occurs with a partial correction of hemoglobin concentration to 110–120 g/L. Increases beyond this level do not appear to offer additional benefits, and is associated with increased cardiovascular mortality (Drueke et al., 2006; Singh et al., 2006). Treatment of anemia in patients with pre-dialysis Stage 3 or 4 chronic kidney disease also increases the hemoglobin concentration and reduces the need for blood transfusions, but has not shown clear evidence for a beneficial or adverse effect on the progression of chronic kidney disease, or on the timing of initiation of dialysis, or on subsequent mortality (Cody et al., 2007).
Anemia and Quality of Life: Association with Diagnosis and Treatment of Anemias
7
111
Anemia and Cancer
In anemic patients with cancer who were treated with recombinant human erythropoietin, a meta-analysis of 23 published trials demonstrated a clear and statistically significant improvement in quality of life scores measured by various instruments compared to control groups who experienced no change or a decline in scores (Anonymous, 2005). Moreover, there is a linear association of improvement in quality of life with duration of anemia treatment. Although the recombinant human erythropoietin subjects improved the subject’s reported quality of life score, their performance status continued to show a gradual decrease over time. Thus, the subject’s perception of quality of life occurred in a setting of inevitable decline in performance status due to the underlying cancer.
8
Chemotherapy Induced Anemia
In patients with breast cancer receiving chemotherapy who were treated with recombinant human erythropoietin, scores on the Functional Assessment of Cancer Therapy for anemia and fatigue were higher than subject who did not receive recombinant human erythropoietin therapy (Chang et al., 2005). In five trials among cancer patients on and off chemotherapy, an increase in hemoglobin concentration was associated with improvements in fatigue, which, in turn, was associated with improved physical, functional, emotional and overall well-being (Cella et al., 2004). Clinical guidelines for cancer patients found Level I evidence showing that quality of life is significantly improved with erythropoietin therapy in patients with both chemotherapyinduced anemia and anemia of chronic disease who had an increase in hemoglobin concentration (Bokemeyer et al., 2004). Other studies confirm that treatment of anemia increases hemoglobin concentration, improves quality of life, and may decrease mortality (Cella, 1998; Quirt et al., 2001; Valderrabano, 2000). In a meta-analysis of the patients with anemia, a consistent and significant positive correlation has been shown between improvement in hematocrit levels and improvements in adjusted health-related quality of life scores. Patients treated with recombinant human erythropoietin showed statistically significant improvements in quality of life on the Functional Assessment of Cancer Therapy - Fatigue subscale (four studies) and the Cancer Linear Analogue Scale fatigue self assessment score (four studies) compared to standard care or placebo (Ross et al., 2006). Treatment on anemia due to cancer or cancer chemotherapy with recombinant human erythropoietin improves the hemoglobin concentration and reduces the risk of blood transfusion by 62%. Evidence-based Clinical Practice Guidelines of the American Society of Clinical Oncology and the American Society of Hematology recommend either recombinant human erythropoietin or red blood cell transfusion for patients with chemotherapy-associated anemia whose hemoglobin concentration has declined to a level 100 g/L, depending on the severity of anemia or clinical circumstances (Rizzo et al., 2002). There is concern over potential stimulating effects of recombinant human erythropoietin on certain cancers.
9
Anemia and Hematological Malignancies
Anemia is a common finding in patients with hematologic malignancies such as multiple myeloma and myelodysplastic syndrome, and has been associated with decreased quality of life
1889
1890
111
Anemia and Quality of Life: Association with Diagnosis and Treatment of Anemias
(Casadevall et al., 2004; Hellstrom-Lindberg et al., 2003). Treatment of the anemia with red cell transfusion or recombinant human erythropoietin was associated with an improvement in quality of life scores, particularly in the recombinant human erythropoietin group. Rapid fluctuations in the correction of hemoglobin was associated with a lower quality of life (Caocci et al., 2008).
10
Anemia and Cardiovascular Disease
Treatment of anemia in patients with congestive heart failure and an ejection fraction of less than 40% has demonstrated a 42% improvement in New York Heart Association classification, compared with the control patients who had a decrease of 11% (Silverberg et al., 2001). Recombinant human erythropoietin also significantly enhances exercise capacity in patients with congestive heart failure. In a 3 month trial, significant increases in hemoglobin, peak oxygen consumption, and 6 min exercise duration were observed in the erythropoietin treated group but no significant changes occurred in the control group (Mancini et al., 2003). Correction of anemia also produces a decrease in the left ventricular hypertrophy (Levin et al., 1996). As may be expected with increased blood volume, erythropoietin therapy may increase blood pressure, necessitating close monitoring in patients with known cardiovascular disease.
11
Anemia and AIDS
In 650 subjects with human immunodeficiency virus treated with recombinant human erythropoietin, improvement in the linear Analog Scale Assessment Quality of Life score, energy score, activity score, and the Medical Outcomes Study (MOS)-HIV physical and mental health summary scores was observed from baseline to final measurement. The improvement coincided with the increase in hemoglobin concentration (Saag et al., 2004).
12
Anemia and Rheumatoid Arthritis
In patients with rheumatoid arthritis, ten published articles reported an improvement in disease markers including swollen, painful, and tender joints, pain, muscle strength, and energy levels with treatment and resolution of anemia. Two trials found a significant improvement in quality of life scores in patients who responded to treatment for anemia (Wilson et al., 2004).
13
Difficulties with Measuring Quality of Life and Anemia
Quality of life remains difficult to measure due to a lack of standardized instruments. The use of non-scandalized instruments to measure quality of life makes interpretation of the data difficult to compare across studies. Validated scales were not used in five of eight cancer treatment trials in which quality of life or exercise capacity were assessed. In those trials where standardized scales were used, reporting was incomplete. Reporting of quality of life
Anemia and Quality of Life: Association with Diagnosis and Treatment of Anemias
111
improvement is often based upon individual components of a scale rather than an complete analysis of all scale variables. Often only the positive subscales were reported and little detail given for the overall scores. Similar problems exist in other disease conditions. In studies of quality of life in chronic kidney disease, only one trial reported means and standard deviations for quality of life measures (> Table 111‐4) (Kleinman et al., 1989). . Table 111‐4 Difficulties with reporting quality of life measures Lack of standardized instruments Lack of validated scales Reporting individual subscales rather than global scores Failure to report means and standard deviations Interpreting reported quality of life measures in published trials is limited by choice of instruments and failure to report adequate statistical measures
For these reasons, it is frequently impossible to perform a pooled analysis of subjects given the wide variability of assessment and the uncertainties regarding the validity of the instruments. It does appear that individual items of quality of life assessment improve with improvement in hemoglobin concentration, suggesting an overall positive effect of correction of anemia with improvement in quality of life. It is very likely that correction of anemia with human recombinant erythropoietin does improve quality of life, but it is unclear whether this is a global improvement or only an improvement of some aspects of the quality of life measure used in these studies.
14
Summary and Conclusions
Anemia appears to have a direct effect on health-related quality of life across the spectrum of many disease conditions. Anemia appears to be an independent predictor of mortality, whether or not there is an associated underlying disease. In anemia which is responsive to treatment, including iron deficiency, anemia of chronic kidney disease, anemia of malignancy, chemotherapy induced anemia, AIDS related anemia, and anemia associated with rheumatoid arthritis, improvement in health-related quality of life has been demonstrated. In these conditions, treatment with iron and/or erythropoietin therapy increases hemoglobin concentration, improves quality of life, and may decrease mortality. In other common anemia’s, including anemia of chronic disease and unexplained anemia, where correction of the anemia may be more recalcitrant, little data on health-related quality of life is available. The measurement of health-related quality of life is fraught with methodological difficulties and comparison across trials is often impossible. But the preponderance of evidence suggests that there is a direct and independent effect of hemoglobin concentration on the symptom scores used to assess quality of life in a number of disease conditions.
Summary Points Anemia appears to have a direct effect on health-related quality of life across the spectrum of many disease conditions.
1891
1892
111
Anemia and Quality of Life: Association with Diagnosis and Treatment of Anemias
Anemia appears to be an independent predictor of mortality, whether or not there is an associated underlying disease.
In anemia which is responsive to treatment, improvements in health-related quality of life has been demonstrated.
Methodological difficulties confound the assessment of health-related quality of life. The preponderance of evidence suggests that there is a direct and independent effect of hemoglobin concentration on the symptom scores used to assess quality of life in a number of disease conditions.
References Ando K, Morita S, Higashi T, Fukuhara S, Watanabe S, Park J, Kikuchi M, Kawano K, Wasada I, Hotta T. (2006). Qual Life Res. 15: 1559–1563. Ania BJ, Suman VJ, Fairbanks VF, Rademacher DM, Melton LJ III. (1997). JAm Geriatr Soc. 45: 825–831. Anonymous. (2005). Curr Med Res Opin. 21 Suppl 2: S16–S18. Artz AS, Fergusson D, Drinka PJ, Gerald M, Bidenbender R, Lechich A, Silverstone F, McCamish MA, Dai J, Keller E, William B, Ershler WB. (2004). J Am Geriatr Soc 52: 423–427. Bogdanos J, Karamanolakis D, Milathianakis C, Repousis P, Tsintavis A, Koutsilieris M. (2003). Anticancer Res. 23: 1757–1762. Bokemeyer C, Aapro MS, Courdi A, Foubert J, Link H, Osterborg A, Repetto L, Soubeyran P. (2004). Eur J Cancer. 40: 2201–2216. Casadevall N, Durieux P, Dubois S, Hemery F, Lepage E, Quarre` MC, Damaj G, Giraudier S, Guerci A, Laurent G, Dombret H, Chomienne C, Ribrag V, Stamatoullas A, Marie JP, Vekhoff A, Maloisel F, Navarro R, Dreyfus F, Fenaux P. (2004). Blood. 104: 321–327. Caocci G, Baccoli R, La Nasa G. (2008) in press. Cella D. (1998). Semin Oncol 25(3 Suppl 7): 43–46. Cella D, Kallich J, McDermott A, Xu X. (2004). Ann Oncol. 15: 979–86. Chang J, Couture F, Young S, McWatters KL, Lau CY. (2005). J Clin Oncol. 23: 2597–2605. Chaves PH, Ashar B, Guralnik JM, Fried LP. (2002). J Am Geriatr Soc. 50: 1257–1264. Clyne N, Jogestrand T. (1992). Nephron. 60: 390–396. Cody J, Daly C, Campbell M, Donaldson C, Khan I, Rabindranath K, Vale L, Wallace S, MacLeod A. (2007). The Cochrane Database of Systematic Reviews 4, [no page #]. Denny SD, Kuchibhatla MN, Cohen HJ. (2006). Am J Med 119: 327–334.
Dharmarajan TS, Avula S, Norkus EP. (2006). J Am Med Dir Assoc. 7: 287–293. Di Fazio I, Franzoni S, Fisoni GB, Gatti S, Cornali C, Stofler PM. (2006). J Am Med Dir Assoc. 7: 208–211. Drueke TB, Locatelli F, Clyne N, Eckardt KU, Macdougall IC, Tsakiris D, Burger HU, Scherhag A. (2006). N Engl J Med. 355: 2071–2084. Ellegala DB, Alden TD, Couture DE, Vance ML, Maartens NF, Laws ER Jr. (2003). J Neurosurg. 98: 974–977. Ezekowitz JA, McAlister FA, Armstrong PW. (2003). Circulation. 107: 223–225. Ferrucci L, Guralnik JM, Bandinelli S, Semba RD, Lauretani F, Corsi A, Ruggiero C, Ershler WB, Longo DL. (2007). J Compilation. 136: 849–855. Ganguli A, Singh NP, Ahuja N. (2003). Nephrol Dialy Transpl. 18 Suppl 4: 692. Hellstrom-Lindberg E, Gulbrandsen N, Lindberg G, Ahlgren T, Dahl IM, Dybedal I, Grimfors G, Hesse-Sundin E, Hjorth M, Kanter-Lewensohn L, Linder O, Luthman M, Lofvenberg E, Oberg G, Porwit-MacDonald A, Radlund A, Samuelsson J, Tangen JM, Winquist I, Wisloff F (2003). Br J Haematol. 120: 1037–1046. Hsu CY, McCulloch CE, Curhan GC. (2002). J Am Soc Nephrol. 13(11): 2783–2786. Hsu CY, Bates DW, Kuperman GJ, Curhan GC. (2001). Kidney Int. 59: 725–731. Izaks GJ, Westendorp RG, Knook DL. (1999). JAMA. 281: 1714–1717. Jelkmann W. (1998). J Interferon Cytokine Res. 18: 555–559. Joosten E, Pelemans W, Hiele M, Noyen J, Verhaeghe R, Boogaerts MA. (1992). Gerontology 38: 111–117. Kamenetz Y, Beloosesky Y, Zelter C, Gotlieb D, Magazanik A, Fishman P, Grinblat J. (1998). Aging Clin Exp Res. 10: 32–38.
Anemia and Quality of Life: Association with Diagnosis and Treatment of Anemias Kazmi WH, Kausz AT, Khan S, Abichandani R, Ruthazer R, Obrador GT, Pereira BJ. (2001). Am J Kidney Disord. 38: 803–812. Kleinman KS, Schweitzer SU, Perdue ST, Bleifer KH, Abels RI. (1989). Am J Kidney Disord. 14: 486–495. Krishnan G, Grant BJ, Muti PC, Mishra A, Ochs-Balcom HM, Freudenheim JL, Trevisan M, Schunemann HJ. (2006). BMC Pulmonary Med. 6: 23. Lefebvre P, Vekeman F, Sarokhan B, Enny C, Provenzano R, Cremieux PY. (2006). Curr Med Res Opin. 22: 1929–1937. Levin A, Singer J, Thompson CR, Ross H, Lewis M (1996). Am J Kidney Disord. 27: 347–354. Lim VS, DeGowin RL, Zavala D, Kirchner PT, Abels R, Perry P, Fangman J. (1989). Ann Intern Med. 110: 108–114. Looker AC, Dallman PR, Carroll MD, Gunter EW, Johnson CL. (1997). JAMA. 277: 973–976. Mancini DM, Katz SD, Lang CC, LaManca J, Hudaihed A, Androne AS. (2003). Circulation. 107: 294–299. McClellan WM, Flanders WD, Langston RD, Jurkovitz C, Presley R. (2002). J Am Soc Nephrol. 13: 1928–1936. Means RT Jr, Krantz SB. (1991). Blood. 78: 2564–2567. Milward EA, Grayson DA, Creasey H, Janu MR, Brooks WS, Broe GA. (1999). Neuroreport 10: 2377–2381. Moreno F, Aracil FJ, Perez R, Valderrabano F. (1996). Am J Kidney Disord. 27: 548–556. Nemeth E, Rivera S, Gabayan V, Keller C, Taudorf S, Pedersen BK, Ganz T. (2004). J Clin Invest 113: 1271–1276. Nissenson AR, Collins AJ, Hurley J, Petersen H, Pereira BJ, Steinberg EP. (2001). J Am Soc Nephrol 12: 1713–1720. Patterson AJ, Brown WJ, Roberts DC. (2001). J Am Coll Nutr. 20: 337–342. Perlman RL, Finkelstein FO, Liu L, Roys E, Kiser M, Eisele G, Burrows-Hudson S, Messana JM, Levin N, Rajagopalan S, Port FK, Wolfe RA, Saran R. (2005). Am J Kidney Disord. 45: 658–666. Quirt I, Robeson C, Lau CY, Kovacs M, Burdette-Radoux S, Dolan S, Tang SC, McKenzie M, Couture F. (2001). J Clin Oncol. 19: 4126–4134. Rizzo JD, Lichtin AE, Woolf SH, Seidenfeld J, Bennett CL, Cella D, Djulbegovic B, Goode MJ, Jakubowski
111
AA, Lee SJ, Miller CB, Rarick MU, Regan DH, Browman GP, Gordon MS. (2002). J Clin Oncol. 20: 4083–4107. Ross SD, Allen IE, Henry DH, Seaman C, Sercus B, Goodnough LT. (2006). Clin Ther. 28: 801–831. Roth D, Smith RD, Schulman G, Steinman TI, Hatch FE, Rudnick MR, Rudnick MR, Sloand JA, Freedman BI, Williams WW Jr, Shadur CA. (1994). Am J Kidney Disord. 24: 777–784. Saag MS, Bowers P, Leitz GJ, Levine AM. (2004). AIDS Res Hum Retroviruses. 20: 1037–1045. Salive ME, Cornoni-Huntley J, Guralnik JM, Phillips CL, Wallace RB, Ostfeld AM, Cohen HJ. (1992). J Am Geriatr Soc. 40: 489–496. Silverberg OS, Wexler D, Sheps D, Blum M, Keren G, Baruch R, Schwartz D, Yachnin T, Steinbruch S, Shapira I, Laniado S, Iaina A. (2001). J Am Coll Cardiol. 37: 1775–1780. Singh AK, Szczech L, Tang KL, Barnhart H, Sapp S, Wolfson M, Reddan D. (2006). N Engl J Med 355: 2085–2098. Steinberg KE. (2006). J Am Med Dir Assoc 7: 327. Tangalos EG, Hoggard JG, Murray AM, Thomas DR. (2004). J Am Med Dir Assoc 5(4 Suppl): H1–H6. Teehan BP. (1991). Am J Kidney Disord. 18: 50–59. Thomas DR, Cepeda OA. (2006). Aging Health. 2: 303–312. Thomas DR. (2007). J Am Med Dir Assoc 8: 80–82. Thomas DR. (2004). J Gerontol Ser A-Biolog Med Sci. 59: 238–241. Valderrabano F. (2000). Nephrol Dialysis Transpl. 15 (Suppl): 23–28. Verdon F, Burnand B, Stubi CL, Bonard C, Graff M, Michaud A, Bischoff T, de Vevey M, Studer JP, Herzig L, Chapuis C, Tissot J, Pecoud A, Favrat B. (2003). BMJ. 326: 1124. Wallis JP, Wells AW, Whitehead S, Brewster N. (2005). Transfusion Med. 15: 413–418. Weber JP, Walsh PC, Peters CA, Spivak JL. (1991). Am J Hematol 36: 190–194. Wilson A, Yu HT, Goodnough LT, Nissenson AR. (2004). Am J Med. 116 Suppl 7A: 50S–57S. Wu WC, Rathore SS, Wang Y, Radford MJ, Krumholz HM. (2001). N Engl J Med. 345: 1230–1236.
1893
112 Quality of Life in Hemophilia S. V. Mackensen . A Gringeri 1 1.1 1.2
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1896 Hemophilia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1896 Quality of Life and Hemophilia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1898
2 2.1 2.1.1 2.1.2 2.1.3 2.1.4 2.1.5 2.2 2.2.1 2.2.2 2.2.3 2.3
Burden of Hemophilia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1899 Clinical Burden . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1899 Bleeding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1899 Treatment Complications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1899 Venous Access . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1899 Pain and Arthropathy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1900 Presence of Inhibitors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1900 Psychological Burden . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1901 Prenatal Diagnosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1901 Stress and Coping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1902 Anxiety and Depression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1903 Economical Burden . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1904
3 3.1 3.1.1 3.1.2 3.1.3 3.2 3.2.1 3.2.2 3.2.3 3.2.4 3.3
Quality of Life Instruments in Hemophilia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1905 Pediatric Instruments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1906 Hemo-QOL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1906 CHO-KLAT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1909 Quality of Life Questionnaire for Young Patients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1909 Adult Instruments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1910 Hem-A-QOL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1910 Medtap Questionnaire (Hemo-QOL-A) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1911 Hemofilia-QOL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1912 Hemolatin-QOL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1913 Treatment Satisfaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1913
4 4.1 4.2 4.3 4.4 4.5
Findings from Quality of Life Assessments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1914 Comparison with the General Population . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1914 Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1914 Severity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1915 Complications & Side Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1915 General Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1915
5
Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1916 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1916
#
Springer Science+Business Media LLC 2010 (USA)
1896
112
Quality of Life in Hemophilia
Abstract: Haemophilia is a congenital bleeding disorder affecting about 1 out of 10,000 people in the general population. Haemophilia is characterized by post-traumatic and spontaneous bleeds, mainly in joints and muscles. Treatment is based on replacement of lacking factors by intravenous infusions when bleeding occurs or preventively. HR-QOL assessment is increasingly recognised as an important health outcome measure in haemophilia, which can help to optimise the treatment and to understand which treatment strategy fits the patients’ needs best. HR-QOL instruments can be distinguished between generic vs. disease-specific measures and instruments for adults and for children. Although the modern management of haemophilia has improved over the last decades patients still can suffer from a burden of their disease regarding clinical (such as treatment complications, presence of inhibitors, pain and arthropathy), psychological (stress and coping, anxiety and depression, stigmatisation and discrimination) and economical aspects. For an adequate assessment of HR-QoL in haemophilic patients validated instruments are necessary. Only in the last 6 years disease-specific HR-QoL instruments have been developed for haemophilic children, namely the European Haemo-QoL, the Canadian CHO-KLAT, the American Quality of Life Questionnaire for Young Patients and for adult patients: the Medtap questionnaire (USA, Canada, Spain and Germany), the South American Hemolatin-QoL, the Italian Haem-A-QoL and the Spanish Hemofilia-QoL. Since treatment of haemophilia can impact on patients’ HR-QoL the assessment of treatment satisfaction becomes quite important providing information for clinical trials and disease management programs. The first haemophilia-specific treatment satisfaction questionnaire for haemophilia patients is described (Hemo-SatA). Development and psychometric testing of these questionnaires is described in detail and revealed that all instruments are reliable and valid instruments which can be used in clinical trials. The choice of an instrument should be based on study-related and instrument-related issues. Results of HR-QoL studies are described concerning treatment effects, severity of the disease, complications and side effects and comparison with the general population. List of Abbreviations: CHQ, Child Health Questionnaire; GHQ, General Question of the Child Health Questionnaire; Hem-A-QOL, Hemophilia-specific quality of life instrument for adults; Hemo-QOL, Hemophilia-specific quality of life instrument for children and parents; Hemofilia-QOL, Hemophilia-specific quality of life instrument for adults; Hemolatin-QOL, Hemophilia-specific quality of life instrument for adults; Hemo-SatA, Hemophilia-specific treatment satisfaction questionnaire for adults and parents; HIV, Human Immune Deficiency Virus; HRQOL, > Health-Related Quality of Life; KINDL, Generic quality of life questionnaire for children; MCS, Mental Component Summary Score of the SF-36; Medtap, Hemophiliaspecific quality of life instrument for adults; PCS, Physical Component Summary Score of the SF-36; PedsQL, Generic quality of life questionnaire for children; QOL, Quality of life; SF-36, Generic Quality of Life questionnaire for adults (Short-form 36); TS, Treatment Satisfaction
1
Introduction
1.1
Hemophilia
Hemophilia is a congenital bleeding disorder affecting about 1 out of 10,000 people in the general population (Bulletin from the WHO, 1991). It is caused by a genetic defect in the genes coding for coagulation factor VIII (hemophilia A) (Gitschier et al., 1984) or factor IX
Quality of Life in Hemophilia
112
(hemophilia B) (Choo et al., 1982), essential for thrombin generation and consequently clot formation (Mann, 1999). The gene defect ends up in the production of no clotting protein at all or in a defective protein with low or absent activity (Gianelli and Green, 1996). Based on the residual plasma activity of these clotting proteins, we can distinguish severe hemophilia when the plasma clotting activity levels are below 1%, moderate hemophilia, when activity levels range between 1 and 5%, and mild hemophilia with levels between 5–25 (White et al., 2001). Severity of clotting defects correlates with phenotypic manifestation of these diseases, characterized by post-traumatic and spontaneous bleeds. In particular, patients more frequently suffer from bleeding in joints and muscles, which are progressively damaged by this bleeding: it eventually results in irreversible arthropathy and invalidity (Bulletin from the WHO, 1991). Bleeding events require prompt, appropriate and repeated treatment together with frequent and multidisciplinary monitoring. The mainstay of hemophilia management is based on intravenous replacement therapy with concentrates of coagulation factors (Report of the WHO, 2002; Mannucci, 2003). Treatment of hemophilia is based on replacement of missing factors by intravenous infusions when bleeding occurs or preventively. So called ‘‘on-demand’’ treatment is administered once or twice a day until complete healing of the episode: mild or minor bleeds can require only one or two infusions, while more severe life- or limb-threatening bleeds can require up to a month of daily infusion. Patients are on prophylaxis when they are treated at regular intervals 2–4 times weekly for a long period of time (months or years) in order to decrease or abolish the occurrence of bleeding episodes (Berntorp et al., 1995, 2003). Prophylaxis is today considered the treatment of choice of children with hemophilia: it is recommended to start in the first 2–3 years of life before repeated bleeds in joints (‘‘primary prophylaxis’’), in order to prevent permanent joint damage, or at least as soon as possible (‘‘secondary prophylaxis’’) and continued at least till the adulthood (Berntorp et al., 1995). In fact, prophylaxis has been demonstrated to be effective not only in lowering the bleeding occurrence rate, but also in preventing the development of hemophilic arthropathy with its consequence of pain and invalidity (Nilsson et al, 1996). It has been shown that a prophylaxis started after the age of 6–7 years is not capable to prevent progressive arthropathy (Nilsson et al, 1996).
. Table 112-1 Clinical burden of hemophilia Disease-related issues
Treatment-related issues
Spontaneous and post-traumatic bleeds
Frequent intravenous injections
Life-threatening bleeds
Central venous catheter
Joint bleeds
Catheter-related infections
Acute pain
Catheter-related thrombosis
Unpredictability of bleeding episodes
Allergic reactions
Hemophilic arthropathy
Blood-borne virus infections
Chronic pain
Inhibitor development
Invalidity
NSAID-related side effects
Mortality
1897
1898
112
Quality of Life in Hemophilia
Replacement factors are derived from plasma fractionation or produced by DNA engineered cells of Chinese hamster ovarian glands or baby hamster kidney (Key and Negrier, 2007). The half life of these concentrates range from 10–12 h to 18–20 h for factor VIII and FIX respectively (Morfini, 2003). These relatively short half-lives force patients to be infused frequently up to 2–3 times a day for a bleeding episode or every 24–36 h for maintaining trough levels above 1% (Bjo¨rkman, 2003). In 20% of patients, factor infusions can cause the development of inhibitory antibodies capable to neutralize the infused factors and making treatment completely ineffective (Wight and Paisley, 2003). The use of other coagulation factors (factor VII or a complex of Factor II, VII, IX and X) can by-pass the factor deficiency and the presence of factor inhibitors, but unfortunately they are less effective and difficult to manage and to monitor (Hay et al., 2006).
1.2
Quality of Life and Hemophilia
In medicine the term quality of life comprises the impact of the disease or the treatment on different aspects of life, better known as ‘health-related quality of life’ (HR-QOL). HR-QOL is not only influenced by a disease and its treatment but also by personal characteristics such as coping or internal locus of control as well as by living conditions and socio-economic status as well as social support (Bullinger, 1991). HR-QOL is more and more regarded as one of the most relevant health outcome measures in medicine (Cella et al., 1990). Health outcome data can help to optimize the treatment, which is essential in a cost-intensive chronic disease such as hemophilia where traditional outcome measures such as mortality and mobility are no longer influenced by diverse treatment options. Based on the definitions of quality of life (WHOQOL Group, 1993; Bullinger, 1991) the direct perception of patients is one of the most important issues in QOL assessment. The perspective plays an important role, since often observers overestimate some aspects of QOL of patients, whereas psychological aspects are often underestimated. Therefore, it is recommended to use self-rated or subjective measures where the patient is asked directly (Bowling, 1991). In some cases other-rated or ‘‘proxy’’ measures are used, e.g. in young children or patients unable to answer (e.g. mentally impaired patients). In this case parents or other caregivers related with the patient should be asked. In general instruments can be distinguished between generic and disease-specific measures. Generic instruments have the benefit that they can be used irrespectively of a specific disease, allowing the comparison of the quality of life of hemophilia patients with patients with other conditions or with the general population. Disease-specific instruments are more sensitive towards changes and can provide a detailed pattern of symptoms and impairments related to the disease (von Mackensen and Gringeri, 2005). Measures for adults and for children must be distinguished when choosing a QOL instrument. Instruments for children should be especially developed considering their developmental status and relevant aspects for different age groups. In small children the parents’ reports of children’s well-being are necessary, while in grown-up children the comparison between children’s and parent’s perspectives turns out to be of interest by itself (Eiser and Morse, 2001). Besides severe pain, arthropathy and disability hemophilia can affect as well patients’ health-related quality of life (HR-QOL). The modern management of hemophilia had visibly influenced not only clinical symptoms, orthopedic outcome and survival of patients but also their perceived HR-QOL (Miners et al., 1999).
Quality of Life in Hemophilia
2
Burden of Hemophilia
2.1
Clinical Burden
2.1.1
Bleeding
112
The impaired clotting activity of patients’ plasma makes the patient susceptible to terribly painful bleeding events, which can occur spontaneously or after insignificant traumas (Bulletin from the WHO, 1991). Bleeding can occur everywhere in the body: intracerebral hemorrhages are not infrequent and they can occur at any time in the life of the patients, but they are particularly frequent in the first decade (Klinge et al., 1999). Other more rare sites of internal bleeds are represented by the gastro-intestinal tract. Joints and muscles are the most frequent sites of bleeding, probably due to the stress provoked by upright position and movement (Ljung et al., 1990). The target joints of bleeding in children are mainly the ankles, followed by the knees and the elbow; adults are more frequently bleeding in knees (Donadel-Claeyssens, 2006). In general, hips and shoulders are less frequently affected, while wrists and fingers are rarely involved (Ljung et al., 1990). The unpredictability of bleeding leaves the patients in a condition of complete uncertainty, unable to prevent the bleeding and to plan the future activities without being influenced by its potential occurrence.
2.1.2
Treatment Complications
Beside the burden of intravenous injections since the very first year of age, patients have suffered from the risk of allergic reactions to factor concentrates (Giangrande, 2003; Jadhav and Warrier, 2000), and of the much more dangerous blood-borne virus infections, these concentrates being derived from donor plasma. These viral infections have severely struck the hemophilia population: in the early eighties with an epidemics of human immunodeficiency virus infection (CDC Report, 1982), and till the early nineties with hepatitis infections, such as hepatitis B, hepatitis C (Troisi et al., 1993) and more recently hepatitis A (Vermilen and Peerlink, 1994). Highly purified plasma-derived and more recently recombinant factor VIII and factor IX concentrates, available now for more than a decade, has enormously lowered these risks. In particular, with the introduction of new virucidal techniques (including pasteurization, chemical treatment and filtration), donor screening, and protein purification methodologies, the incidence of transmission of blood-borne viruses, such as hepatitis viruses or human immunodeficiency virus, is negligible (Bulletin from the WHO, 1991; Furie et al., 1994). On the other hand, other threats continue to be perceived for plasma-derived concentrates, represented by the definitely transmitted B19 parvovirus (Prowse et al., 1997), or prions causing the variant Creutzfeld-Jacob disease. These threats sustained the perception of risk, real or only theoretical, of other unknown or future blood-borne viruses.
2.1.3
Venous Access
Unfortunately, venous access for so frequent and prolonged treatment is not always available, particularly in children, and it often requires central venous devices that can offer a
1899
1900
112
Quality of Life in Hemophilia
user-friendly access. Percutaneous central venous catheters, such as Broviac and Hickman catheters, or indwelling catheters, such as Port-a-Caths, are the most used (Komvilaisak et al., 2006). Of course they require surgery and general anesthesia, and families must be educated in their use. Nevertheless, their use is at high risk of catheter bacteria contamination, which can cause septic infections, or thrombosis of the vessel where they are placed, with additional burden for the patients and the requirement of systemic antibiotic therapy (Ljung, 2007; Ewenstein et al., 2004). Artero-venous fistula has recently been used in order to provide a safer venous access (Santagostino et al., 2003).
2.1.4
Pain and Arthropathy
The recurrent presence of blood in joints causes progressive joint degeneration with increasing function impairment and consequent invalidity and chronic pain (Aledort et al., 1994). The latter requires the frequent use of non-steroidal anti-inflammatory drugs, with the load of side effects. With repeated bleeding in a joint, the synovium is a state of chronic inflammation, called synovitis that eventually results in hypertrophy, causing the joint to appear grossly enlarged and swollen (Rosendaal et al., 200). This swelling is usually not tense, nor is it particularly painful, initially. Muscle atrophy can often occur at this stage, while a relatively good range of motion of the joint is preserved. Synovitis, in turn, facilitates the occurrence of recurrent bleeds due to inflammation, synovial hypertrophy and hypervascularization. Persistent chronic synovitis and recurrent hemarthroses results in irreversible damage to the joint cartilage. It is characterized by advancing cartilage loss, a progressive arthritis with secondary soft tissue contractures, muscle atrophy, and angular deformities (Rosendaal et al., 200). The pain now is present, excruciating and persistent. With progressing chronicity of the arthropathy, there is due to the progressive fibrosis of the synovium and the capsule produces less swelling, but loss of motion is common with flexion contractures. This is caused by advancing cartilage space narrowing, irregular articular bone surfaces, due in turn to development of bone erosions and subchondral bone cysts (Rosendaal et al., 200). Pain at this stage may or may not be present, but invalidity is invariably reported.
2.1.5
Presence of Inhibitors
Another problem is represented by the development of inhibitory antibodies against the clotting factor infused (Wight and Paisley, 2003). These antibodies reduce or abolish the efficacy of replacement therapy, precluding the possibility to stop the hemorrhage and to rapidly absorb blood from the bleeding site, with the already mentioned consequence on the joint status. Factor inhibitors occurred frequently: cumulative incidence is reported to be up to 35% of previously untreated patients, particularly when recombinant factors are administered (Wight and Paisley, 2003). The presence of inhibitors greatly complicates hemophilia management (Gringeri et al., 2003) and the failure to adequately treat or prevent bleeding in patients with inhibitors leads to increased morbidity and mortality over the lifetime of the patient (Darby et al., 2007). Current management of patients with inhibitors has improved clinical outcomes (Gringeri et al., 2003), however, they are particularly difficult as they are neither satisfactory in all clinical situations nor applicable for all patients. The difficulty associated with treating patients with inhibitors mandates that a number of variables be taken into account when a
Quality of Life in Hemophilia
112
clinician makes therapeutic decisions. The site of the bleed, product efficacy and safety, patient age, previous exposure to plasma-derived factor concentrates and cost are important considerations (Berntorp et al., 2006; Teitel et al., 2007; Hay et al., 2006; Gringeri and Mannucci, 2005). In addition, both inhibitor titer and the patient’s responsiveness to factor replacement therapy are important in developing treatment strategies. Patients who do not experience anamnesis (i.e. a spike in titer upon exposure to factor concentrate) and who have low-titer inhibitors (5 Bethesda units [BU]) may be treated with higher than normal doses of factor concentrates given at standard or more frequent intervals, thereby overwhelming the inhibitor (Rizza, 1984). Decreasing plasma antibody titers via plasmapheresis or immunoadsorption may also be considered as a rescue treatment option (Freiburghaus et al., 1998). Patients with high-titer inhibitors (>5 BU) are the most difficult to manage and often require treatment with bypassing agents (Lloyd et al., 2003). The two most commonly used bypassing agents available for the treatment of patients with hemophilia and high-titer inhibitors are Factor Eight Inhibitor Bypassing Activity (FEIBA; Baxter AG, Vienna, Austria), an activated prothrombin complex concentrate (aPCC), and recombinant activated factor VII (rFVIIa; NovoSeven1, Novo Nordisk A/S, Bagsvaerd, Denmark) (Lusher et al., 1998). Both agents have their advantages, i.e. efficacy and feasibility as home treatment (Negrier et al., 1997; Santagostino et al., 1999), and disadvantages, namely the safety profile (Ehrlich et al., 2002; Roberts et al., 2004) and the lack of a reliable bedside monitoring test (Barrowcliffe, 2004). In conclusion, hemophilia represents a very high burden for patients and families since the disease manifestations are unpredictable, painful and leading inexorably to invalidity, and its treatment is demanding and often complicated by adverse events.
2.2
Psychological Burden
2.2.1
Prenatal Diagnosis
Prenatal diagnosis (PND) is an important issue in the comprehensive care of hemophilia patients and carriers. In a study conducted in the US (Kraus and Brettler, 1988) hemophilia had an impact on family planning, which was influenced by parental fulfillment, availability of medical care and education for their affected child. Even though 43% of mothers would consider prenatal diagnosis, only 17% would have terminated a pregnancy if the fetus were found to have hemophilia. The majority of mothers in the group did not view having a child with hemophilia as an insurmountable burden on their lives. The authors concluded that the disease appeared to have little impact on family planning. Same attitude seems to be shared by Italian carriers of hemophilia, who would accept to terminate a pregnancy of a hemophilic fetus only in 16.7% of them (Karimi et al., 2004). By contrast, 58.2% of Iranian women would be available to terminate the pregnancy if the fetus were hemophilic. The authors concluded that the greater rate of acceptability of abortion in Iranians may be due to differences in the quality of patient care in the two countries (Karimi et al., 2004). In Sweden carriers who did not choose prenatal diagnosis (PD) often abstained from further pregnancies after the birth of a hemophilic child, and they had significantly fewer children than the remaining carriers, as well as fewer children than women in the control group. Carriers who have experienced the complications of hemophilia or its treatment appear to be more in favor of PD than women whose hemophilic children have received modern treatment without complications (Tedgard et al., 1999).
1901
1902
112 2.2.2
Quality of Life in Hemophilia
Stress and Coping
Hemophilia is a threatening disease which can have a high impact on patients’ lives. Therefore coping with the disease and its resulting stress becomes quite important. > Coping strategies are different ways how people deal with stressful events both behavioral and psychological. Coping strategies can be distinguished between problem-focused and emotion-focused strategies or between active and avoidant coping strategies. ‘‘Problem-focused coping refers to efforts to improve the troubled person-environment relationship by changing things, for example, by seeking information about what to do, by holding back from impulsive and premature actions, and by confronting the person or persons responsible for one’s difficulty. Emotion-focused coping refers to thoughts or actions whose goal is to relieve the emotional impact of stress. These are apt to be mainly palliative in the sense that such strategies of coping do not actually alter the threatening or damaging conditions but make the person feel better.’’ (Monat and Lazarus, 1991). Personal characteristics and the type of stressful event are determining the type of coping strategy used, e.g. people typically employ problem-focused coping to deal with potential controllable problems (e.g. work-related or family-related problems), whereas stressors perceived as less controllable, such as chronic physical health problems, prompt more emotionfocused coping. Active coping strategies are considered better for dealing with stressful events, while avoidant coping strategies appear to be a psychological risk factor for adverse responses to stressful life events (Holahan and Moos, 1987). A Korean study showed that young adults with hemophilia attempt to live a normal life, by means of the following coping strategies found in these population: (1) pretending as if he is not hemophiliac; (2) relieving the burden; (3) maintaining best physical conditions; (4) becoming independent; and (5) reconciliation with their mothers (Yi et al., 2003). In a comparison between adult hemophilia patients and normal controls in Italy it was found that hemophilia patients had less self-esteem than healthy controls (Canclini et al., 2003). In an European study, hemophilic children were shown to use more positive coping strategies such as ‘trying to spend a lot of time with other people’ (77.5%), ‘trying to see good sides of things’ (62.6%) and ‘trying to do something or talk to someone’ (59.8%); only few children showed negative coping strategies such as ‘blaming themselves for the disease’ (10.8%) and ‘being furious about it’ (7.3%) (Bullinger et al., 2003). Moreover, results of a Dutch study revealed that internal locus of control1 and favorable psychological characteristics could reduce the perceived seriousness of hemophilia (Triemstra et al., 1998). Hemophilia has not only an impact on the concerned patient, but as well on his family, especially on his parents. In the UK, parents of hemophilic children were asked about how hemophilia impacted their lives (Beeton et al., 2007). Four main areas were emerged: ‘initial experience’, ‘managing the condition’, ‘engaging with others’ and ‘developing mastery’. Parents seemed more stressed than their children with hemophilia and reported an impact of the disease as well on their partnership and siblings of the hemophilic child. After the initial difficulties in dealing with the disease, parents’ strategies improved over time. Authors
1
Locus of control is considered an important aspect of personality and can be destinguished between internal locus of control (a person believes that he controls himself and his life) or external locus of control (a person believes that his environment, some higher power or other people control his decisions and his life) (Rotter, 1954).
Quality of Life in Hemophilia
112
recommended professional support for parents with a hemophilic child in order they could restructure their lives. In Italy parents of hemophilic children underwent a counseling programme. It could be demonstrated that after the counseling programme emotion-focused coping strategies could be decreased and problem-focused coping strategies increased, while avoidance-focused coping strategies remained the same (Bottos et al., 2007). In a study conducted in the UK active coping styles (‘problem focused’, ‘social support’ and ‘cognitive restructuring’) were used significantly more by the parents; but they ‘blamed themselves’ and ‘blamed others’ significantly more than did children. The passive strategy of ‘wishful thinking’ was used more frequently by children (Miller et al., 2000).
2.2.3
Anxiety and Depression
The clinical burden of hemophilia leads often to anxiety and depression of patients and their families. As far as anxiety is concerned two different types should be distinguished, namely state and trait anxiety (Spielberger, 1972). State anxiety is defined as an unpleasant emotional arousal, which varies in intensity and time in face of threatening demands or dangers, while trait anxiety reflects the existence of stable individual differences, which are relatively durable and independent from the context, reflecting the tendency to respond in the anticipation of threatening situations. Since the 1980s, mortality in the hemophilia population has been dominated by human immunodeficiency virus (HIV) infection, which was transmitted via factor replacement therapy. Nowadays only about 10% of adult hemophilia patients’ wit HIV infections are still alive. This disease represents a huge burden not only for the concerned patients, but it induces anxiety in the non-infected patients too. In an Australian study it was demonstrated that HIV seropositive patients without symptoms of disease progression had higher levels of State Anxiety compared with seronegative subjects. The State Anxiety scores were predicted by HIV infection or alternatively CD4 + T-cell levels (Jones et al., 1995). In Italy, one third of the adult hemophilia population was still worrying about the risk of infections irrespective of being HIV seropositive or seronegative (von Mackensen et al., 2007). Another cause of anxiety is represented by the risk of inhibitor development. In Italy, it has been found that more than 40% of adult patients with hemophilia worried about inhibitors (von Mackensen et al., 2007). In a qualitative study in the UK it was found that anxiety and depression were associated with the daily management of symptoms, unpredictable bleeds and concerns about the future. Patients reported that their disease had an impact on their work, education, social activities and their family life, especially they felt stigmatized and discriminated in working settings (Barlow et al., 2007). Problems with work were reported as well by Hartl et al. (2008); the Authors found that Austrian hemophilia patients were significantly more unemployed than healthy controls. In a French study it could be demonstrated that the way that information was provided about treatment-risks could induce more anxiety in patients: 12% of hemophilia patients receiving written information showed anxiety and 6.6% depression (Magli-Barioz et al., 2004). After delivery of the information sheet only 70% of the patients believed that the HIV was inactivated during preparation of the product. The Authors recommended giving patients oral information about the disease and its treatment instead or in addition to written information. A counseling programme in Italy was able to reduce the state anxiety levels in parents of hemophilic children (Bottos et al., 2007). Another study conducted in Italian couples of
1903
1904
112
Quality of Life in Hemophilia
hemophilic children showed that mothers were more depressed and anxious about the disease of their children than fathers (Saviolo-Negrin et al., 1999). Canclini et al. (2003) did not find differences concerning depression between hemophilia patients and controls in Italy. Another study in Italy demonstrated that elderly hemophiliacs reported in the depression scale significant more impairments than normal controls, their symptoms could be classified as mild depression (von Mackensen et al., 2008).
. Table 112-2 Psychological burden of hemophilia Clinical aspects Prenatal diagnosis
Psychological burden Problems (work, education, partnership, family)
Health condition
Coping
Complications (inhibitors, invalidity)
Anxiety
Treatment (regimen, venous access)
Depression
Side effects (infections, allergic reaction)
Stigmatization & discrimination
2.3
Economical Burden
Although rare, hemophilia has a high impact on society and healthcare system, because it is costly and challenging for its complexity. In fact, all the different aspects of hemophilia management entail the absorption of a huge amount of human and economic resources (Gringeri et al., 2003; Schramm et al., 2002) and greatly influence patients’ well-being (Miners et al., 1999; Barr et al., 2002). The cost of hemophilia care in a time of ever-increasing awareness of limited medical resources is threatening the maintenance of the achieved level of care (Aledort, 2000). On the other hand, many developing countries cannot afford these costs, leaving patients without an adequate treatment (Bolton-Maggs, 2006; Evatt, 2005). As a matter of facts, the treatment of hemophilia is highly expensive: it was calculated that the average annual overall costs of replacement therapy in adult patients with hemophilia was ranging from $ 379 to 1’015 per Kg body weight per year for on-demand treatment, and from $ 1’730 to 6’045 per Kg body weight per year for prophylaxis (Schramm et al., 2002). It corresponds to an average annual cost of $ 24’635–392’925 per patient, hemophilia scoring the fifth expensive chronic disease. The costs of care of patients may be even higher when patients develop an inhibitor, due to the extremely high costs of by-passing agents and their partial efficacy. In a study carried out in Italy, it was found that the average cost of an adult patient with inhibitors was about €18.000 monthly, twice of the costs for patients without inhibitors (€ 8.900) (Gringeri et al., 2003). More than 97% of the costs required for hemophilia treatment are allocated to factor concentrates; only 1–3% is costs for medical visits, hospitalizations, surgery or treatment for blood-borne viral infections. Lacey (2002) found that additional cost for inhibitor patients was $ 51,553 per QALY gained which is comparable with that derived for currently reimbursed healthcare strategies, including hospital dialysis (i.e., $ 57,053). The treatment of hemophilia, with or without inhibitors, places a burden not only on the healthcare system but also directly on the patient. In effect, these other expenses directly out of
Quality of Life in Hemophilia
112
the patient’s and family’s pockets have never been quantified: the costs for attending frequent medical evaluations in terms of costs of traveling, working days lost by the patients and by the care givers, costs for assisted physical activity, which does not represent leisure costs, but a need in order to maintain and improve joint and muscle functioning. These huge overall costs for the society should be considered in the context of rarity of the disease in the general population, of its morbidity and mortality, and last but not least of quality of life. In truth, if we consider that the cost of one patient with inhibitor is about 220,000 € a year we can reasonably be shocked, but we must keep in mind that the prevalence of this complication is about 0.5 out of 100,000 inhabitants. Consequently, the overall bill for a country like Italy, with 60 million inhabitants, will be around 66 million €, more than 100 times less compared to diabetes (7 billion €), which costs only 2’500 €/patient a year, but with 2.8 million patients in Italy. Hemophilia with inhibitors costs 85 times less than chronic ischemic cardiopathy (5.7 billion €), with which 1.2 million people are ill in Italy, or 20 times less than end-stage renal disease (1.3 billion), which hits 45,000 Italians. In addition, this cost of patients with hemophilia and inhibitors becomes less impressive when one calculates that the cost of care met by each Italian citizen for all these patients is approximately € 0.7 per year (Gringeri et al, 2003).
3
Quality of Life Instruments in Hemophilia
For an adequate assessment of HR-QOL in patients with hemophilia validated instruments are necessary (von Mackensen and Gringeri, 2005). Most of the publications related to QOL and hemophilia mentioned QOL as an important condition of the patient. In these studies it was mainly stated that e.g. prophylactic treatment or home therapy is improving the QOL of hemophilic patients, but the assessment of QOL with standardized instruments was lacking. Only in few studies measurements for the assessment of QOL were used. In those studies including QOL assessment mainly generic instruments have been used. Since generic instruments are not able to provide a clear pattern of symptoms or impairments related to a specific disease and are not sensitive enough to treatment consequences (Bowling, 2001), we want to describe in this chapter only hemophilia-specific instruments. . Table 112-3 Internationally developed hemophilia-specific HRQOL Questionnaires Acronym
Name
Age-group
Reference
Country of origin
HEMO-QOL
Hemophilia Quality of Life Questionnaire
Pediatric
v. Mackensen et al., 2004
Medtap
n.a.
Adult
Flood et al., 2003 USA, Canada, Germany, Spain
Hemolatin-QOL Disease-specific quality-of-life Adult questionnaire to adults living with hemophilia Na = not available
Remor et al., 2004
Europe
South America
1905
1906
112
Quality of Life in Hemophilia
Only in the last 6 years disease-specific instruments have been developed for the assessment of quality of life in hemophilic children and adults. Some of these questionnaires have been developed internationally (see > Table 112-3) such as the Hemo-QOL in six European countries (Germany, Italy, Spain, UK., France, Netherlands) (von Mackensen et al., 2004), the Medtap questionnaire in USA, Canada, Spain and Germany (Flood et al., 2003) and the HemolatinQOL (Remor et al., 2004) in Latin America, whereas others have been developed only on a national level (see > Table 112-4) such as the CHO-KLAT in Canada (Young et al., 2004), the American Quality of Life Questionnaire for Young Patients (Manco-Johnson et al., 2004), the Hem-A-QOL in Italy (von Mackensen et al., 2004) and the Spanish Hemofilia-QOL (Arranz et al., 2004). Following these instruments will be described distinguishing pediatric and adult measures.
. Table 112-4 Nationally developed hemophilia-specific HRQOL Questionnaires Acronym
Name
Age-group
Reference
Country of origin
CHO-KLAT
Canadian hemophilia outcomes-kids’ life assessment Tool
Pediatric
Young et al., 2004
Canada
Na
Quality of life for young patients
Pediatric
MancoJohnson et al., 2004
USA
Hemofilia-QOL Disease-specific quality-of- Adult life questionnaire to adults living with hemophilia
Arranz et al., 2004
Spain
Hem-A-QOL
v. Mackensen Italy et al., 2004
Hemophilia QOL questionnaire for adults
Adult
Na = not available
3.1
Pediatric Instruments
All questionnaires for children are self-administered, but the American Quality of Life for Young Patients questionnaire, which is designed for smaller children, aged 2–6 years, and completed by their parents.
3.1.1
Hemo-QOL
The first hemophilia-specific questionnaire (Hemo-QOL) was developed using parents’ assessment of children’s quality of life as well as clinical expert consensus on relevant dimensions in order to construct a quality of life instrument for children. The Hemo-QOL is available as a self-administered questionnaire for hemophilic children as well as a proxy version for their parents.
Quality of Life in Hemophilia
112
The Hemo-QOL has been pilot-tested in 52 patients regarding completion time, acceptance and comprehensibility of the questionnaire using cognitive debriefing techniques (Bullinger et al., 2002). The questionnaire was validated in 339 hemophilic children from six European countries (German, French, Italian, Spanish, Dutch, English) (von Mackensen et al., 2004) and is available for download from the project website (www.haemoqol.org). Linguistic validations of the Hemo-QOL currently exist for 30 languages. Three different age-group versions (I:4–7, II:8–12, III:13–16 years) are available for children in order to better capture all the aspects of their well-being in the different developmental phases and 3 proxy versions for their parents. These versions differ in the number of items (I: 21, II: 64, III: 77 items) and domains (I: 8, II: 10, III: 12 domains). Common domains are ‘physical health’, ‘feeling’, ‘view’, ‘family’, ‘friend’, ‘others’, ‘sport & school’ and ‘treatment’, additional domains for age group II and III are ‘perceived support’ and ‘dealing’ and for adolescents the domains ‘future’ and ‘relationship’ are added. Psychometric characteristics of the Hemo-QOL revealed acceptable values for > reliability in terms of internal consistency (ranging for the total score from Cronbach’s a = .85–.91 for the different age group versions I-III) and Test-Retest reliability for age group II and III (ranging from r = .90–.92). > Validity was satisfactory in terms of convergent and discriminant validity. As regards convergent validity, the Hemo-QOL was correlated with scales measuring similar concepts, namely the KINDL, the chronic-generic subscale of the KINDL and the General Health Question (GHQ) of the Child Health Questionnaire. Correlations for the GHQ with the scales of the Hemo-QOL were satisfactory. Correlations between HemoQOL and the KINDL were moderate to high (depending from the age group) in that scales correlated within the respective physical, social or psychological domains. High correlations were found between most of the subscales of the Hemo-QOL and the chronic-generic subscale of the KINDL. Correlations were negative because a high value in the Hemo-QOL implies a high impairment in quality of life while a high value of the KINDL and GHQ imply a good quality of life. Discriminant validity was assessed using clinical information differentiating HemoQOLscoreswithregardtotreatmentandcondition-relatedinformation.Differenceswerefoundfor numberofjointandmajorbleeds(vonMackensenetal.,2004;Gringerietal.,2004)indicatinghigher impairment inqualityoflifeforchildrenwithmorebleeds. Theseindicatorswerelesssatisfactory in young children compared to older children (see > Table 112-5).
. Table 112-5 Psychometric Characteristics of the pediatric Hemo-QOL questionnaire (n = 312) Scale
No of items Min Max Mean
SD
a
Test-retest (r)
Convergent validity (chronic generic) (r)
Age group I (n = 90) Physical health
4
4
16
7.03
3.15 0.55
–
.33
Feeling
3
3
15
5.13
3.13 0.82
–
.38
View
2
2
10
3.53
2.40 0.69
–
0.42 0.44
Family
4
4
20
9.47
3.91 0.66
–
Friend
1
1
5
1.74
1.21 –
–
–
Others
2
2
10
3.87
2.20 0.56
–
0.53
1907
1908
112
Quality of Life in Hemophilia
. Table 112-5 (continued) Scale
No of items Min Max Mean
SD
a
Test-retest (r)
Convergent validity (chronic generic) (r)
Sport
3
3
12
5.48
2.64 0.49
–
–
Treatment
2
2
10
4.00
2.34 0.45
–
–
21
21
81
40.25 13.65 0.85
–
n.a.
27
11.64
Total
Age group II (n = 117) Physical health
7
7
4.80 0.78
.87
.33
Feeling
7
7
View
9
9
39
9.56
3.36 0.69
0.87
.29
14.89
5.86 0.79
0.84
Family
5
5
21
9.28
.37
4.02 0.68
0.76
.34
Friend
4
4
20
11.03
4.47 0.71
0.81
Perceived support
4
4
20
11.84
4.07 0.66
0.74
Others
6
6
25
8.50
3.33 0.74
0.76
– .36 –
Sport
8
8
28
15.49
5.53 0.67
0.79
Dealing
7
7
35
16.32
5.54 0.66
0.61
–
Treatment
7
7
25
12.13
4.06 0.60
0.67
–
64
79
200 120.68 22.66 0.85
0.90
n.a.
Total
.28
Age group III (n = 95) Physical Health
7
7
27
12.51
4.46 0.76
.79
.45
Feeling
8
8
36
11.71
5.15 0.87
.89
.56
10
10
43
18.00
6.92 0.86
.86
.52
8
8
26
13.62
4.37 0.69
.76
.51
View Family Friend
4
4
20
11.20
3.90 0.69
.81
Perceived Support
4
5
20
12.40
3.75 0.63
.57
Others
6
6
22
9.07
3.85 0.75
.71
Sport
9
9
35
19.18
7.24 0.76
.78
Dealing
7
7
35
15.94
5.38 0.68
.71
Treatment
8
8
29
15.41
5.14 0.67
.63
Future
4
4
16
8.66
3.00 0.52
.64
9
2.75
1.46 0.73
.90
240 150.42 30.62 0.91
.92
Relationship Total
2
2
77
92
– .45 – .37 – .33 .24 n.a.
a = Cronbach’s a(internal consistency); r = Pearson product moment correlation; n.a. = not available
In addition to the three long versions, two short versions have been developed for age group I (16 items) and for age group II/III (35 items), which still have to be validated in clinical trials. An ultra-short version with 8 items (Hemo-QOL Index) has been constructed recently allowing the comparison between different age-groups (Pollak et al., 2006).
112
Quality of Life in Hemophilia
3.1.2
CHO-KLAT
Another hemophilia-specific questionnaire for children is the Canadian Hemophilia Outcomes – Kids Life Assessment Tool (CHO-KLAT) (Young et al., 2004). The CHO-KLAT was developed including the perspectives of children with hemophilia giving the priority to children to ensure that the measure reflected the perspectives of children and could be used by self-report by children. The CHO-KLAT is available for children 5–18 years of age and includes parents- and self-report forms. The CHO-KLAT consists of 35 items and has a single summary score representing overall quality of life based on questions regarding: treatment, physical health, family, future, feeling, understanding of hemophilia, other people & friends, and control over your life. The instrument was pilot tested in a group of 52 children and is currently field tested in a larger group. The CHO-KLAT is available in English and French and is currently being translated into German, Spanish and Mandarin. Psychometric characteristics of the CHO-KLAT proved to be very good in terms of testretest reliability and inter-rater reliability and showed high convergent validity in terms of correlation with the hemophilia-specific Hemo-QOL questionnaire, but lower correlation with the generic PedsQL questionnaire (see > Table 112-6). In terms of discriminant validity no significant differences were found for clinical subgroups comparing patients with moderate and severe hemophilia (Young et al., 2006).
3.1.3
Quality of Life Questionnaire for Young Patients
The hemophilia-specific quality of life questionnaire for young patients is not a self-rated instrument, but a proxy instrument for parents of hemophilic children aged 2–6 years (Manco-Johnson et al., 2004). The questionnaire was developed based on interviews with physicians, nurses and parents of children with hemophilia. The questionnaire has been pilot-tested in 44 parents and validated in 103 parents of hemophilic children and 249 parents of healthy controls in the US. The questionnaire consists of 39 items pertaining the following 9 domains: ‘somatic symptoms’, ‘physical functioning, ‘sleep disturbance’, ‘stigma’, ‘social functioning’, ‘fear’, ‘resentment/reaction’, ‘energy level’, ‘mood/behavior’ and ‘restrictions’. Psychometric characteristics revealed good values for reliability in terms of internal consistency (ranging from Cronbach’s a = .73–.94). Convergent validity was analyzed correlating . Table 112-6 Psychometric characteristics of the pediatric CHO-KLAT questionnaire (n = 52) [table modified, Young et al., 2006] Reliability (ICC)a CHO-KLAT version Children Parents a
No of items 35
Convergent validity
Testretest
Child-parent concordance
HemoQOL (r)
PedsQL (r)
74.56 14.08
0.74
0.75
0.78
0.59
74.52 11.60
0.83
–
–
Mean
SD
ICC: Intra-Class Coefficient Model calculating values for test-retest and inter-rater reliability
1909
1910
112
Quality of Life in Hemophilia
hemophilia-specific subscales with the Impact-on-Family Scale or the Functional Status; correlations were moderate in magnitude. Discriminant validity was assessed analyzing differences for patients receiving primary prophylaxis compared to patients on on-demand treatment. Patients on primary prophylaxis scored significantly better on ‘somatic symptoms’ and ‘physical functioning’.
3.2
Adult Instruments
3.2.1
Hem-A-QOL
The Hem-A-QOL was designed for adult patients with hemophilia and developed based on patient-based focus groups (n = 32) and focus groups organized with physicians and nurses. The Hem-A-QOL was pilot tested in 106 Italian hemophilia patients (von Mackensen et al., 2004) and has been field tested (n = 233) in the Italian COCHE study (Cost of Care in Hemophilia) (von Mackensen et al., 2005). The questionnaire has been linguistically validated in 20 languages. The Hem-A-QOL consists of 46 items pertaining to 10 dimensions (‘physical health’, ‘feelings’, ‘view’, ‘sport & leisure time’, ‘work & school’, ‘dealing’, ‘treatment’, ‘future’, ‘family planning’, ‘relationship’) and a total score. The Hem-A-QOL has a core instrument with 27 shared items with the pediatric HemoQOL that allows a comparison between HRQOL of adults and children. A specific version of the Hem-A-QOL has been developed for elderly patients (von Mackensen et al., 2008). The psychometric characteristics showed quite good reliability values in terms of internal consistency ranging from Cronbach’s a = .74–.88 for the subscales and a = .96 for the total score (see > Table 112-7) and test-retest ranging from r = .57–.93. Convergent validity . Table 112-7 Psychometric characteristics of the Hem-A-QOL questionnaire (n = 233)
Scale
SD
Internal consistency (a)
12.88
5.0
0.87
0.79
.70
No of items Min Max Mean
Test- Convergent retest validity (PCS) (r) (r)
Physical Health
5
2
25
Feeling
4
1
20
8.08
4.2
0.88
0.91
.59
View
5
1
24
11.57
5.0
0.83
0.86
.64
Sport & Leisure
5
1
25
15.68
5.9
0.79
0.80
.61
Work & School
4
1
20
7.61
4.1
0.86
0.82
Dealing
3
1
15
5.50
2.5
0.76
0.57
Treatment
8
1
35
18.56
6.8
0.77
0.91
.47
Future
5
1
25
13.40
5.1
0.74
0.74
.63
Family Planning
4
1
20
5.72
3.6
0.76
0.83
.31
5.23
.70 n.s.
Relationship & Sexuality
3
1
15
2.9
0.82
0.93
.26
Total
46
6
193 100.29 33.9
0.96
0.87
.74
a = Cronbach’s a(internal consistency); r = Pearson product moment correlations
Quality of Life in Hemophilia
112
analysis revealed that the dimensions of the Hemo-A-QOL correlated significantly with the dimensions of the SF-36, only the dimension’dealing’ was not correlated with the Physical Component Summary Score (PCS) of the SF-36. High correlations were found for the dimensions ‘physical health’, total score of the Hem-A-QOL and the Physical Summary Component (PCS) of the SF-36 and the dimensions ‘feeling’, ‘view’ and the Mental Component Summary Score (MCS) of the SF-36. Discriminant validity testing showed significant differences for severity of hemophilia in almost all dimensions of the Hem-A-QOL.
3.2.2
Medtap Questionnaire (Hemo-QOL-A)
The MedTap questionnaire was developed based on literature review, focus groups with 31 adult hemophilia patients and two expert meetings with hematologists and outcome researchers. The questionnaire was pilot tested in five English-speaking patients from the US and has been field tested in 221 patients from the US, Canada, Spain and Germany (Flood et al., 2003; Reutz et al., 2008). The questionnaire is available in English, Spanish, German and French. The MedTap questionnaire consists of 41 items pertaining to six dimensions (‘physical functioning’, ‘role functioning’, ‘worry’, ‘consequences of bleeding’, ‘emotional impact’, and ‘treatment concerns’) and four independent items. The psychometric characteristics showed good to excellent values for reliability in terms of internal consistency (Cronbach’s alpha ranging from a = .75–.95) and test-retest reliability with intraclass correlation coefficients > 0.80 for all dimensions, but ‘emotional impact’ (0.79) (see > Table 112-8; Revtz et al., 2008). Validity was analyzed in terms of concurrent and discriminant validity. Concurrent validity between the MedTap questionnaire subscales and the Physical Component Summary Score of the SF-36 were good (correlations ranging from r = .35–.85). The MedTap questionnaire was able to discriminate between clinical subgroups
. Table 112-8 Psychometric characteristics of the medtap questionnaire (n = 221) [table modified, Rentz et al., 2008] Scale Physical functioning Role functioning
No of items
Mean SD
Internal consistency (a)
Test-retest Convergent validity (ICC) (PCS) (r)
9
66.82 23.9
0.90
0.93
0.85 0.57
11
79.38 17.3
0.88
0.91
Worry
5
73.57 24.2
0.81
0.84
0.44
consequences of bleeding
7
72.21 22.0
0.87
0.87
0.46
emotional impact
6
76.92 18.1
0.75
0.79
0.48
treatment concerns
3
60.05 30.6
0.81
0.82
0.35
41
73.12 17.0
0.95
0.93
0.67
Total
a = Cronbach’s a(internal consistency); ICC = Intraclass correlation coefficient; convergent validity (correlation coefficient r)
1911
1912
112
Quality of Life in Hemophilia
such as severity and HIV infection. Subjects with severe hemophilia reported worse QOL compared to patients with moderate and mild hemophilia. Same was found for patients with HIV infections compared to HIV-negative patients.
3.2.3
Hemofilia-QOL
The hemophilia-specific quality-of-life assessment measure for adults (Hemofilia-QOL) has been developed for patients with hemophilia living in Spain. The development was based on semi-structured interviews with Spanish hemophilia patients (n = 43) and health professionals (n = 26). The self-report modular instrument was pilot tested in 35 hemophilia patients (Arranz et al., 2004) and has been validated in 121 adult patients (Remor et al., 2005). The questionnaire is available in Spanish. The Hemofilia-QOL questionnaire assesses nine relevant HRQOL domains for patients with hemophilia (‘physical health’, ‘daily activities’, ‘joint damage’, ‘pain’, ‘treatment satisfaction’, ‘treatment difficulties’, ‘emotional functioning’, ‘mental health’, ‘relationships & social activity’) and consists of 36 items. Psychometric characteristics were analyzed for reliability (in terms of internal consistency and test–retest reliability) and validity (in terms of concurrent, criterion and discriminant validity). The Hemofilia-QOL questionnaire had acceptable internal consistency ranging from Cronbach’s a = .58–.88 for the subscales and a = .95 for the total score and test-retest ranging from r = .79–.92 (see > Table 112-9). The Hemofilia-QOL showed excellent concurrent validity values (significant correlations with the dimensions of the SF-36 ranging from r = .17–.77) and external clinical criterion validity (hemophilia clinical status) and sensitivity (health status changes) as well.
. Table 112-9 Psychometric characteristics of the Hemofilia-QOL questionnaire (n = 121) [table modified, Remor et al., 2005] Scale
No of items
Min Max Mean SD
Internal consistency (a)
Test-retest (r)
Physical health
8
3
32
21.2
7.1
0.88
0.86
Daily activities
4
0
16
10.2
4.5
0.88
0.79
Joint damage
3
0
11
7
2.8
0.79
0.79
Pain
2
0
8
4.7
2.2
0.78
0.81
Treatment satisfaction
2
0
8
6.2
1.5
0.66
0.91
Treatment Difficulties
4
2
16
12.1
3.1
0.58
0.85
Emotional functioning
5
2
20
13.2
4.3
0.74
0.83
Mental health
3
0
12
7.7
2.9
0.74
0.82
Relationships & social activity
5
3
20
15.5
4.5
0.85
0.85
36
28
138
26.7
0.95
0.92
Total
99
a = Cronbach’s a (internal consistency); r = Pearson product moment correlations
Quality of Life in Hemophilia
112
Discriminant validity testing showed significant differences for chronic pain in all dimensions of the Hemofilia-QOL, and in half of the dimensions for hemarthrosis, bleedings and infections.
3.2.4
Hemolatin-QOL
The Hemolatin-QOL was designed for adult patients with hemophilia from Latin-America and developed based on patient interviews (n = 50) in eight countries (Argentina, Brazil, Colombia, Cuba, Guatemala, Panama, Uruguay and Venezuela). The Hemolatin-QOL consists of 47 items pertaining to nine dimensions (‘pain’, ‘physical health’, ‘emotional functioning’, ‘social support’, ‘activities & social functioning’, ‘medical treatment’, ‘mental health’, ‘satisfaction with condition’, ‘general well-being’) and one single item for ‘general health’ (Remor, 2005). The questionnaire is available in Spanish and Portuguese for South America. Psychometric evaluation is in progress within several LatinAmerican countries (Remor et al., 2004).
3.3
Treatment Satisfaction
Treatment satisfaction can be defined as the individual rating of important attributes of the process and outcomes of his treatment experience (Weaver et al., 1997). Only recently the assessment of treatment satisfaction has become of interest. This research area can provide important information for clinical trials and disease management programs. Since treatment of hemophilia can impact on the quality of life of hemophilic patients the assessment of satisfaction with this treatment becomes quite important. The Hemo-SatA is the first hemophilia-specific treatment satisfaction questionnaire for patients with hemophilia, which was developed in Italy (von Mackensen et al., 2002; 2004) and which is available in 19 languages. The questionnaire was pilot tested in 51 patients and has been validated in the Italian COCHE Study (n = 233) (von Mackensen et al., 2005). The Hemo-SatA is a well accepted short questionnaire for the assessment of satisfaction with hemophilia treatment in adults and for parents of hemophilic children (Hemo-SatP). . Table 112-10 Psychometric characteristics of the Hemo-SatAquestionnaire (n = 233) Scale
No of items
Min
Max
Mean
SD
Internal consistency (a)
10
10
43
22.20
6.1
0.73
Efficacy
6
6
22
11.29
3.5
0.75
Burden
4
4
15
7.29
2.7
0.68
Specialist/Nurse
7
7
29
11.27
4.4
0.90
Centre
5
5
19
8.05
3.2
0.86
General Satisfaction
2
2
9
3.33
1.2
0.72
34
34
112
63.07
15.3
0.90
Ease & convenience
Total
a = Cronbach’s a (internal consistency)
1913
1914
112
Quality of Life in Hemophilia
The Hemo-SatA consists of 34 items pertaining to six dimensions (‘ease & convenience’, ‘efficacy’, ‘burden’, ‘specialist’, ‘centre’, ‘general’). The questionnaire showed quite satisfactory psychometric characteristics in terms of reliability (Cronbach’s alpha values ranging from .68 –.90) and validity (convergent: correlation with life satisfaction scale; discriminant: differences for clinical subgroups concerning severity and target joints) (Gringeri et al., 2006).
4
Findings from Quality of Life Assessments
4.1
Comparison with the General Population
A Dutch study in hemophilia patients showed that adult hemophiliacs did not differ in their QOL from the general population (Rosendaal et al., 1990). German hemophiliacs showed significantly more limitations in the quality of life domains ‘physical activities’, ‘pain’ and ‘general health’ compared to healthy men (Szucs et al., 1996). In a Canadian study hemophilia patients reported a greater burden of morbidity than the general population (Barr et al., 2002). Shapiro et al. (2001) found that American hemophilic children with few bleeding events had higher physical functioning scores and reported less school absenteeism and were similar to the general population. In the Netherlands hemophilic children on prophylaxis were compared to their healthy peers with regard to motor performance and disability. Although 90% of the hemophilic children had no disabilities in activities of daily living (ADL) and were comparable to their healthy peers, 79% reported that the disease impacted on their lives, mainly in terms of pain and restrictions in sports (Schoenmakers et al., 2001). In a recent study it was demonstrated that adult hemophiliacs had a higher level of co-morbidity and joint damage than healthy controls (Walsh et al., 2008). Existing joint damage and presence of heart disease were found to be the strongest associates of lower physical health-related quality of life. Compared to the Canadian population, affected males had lower scores in almost all dimensions of their QOL.
4.2
Treatment
Concerning treatment regimen it was found in a European study in more than 1,000 adult hemophilia patients that patients on prophylactic treatment had a better QOL than patients on on-demand treatment (Royal et al., 2002). In an Italian study in adult hemophilia patients the opposite was found: patients on prophylactic treatment reported worse QOL compared to on-demand patients (von Mackensen et al., 2005). This contradictory finding can be explained by the deteriorated clinical situation of patients on prophylaxis in terms of high pain scores and impairments in the orthopedic score. In Italy adult patients receive only prophylactic treatment when they are clinically in a very severe condition. In the UK prophylactic treatment proved to decrease significantly the average number of bleeds in children with severe hemophilia compared to prior prophylaxis and families reported an improved health perception (Liesner et al., 1996). Comparing the effects of two prophylactic treatment regimes in children with hemophilia from Sweden and the Netherlands it was found that clinical scores and QOL were similar in both prophylactic groups (Fischer et al., 2002). Total knee arthroplasty (TKA) was found to reduce the burden of disease comparable to patients with osteoarthritis undergoing hip arthroplasty. Clinical and functional improvements after TKA lead to significantly increased quality of life and patient satisfaction (Schick et al., 1999).
Quality of Life in Hemophilia
4.3
112
Severity
In a British study hemophilia patients with severe hemophilia reported poorer levels of QOL than patients with moderate or mild hemophilia (Miners et al., 1999). Authors suggested that early primary prophylaxis might increase the QOL in these patients. Patients with severe hemophilia from France (n = 116) had acceptable values in their physical function and social relation, whereas QOL scores in the pain dimension of the SF-36 were low (Molho et al., 2000). In a Finnish study it was shown that QOL levels were associated to the clinical severity of patients (Solovieva, 2001). The burden of morbidity in mild, moderate and severe hemophiliacs was demonstrated to be associated linearly with the severity of hemophilia (Barr et al., 2002). QOL in adult Spanish hemophilia patients was negatively affected by severe orthopedic impairment related to hemophilia (Aznar et al., 2000).
4.4
Complications & Side Effects
The COCIS study in Italian hemophiliacs revealed that QOL of patients with inhibitors was similar to that of patients with severe hemophilia without inhibitors (Gringeri et al., 2003). In comparison to other diseases, physical functioning was similar to that of patients with diabetes and on dialysis, whereas mental well-being was comparable to that in the general population. HIV-positive boys with hemophilia had significantly higher rates of anxiety disorders than did the other comparison groups with a high rate of separation anxiety disorder. Anxiety disorders appear to be common in HIV-positive children with hemophilia; however, they report little intra-familial stress (Busing et al., 1993). Fathers of HIV-positive boys with hemophilia in the US stated elevated levels of parenting and psychological distress (Wiener et al., 2001). A significant interaction between HIV status and frequency of stressful life events was found among mothers of HIV-infected children and adolescents with hemophilia. By contrast, severity of hemophilia was unrelated to distress. Authors concluded that the presence of HIV infection may increase the impact of negative life events on the psychological distress experienced by mothers (Drotar et al., 1996).
4.5
General Findings
In an Italian study in adult hemophilic patients low values were found in the general health perceptions and higher scale values in social functioning (Trippoli et al., 2001). Another Italian study revealed that adult patients were mainly impaired in the dimension ‘sports & leisure’ followed by ‘future’ and ‘physical health’ of their HRQOL (von Mackensen et al., 2005). 56.7% of the adult patients had always or often to refrain from sports, moreover 75.5% from soccer (von Mackensen, 2007). In a European study HR-QOL of children aged 4–16 years was shown to be satisfactory: young children were only bothered in the dimension ‘family’ and ‘treatment’, whereas older children had higher impairments in the social dimensions, such as ‘perceived support’ and ‘friends’. The initial burden induced by prophylaxis in younger children was highly compensated by improvements in HR-QOL in older children (Gringeri et al., 2004). Canadian parents rated the QOL of their hemophilic children significantly worse compared to the self-rating of
1915
1916
112
Quality of Life in Hemophilia
their children (Sung et al., 2004). It could be shown that QOL of hemophilic children was dependent not only on clinical but also on psychosocial characteristics such as life satisfaction, social support and locus of control (Bullinger and von Mackensen, 2003).
5
Conclusions
In hemophilia, the modern management has already been successful in improving the clinical health outcomes and is recently aiming in improving the more comprehensive well-being of patients with hemophilia. HRQOL assessment has become more and more popular in the field of hemophilia research in the last decade allowing the assessment of patient’s perception of the overall effect of hemophilia care. Benefits of old and new treatment strategies or new drugs can be assessed by asking the patient about the benefits or improvements of well-being he has eventually obtained from his treatment. The quality of the overall structure of hemophilia care can benefit from knowing the patient opinion on his health, how hemophilia care is delivered in order to promote initiatives to correct and to improve it. Based on cross-cultural HRQOL assessment health care systems or hemophilia care resources can be compared among countries aiming to provide adequate care in each country and to harmonize different health care services. Finally, HRQOL assessment can be implemented in the routine assessment of patients with hemophilia, not differently from the annual check-up of joint status or viral screening. It can help to detect specific health care needs of individual patients and to verify whether or not the treatment provided to each single patient is the right one, capable to maintain or to improve his quality of life (Gringeri and von Mackensen, 2008). The here presented hemophilia-specific questionnaires have been psychometrically tested and validated, some of them cross-culturally, but still need to be implemented in clinical trials evaluating the effects of different treatment options, or in product licensing studies or gene therapy trials. Differences among the measures should be investigated as well based on clinical significant differences and minimal important differences (Wyrwich et al., 2005). Due to the multiplicity and diversity of instruments and the lack of gold standards physicians are uncertain which instrument to use. The choice should be based on the aim of the evaluation, the age group of patients, the design of the study, and the characteristics of the instruments available (study-related issues). In addition instrument-related issues (such as feasibility, psychometric characteristics, aspects covered, etc.) should be taken into account for the choice of the instrument to use (Gringeri et al., 2006).
Summary Points Haemophilia is a rare congenital bleeding disorder affecting about 1 out of 10,000 people in the general population.
Haemophilia is characterized by post-traumatic and spontaneous bleeds, mainly in joints and muscles.
Treatment is based on clotting factor replacement by intravenous infusions when bleeding occurs or prophylactically.
HR-QOL instruments can be distinguished between generic vs. disease-specific measures
Quality of Life in Hemophilia
112
Instruments for adults and for children Clinical burden of haemophilia treatment complications, presence of inhibitors, pain and arthropathy Psychological burden of haemophilia stress and coping, anxiety and depression, stigmatisation and discrimination Economical burden of haemophilia Treatment costs, costs of travelling, working days lost by the patients and by the care givers, costs for assisted physical activity
For an adequate assessment of HR-QoL validated haemophilia-specific instruments are necessary.
HR-QoL instrument for haemophilic children Haemo-QoL, CHO-KLAT, the Quality of Life Questionnaire for Young Patients HR-QoL instrument for adult haemophiliacs Medtap questionnaire (Haemo-QoL-A), Hemolatin-QoL, Haem-A-QoL, Hemofilia-QoL Treatment satisfaction questionnaire for haemophiliacs Hemo-SatA (adults) Hemo-SatP (parents) The choice of an instrument should be based on study-related and instrument-related issues.
References Aledort L, Haschmeyer R, Pettersson H. (1994). J Intern Med. 236: 391–339. Aledort LM. (2000). Haemostasis. 30(6): 333–336. Arranz P, Remor E, Quintana M, Villar A, Diaz JL, Moreno M, Monteagudo J, Ugarriza A, Soto I, Perez R, Chacon J, Garcia-Luaces M, Cid A, Balda I, Lopez MF, Gutierrez MJ, Martinez E, Marrero C, Prieto M, Sedano C, Vaca R, Altisent C, Hernandez-Navarro F, Hemofilia-QOL Group. (2004). Hemophilia. 10: 376–382. Aznar J, Magall M, Querol F, Gorina E, Tusell J. (2000). Hemophilia. 6: 170–176. Barlow JH, Stapley J, Ellard DR. (2007) Patient Educ Couns. 68(3): 235–242. Barrowcliffe TW. (2004). Hemophilia. 10(Suppl 4): 109–114. Barr RD, Saleh M, Furlong W, Horsman J, Sek J, Pai M, Walker I. (2002). Am J Hematol. 71(3): 152–160. Beeton K, Neal D, Watson T, Lee CA. (2007). Hemophilia. 13(5): 570–579. Berntorp E. (1995). Hemophilia 1(Suppl. 1): 3–7. Berntorp E, Astermark J, Bjo¨rkman S, Blanchette VS, Fischer K, Giangrande PL, Gringeri A, Ljung RC, Manco-Johnson MJ, Morfini M, Kilcoyne RF, Petrini P, Rodriguez-Merchan EC, Schramm W, Shapiro A, van den Berg HM, Hart C. (2003). Hemophilia 9(Suppl 1): 1–4.
Berntorp E, Boulyjenkov V, Brettler D, Chandy M, Jones P, Lee C, Lusher J, Mannucci P, Peak I, Rickard K, et al. (1995). Bull World Health Organ. 73(5): 691–701. Berntorp E, Shapiro A, Astermark J, Blanchette VS, Collins PW, Dimichele D, Escuriola C, Hay CR, Hoots WK, Leissinger CA, Negrier C, Oldenburg J, Peerlinck K, Reding MT, Hart C. (2006). Hemophilia. 12(Suppl 6): 1–7. Bjo¨rkman S. (2003). Hemophilia. 9(Suppl 1): 101–8. Bolton-Maggs PH. (2006). Br J Haematol. 132(6): 671–682. Bottos AM, Zanon E, Sartori MT, Girolami A. (2007). Hemophilia. 13(3): 305–310. Bowling A. (1991). ‘‘Measuring Health. A Review of Quality of Life Measurement Scales’’, Milton Keynes. Open University Press: Philadelphia. Bowling A. (2001). ‘‘Measuring disease. A review of disease-specific quality of life measurement scales’’, Open University Press, Buckingham. Brown P. (2007). Hemophilia. 13(Suppl 5): 33–40. Bullinger M. (1991). Thero Surg. 6: 143–149. Bullinger M, Mackensen S, The Haemo-QOL Group. (2003). J Pediatr Haematol Oncol. 25(Suppl. 1): 64–67. Bullinger M, von Mackensen S, Fischer K, Khair K, Petersen C, Ravens-Sieberer U, Roccino A,
1917
1918
112
Quality of Life in Hemophilia
Sagnier P, Tusell J, van den Berg M, Vicariot M. (2002). Hemophilia. 8(Suppl. 2): 47–54. Bussing R, Burket RC. (1993). J Am Acad Child Adolesc Psychiatry. 32(3): 562–567. Canclini M, Saviolo-Negrin N, Zanon E, Bertoletti R, Girolami A, Pagnan A. (2003). Hemophilia. 9(5): 619–624. Cella DF, Tulsky DS. (1993). Oncology. 29–38. Centers of Disease Control. (1982). ‘‘Update on Acquired Immune Deficiency Syndrome (AIDS) Among Patients with Hemophilia’’, MMWR Wkly. 31: 644–652. Choo KH, Gould KG, Rees DJG, Brownlee GG. (1982). Nature. 299: 178–80. Darby SC, Kan SW, Spooner RJ, Giangrande PL, Hill FG, Hay CR, Lee CA, Ludlam CA, Williams M. (2007). Blood. 110(3): 815–25. Donadel-Claeyssens S. On behalf of the European Paediatric Network for Hemophilia Management. (2006). Hemophilia. 12: 124–127. Drotar D, Agle DP, Eckl CL, Thompson PA. (1997). J Pediatr Psychol. 22(1): 1–14. Ehrlich HJ, Henzl MJ, Gomperts ED. (2002). Hemophilia. 8: 83–90. Eiser C, Morse R. (2001). Arch Dis Child. 84: 205–211. Evatt BL. (2005). Semin Thromb Hemost. 31(5): 489–494. Ewenstein BM, Valentino LA, Journeycake JM, Tarantino MD, Shapiro AD, Blanchette VS, Hoots WK, Buchanan GR, Manco-Johnson MJ, Rivard GE, Miller KL, Geraghty S, Maahs JA, Stuart R, Dunham T, Navickis RJ. (2004). Hemophilia. 10(5): 629–48. Fischer K, Astermark J, van der Bom J, Ljung R, Berntorp E, Grobbees D, van den Berg HM. (2002). Hemophilia. 8: 753–760. Flood E, Szende A, Rentz A, Gorina E, Larson P, Bullinger M, Schramm W, Wasserman J, Leidy NK. (2003). Journal of Thrombosis and Haemostasis. 1(suppl. 1): abstract number P0657. Freiburghaus C, Berntorp E, Ekman M, Gunnarsson J, Kjellberg BM, Nilsson IM. (2002). Hemophilia. 4: 16–20. Furie B, Limentani SA, Rosenfield CG. (1994). Blood. 84: 3–9. Giangrande PL. (2003). Hemophilia. 9(Suppl 1): 50–4. Giannelli F, Green PM. (1996). Bailliere’s Hematol. 9: 211–228. Gitschier J, Wood WI, Goralka TM, Wion KL, Chen EY, Eaton DH, Vehar GA, Capon DJ, Lawn RM. (1984). Nature. 312: 326–30. Gringeri A, Mannucci PM. Italian Association of Hemophilia Centres. (2005). Hemophilia. 11(6): 611–619. Gringeri A, Mantovani LG, Scalone L, Mannucci PM. (2003). Blood. 102: 2358–2363.
Gringeri A, Mantovani L, von Mackensen S. (2006). Hemophilia: 12(Suppl. 3): 22–29. Gringeri A, von Mackensen S. (2008). Hemophilia. 14(Suppl 3): 19–25. Gringeri A, von Mackensen S, Auerswald G, Bullinger M, Perez Garrido R, Kellermann E, Khair K, Lenk H, Vicariot M, Villar A, Wermes C. (2004). Hemophilia. 1, 10(Suppl. 1): 26–33. Hartl HK, Reitter S, Eidher U, Ramschak H, Ay C, Pabinger I. (2008). Hemophilia. 1: 1–6. Hay CR, Brown S, Collins PW, Keeling DM, Liesner R. (2006). Br J Haematol. 133(6): 591–605. Hay C, Recht M, Carcao M, Reipert B. (2006). Semin Thromb Hemost. 32(Suppl 2): 15–21. Hilgartner M, Knatterud GL, The FEIBA Study Group. (1983). Blood. 61: 36–40. Holahan CJ, Moos RH. (1987). J Abnorm Psychol. 96: 3–13. Jadhav M, Warrier I. (2000). Semin Thromb Hemost. 26(2): 205–8. Jones QJ, Garsia RJ, Wu RT, Job RF, Dunn SM. (1995). J Psychosom Res. 39(5): 597–608. Karimi M, Peyvandi F, Siboni S, Ardeshirim R, Gringeri A, Mannucci PM. (2004). Hemophilia. 10(4): 367–369. Key NS, Negrier C. (2007). Lancet. 370(9585): 439–48. Klinge J, Auberger K, Auerswald G, Brackmann HH, Mauz-Ko¨rholz C, Kreuz W. (1999). Eur J Pediatr. 158: 162–5. Komvilaisak P, Connolly B, Naqvi A, Blanchette V. (2006). Hemophilia. 12(Suppl 6): 87–93. Kraus EM, Brettler DB. (1988). Am J Med Genet. 31(2): 259–267. Lacey L. (2002). Pathophysiol Haemost Thromb 32: (suppl 1):29–32. Liesner RJ, Khair K, Hann IM. (1996). Br J Haematol. 92: 973–978. Ljung R. (2007). Br J Haematol. 138(5): 580–586. Ljung R, Petrini P, Nilsson IM. (1990). Acta Paediatr Scand. 79: 196–200. Lloyd Jones M, Wight J, Paisley S, Knight C. (2003). Hemophilia. 9: 464–520. Lusher J, Ingerslev J, Roberts H, Hedner U. (1998). Blood Coagul Fibrinolysis. 9: 119–28. Magli-Barioz D, Ounnoughe`ne N, Paugy P, Quarre` MC, Vassilief D, Lopez I, Goudemand J, Negrier C, Assogba U, Driss F, Dumont MA, Molho P. (2004). Hemophilia. 10(5): 572–581. Manco-Johnson M, Morrissey-Harding G, EdelmanLewis B, Oster G, Larson P. (2004). Hemophilia. 10: 34–41. Mann KG. (1999). Thromb Haemost. 278: 9370–7. Mannucci PM. (2003). J Thromb Haemos. 1: 1349–1355. Miller R, Sabin C, Goldman E, Clemente C, Sadowski H, Lee CA. (2000). Psychol Health Med. 5: 3–12. Miners AH, Sabin CA, Tolley KH, Jenkinson C, Kind P, Lee CA. (1999). Hemophilia. 5(6): 378–85.
Quality of Life in Hemophilia Molho P, Rolland N, Lebrun T, Dirat G, Courpied J, Croughs T, Duprat I, Sultan Y. (2000). Hemophilia. 6: 23–32. Monat A, Lazarus RS. (1977). ‘‘Stress and Coping: An Anthology’’, Columbia University Press, New York, pp.6. Morfini M. (2003). Hemophilia. 9(Suppl 1): 94–9. Negrier C, Goudemand J, Sultan Y, Bertrand M, Rothschild C, Lauroua P.. The members of the French FEIBA study group. (1997). Thromb Haemost. 77: 1113–1119. Nilsson IM, Berntorp E, Lo¨fqvist T, Pettersson H. (1992). J Intern Med 232(1): 25–32. Pollak E, Mu¨hlan H, von Mackensen S, Bullinger M. the Haemo-QOL Group. (2006). Hemophilia. 12: 384–392. ‘‘Prevention and Control of Hemophilia: Memorandum from a Joint WHO/WFH Meeting’’. (1991). Bull World Health Organ. LXIX. 1: 17–26. Prowse C, Ludlam CA, Yap PL. (1997). Vox Sang. 72(1): 1–10. Remor E. (2005). Revista Interamericana de Psicologı´a/ Interamerican J Psychol. 39(2), 211–220. Remor E, Arranz P, Quintana M, Villar A, Jimenez-Yuste V, Diaz JL, Del Rincon C, Marrero C, Moreno M, Lucia JF, Martinez E, Soto I, Sedano C, GonzalezBoullosa R, Prieto M, Garcia-Luaces M, HernandezNavarro F, in representation of the Hemofilia-QOL Project Group. (2005). Hemophilia. 11: 603–610. Remor E, Young NL, von Mackensen S, Lopatina EG. (2004). Hemophilia. 10(Suppl. 4): 30–34. Rentz A, Flood E, Altisent C, Bullinger M, Klamroth R, Garrido RP, Scharrer I, Schramm W, Gorina E. Cross-cultural development and psychometric evaluation of a patient-reported health-related quality of life questionnaire for adults with haemophilia. Haemophilia 2008; 14(5): 1023–1034. Report of a Joint WHO/WFH/ISTH Meeting. (2002). ‘‘Delivery of Treatment for Hemophilia’’, World Health Organization. London, UK. Rizza CR. (1984). Scand J Haematol. 40(Suppl.33): 187–193. Roberts HR, Monroe DM. 3rd, Hoffman M. (2004). Semin Hematol. 41(suppl 1): 101–108. Rosendaal FR, Smit C, Varekamp I, Bro¨cker-Vriends AH, van Dijck H, Suurmeijer TP, Vandenbroucke JP, Brie¨t E. (1990). J Intern Med. 228: 663–640. Rosendaal G, van der Berg HM, Lafeber FPJG, Bijlsma JWJ. (2000). ‘‘Blood-induced joint damage: an overview of musculoskeletal research in hemophilia’’, In: Rodriguez-Merchan EC, Goddard NJ, Lee CA (eds.) ‘‘Muskuloskeletal Aspects of Hemophilia’’, Blackwell Science Ltd, Oxford, UK, pp. 18–26. Rotter JB. (1954). ‘‘Social Learning and Clinical Psychology’’. New York Prentice-Hall.
112
Royal S, Schramm W, Berntorp E, Giangrande P, Gringeri, A, Ludlam C, Kroner B, Szucs T, for the European hemophilia economics study group. (2002). Hemophilia. 8: 44–50. Santagostino E, Gringeri A, Berardinelli L, Beretta C, Muc¸a-Perja M, Mannucci PM. (2003). Br J Haematol. 123(3): 502–506. Santagostino E, Gringeri A, Mannucci PM. (1999). Br J Haematol. 104: 22–26. Saviolo-Negrin N, Cristante F, Zanon E, Canclini M, Stocco D, Girolami A. (1999). Hemophilia 5(1): 63–68. Schick M, Stucki G, Rodriguez M, Meili E, Huber E, Michel B, Bruhlmann P. (1999). Clin-Rheumatol. 18: 468–472. Schoenmakers M, Gulmans V, Helders P, van den Berg HM. (2001). Hemophilia. 7: 293–298. Schramm W, Royal S, Kroner B, Berntorp E, Giangrande P, Ludlam C, Gringeri A, Berger K, Szucs T. For the European hemophilia economic study group. (2002). ‘‘Clinical outcomes and resource utilization associated with hemophilia care in Europe’’, Hemophilia. 8: 33–43. Shapiro A, Donfield S, Lynn H, Cool V, Stehbens J, Hunsberger S, Tonetta S, Gomperts E.D. Academic Achievement in Children with Hemophilia Study Group. (2001). Pediatrics. 108: 6:e105. Solovieva S. (2001). Hemophilia. 7: 53–63. Spielberger CD. (1972). ‘‘Anxiety: Current trends in theory and research’’, Academic Press. New York, N.Y. Sung L, Young NL, Greenberg ML, McLimont M, Samanta T, Wong J, Rubenstein J, Ingber S, Doyle JJ, Feldman BM. (2004). J Clin Epidemiol. 57(11): 1161–1166. Szucs TD, Offner A, Schramm W. (1996). Hemophilia. 2: 211–217. Tedga˚rd U, Ljung R, McNeil TF. (1999). Br J Haematol. 106(2): 421–426. Teitel J, Berntorp E, Collins P, D’Oiron R, Ewenstein B, Gomperts E, Goudemand J, Gringeri A, Key N, Leissinger C, Monahan P,Young G. (2007). Hemophilia. 13(3): 256–63. The WHOQOL Group. Study protocol for the World Health Organization. Project to develop a Quality of Life assessment instrument (WHOQOL) (1993). Qual Life Res. 2: 153–159. Triemstra AH, Van der Ploeg HM, Smit C, Brie¨t E, Ade`r HJ, Rosendaal FR. (1998). Soc Sci Med. 47(5): 581–593. Trippoli S, Vaiani M, Linari S, Longo G, Morfini M, Messori A. (2001). Haematologica. 86: 722–728. Troisi CL, Hollinger FB, Hoots WK, Contant C, Gill J, Ragni M, Parmley R, Sexauer C, Gomperts E, Buchanan G. (1993). Blood. 81: 412–418. Vermilen J, Peerlink K. (1994). Vox Sang. 84 (Suppl 4): 8–11.
1919
1920
112
Quality of Life in Hemophilia
von Mackensen S. (2007). Hemophilia. 13(2): 38–43. von Mackensen S, Bullinger M, The Haemo-QOL Group. (2004). Hemophilia. 10(Suppl. 1): 17–25. von Mackensen SV., Gringeri A. (2005). ‘‘Quality of Life in Hemophilia’’. In: Lee CA, Berntorp EE, Keith Hoots W. (eds.) Textbook of Hemophilia. Blackwell Publishing Ltd. Oxford, UK, pp. 345–350. von Mackensen S, Gringeri A, Mantovani L, Skovlund SE, Mannucci PM. (2002). Blood. 100(11) (Abstract). von Mackensen S, Gringeri A, Mantovani L, Skovlund S, Santoni L, Tagliaferro A, Tagariello G, Ciavarella N, Iorio A. (2004). Hemophilia. 10(Suppl. 3): 126 (Abstract). von Mackensen S, Gringeri A, Ravera S, (2005). the Haem-A-QOL Group. Haematologica. 90(Suppl. 2): 116–117 [Abstract 0290]. von Mackensen S, Gringeri A, Santoni L, Tagliaferro A, Tagariello G, Ciavarella N, Iorio A, Molinari C. (2004). Blood. 104(11): 2214 [Abstract]. von Mackensen S, Gringeri A, Scalone L, Mantovani L, the COCHE Study Group. (2005). Haematologica. 90(Suppl 2): 310. [Abstract 0778]. von Mackensen S, Scalone L, Ravera S, Mantovani L, Gringeri A. For the COCHE Study Group. (2007). ‘‘Quality of Life Assessment in Adult Patients with Hemophilia – Results of the COCHE Study’’, Oral presentation at the annual congress of the Society
of Thrombosis and Haemostasis Research. Dresden, 21–25 February 2007, FV19. von Mackensen S, Scalone L, Ravera S, Mantovani L, Gringeri A, for the COCHE Study Group. (2005). Value Health 8(6): A127 [Abstract PHM5]. von Mackensen S, Siboni SM, Gringeri A, Tradati F, Franchini M, Tagliaferro A, Mannucci PM. (2008). Hemophilia. 14(s2): 127 [Abstract 22 S 01]. Walsh M, Macgregor D, Stuckless S, Barrett B, Kawaja M, Scully MF. (2008). J Thromb Haemost 6(5): 755–761. Weaver M, Patrick DL, Markson LE, Martin D, Frederic I, Berger M. (1997). A J Manag Care. 3: 579–594. White GC 2nd, Rosendaal F, Aledort LM, Lusher JM, Rothschild C, Ingerslev J. (2001). Thromb Haemost 85: 560. Wiener LS, Vasquez MJ, Battles HB. (2001).J Pediatr Psychol. 26(6): 353–358. Wight J, Paisley S. (2003). Hemophilia 9(4): 418–35. Wyrwich KW, Bullinger M, Aaronson N, Hays RD, Patrick DL, Symonds T. (2005). Qual Life Res 14(2): 285–295. Yi M, Kim WO, Kang HS. (2003) Taehan Kanho Hakhoe Chi. 33:(7)875–884. Young NL, Bradley CS, Wakefield CD, Barnard D, Blanchette VS, McCusker PJ. (2006). Pediatr Blood Cancer. 47: 305–311.
113 Quality of Life in Amenorrhea and Oligomenorrhea William W. K. To 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1922
2 2.1 2.2 2.3
Primary Amenorrhea and Quality of Life . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1923 Anatomical Abnormalities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1923 Chromosomal Abnormalities – Turner Syndrome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1924 Other Chromosomal Abnormalities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1927
3 3.1 3.2 3.3
Oligomenorrhea and Secondary Amenorrhea and the Quality of Life . . . . . . . . . . 1928 Premature Ovarian Failure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1929 Polycystic Ovarian Syndrome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1931 Hypothalamic Menstrual Dysfunction and Associated Disorders . . . . . . . . . . . . . . . . . . 1933
4
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1935 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1935
#
Springer Science+Business Media LLC 2010 (USA)
1922
113
Quality of Life in Amenorrhea and Oligomenorrhea
Abstract: > Oligomenorrhea and > amenorrhea are common forms of menstrual dysfunction that occurs in reproductive age group women. The clinical implications of such menstrual disturbance will depend on the specific diagnosis and on the presence or absence of associated problems, such as infertility or other hormonal or metabolic disturbances. Little attention has been paid in the past to the psychological impact of these symptoms and their effect on the women’s perceived > quality of life (QOL). Current available literature on the QOL associated with amenorrhea and oligomenorrhea has largely focused on specific conditions such as > Turner syndrome, > premature ovarian failure, or > polycystic ovarian syndrome, while many other common and uncommon conditions have not yet been evaluated in detail. In addition, disease specific questionnaires have only been validated for conditions such as polycystic ovarian disease and eating disorders, but not for menstrual dysfunction in general, so that most studies have relied on generic survey tools such as the SF-36. The current available literature showed that oligomenorrhea and amenorrhea resulting from various conditions does have demonstrable negative effects on the perceived QOL of affected women. Such negative effects are most significant in women with associated infertility, metabolic disturbances such as obesity and hyperandrogenism and other medical problems. In women with relatively mild degree of menstrual disturbance with oligomenorrhea, the impact on quality of life largely depends on their interpretation and acceptance of the symptoms. List of Abbreviations: FSH, follicular stimulating hormone; GnRH, gonadotrophin releasing hormone; LH, luteinizing hormone; PCOS, polycystic ovarian syndrome; POF, premature ovarian failure; QOL, quality of life; SF-36, short form 36; TS, Turner syndrome
1
Introduction
Oligomenorrhea and amenorrhea are usually symptoms of ovarian and reproductive dysfunction. Regular monthly menstruation in females has only been observed as the norm in modern society after the introduction of effective contraception to control the number and the timing of pregnancies. The length, regularity and frequency of normal menstrual cycles have been studied in large scale population studies (Harlow and Ephross, 1995). Mean menstrual cycle length between the young reproductive age of 20 and 34 years varies from 28 to 30.7 days (range 19.7–43.5 days, 5th to 95th centiles). Regular monthly cycles represent cyclical ovarian activity, which in turn is dependent on the normal physiological hypothalamic pituitary axis, the uterus and endometrium as end organs, and a patent lower genital tract. Amenorrhea is defined as the complete absence or cessation of menstrual periods for greater than 6 months that is not due to pregnancy. > Primary amenorrhea is defined as no spontaneous onset of menstruation by the age of 16 years, while > secondary amenorrhea is defined as the absence of menstruation for 6 months or longer if the patient has previously experienced regular menses or for 12 months or more if the patient has oligomenorrhea (Critchley, 2003). Oligomenorrhea is the reduction in frequency of menstruation where menstrual intervals may vary between 6 weeks and 6 months. However, regular but long cycles of up to 6 weeks are often physiologically indistinguishable from normal shorter 28 3-day cycles in terms of follicular formation, ovulation and hormonal profile. On the other hand, long irregular cycles often signify chronic anovulation, absence of mid cycle gonadotrophin surges and prolonged hypo-estrogenism. Thus, the definition of oligomenorrhea as such includes a wide spectrum of conditions ranging from virtual normality to true amenorrhea and all its implications.
Quality of Life in Amenorrhea and Oligomenorrhea
113
The World Health Organization defines quality of life (QOL) as ‘‘an individual’s perception of their position in life in the context of the culture and value system in which they live and in relation to their goals, expectations, standards and concerns’’ (Orley et al., 1998). Thus, it can be seen that OQL is influenced by a broad range of associated domains, interlinking physical health with social and psychological functioning. QOL assessments can be either generic or disease specific. Studies of QOL in menstrual dysfunction are not yet comprehensive, with available data in certain specific areas but not others. The tools used for assessment are generally generic surveys, the commonest being the Short-Form 36 developed by Ware (Ware and Sherbourne, 1992). The SF-36 is a generic quality of life questionnaire that could be applied to different health conditions and symptomatologies and has been used commonly for assessing treatment outcomes. Despite its widespread use in many specialties, the use of this questionnaire to assess menstrual dysfunction was not been described before (Manocchia et al., 1998). It thus remains arguable whether there is a lack of sensitivity of the SF-36 with reference to menstrual symptoms, as the broad approach of the SF-36 may not apply to specific quality of life issues encountered by women with menstrual dysfunction. The use of disease specific questionnaires should theoretically be more sensitive to detect perceived quality of life as related to menstrual problems, but such tools have only been developed for specific conditions such as polycystic ovarian diseases (Coffey and Mason, 2003; Guyatt et al., 2004) or those with menstrual problems secondary to inherited bleeding disorders (Kadir et al., 1998). Thus, a menstrual dysfunction specific questionnaire to study QOL has yet to be developed and validated.
2
Primary Amenorrhea and Quality of Life
Amenorrhea can be physiological at certain periods of the female life cycle, including prepuberty, during pregnancy, lactation, and in the postmenopausal period. The incidence of nonphysiological amenorrhea has been estimated to be around 2–5% in reproductive age group women, with secondary amenorrhea being much more common than primary amenorrhea. In clinical practice, investigations for primary amenorrhea will usually commence by around 14 years of age particularly with evidence of delayed development of secondary sexual characteristics. In addition, girls with no menarche within 4 years of the onset of adrenarche and thelarche should also be investigated. The causes of primary amenorrhea are heterogeneous (> Table 113-1), and include simple anatomical causes to chromosomal abnormalities to endocrine causes involving the hypothalamic pituitary axis. In terms of the impact of such conditions on QOL, it is obvious that the influence would depend on various factors such as the presence or absence of external dysmorphology, any impairment on the development of secondary sexual characteristics, and the implications on sexual and reproductive/ fertility function.
2.1
Anatomical Abnormalities
In cases of simple anatomical abnormalities, such as imperforate hymen or cryptomenorrhea, simple surgical drainage of the collected hematocolpos and haematometra (collection of blood in the vagina and uterine cavity) and incision of the vaginal membrane will enable the patient to return to normal menstruation in subsequent months with no long term implications on
1923
1924
113
Quality of Life in Amenorrhea and Oligomenorrhea
. Table 113-1 Causes of primary amenorrhea – classification according to etiology Causes
Conditions
Delayed puberty
Familial or constitutional Nutritional
Anatomical congenital abnormalities
Imperforate hymen (crytomenorrhea)
Karyotype abnormalities
Turner syndrome (45 XO) and mosaics
Mullerian agenesis (absent uterus/ absent vagina) XO/ XY mosaics Testicular feminization syndrome (46 XY)
Hypothalamic pituitary ovarian axis causes
Hyperprolactinemia Premature ovarian failure Hypogonadotrophic hypogonadism Polycystic ovarian syndrome Other endocrine disorders including hypothyroidism, adrenal disorders or hormone secreting tumors
sexual and fertility function. It can thus be expected that the QOL in these affected adolescents should not differ than that of normal women. On the other hand, in more complicated anatomical anomalies of the genital tract, such as congenital absence of the uterus, upper vagina or any combinations of these, despite the absence of hormonal dysfunction, the QOL of affected women would be highly dependent on the implications of the abnormality in their reproductive capacity. This group of abnormalities is commonly known as the MayerRokitansky-Kuster-Huaser syndrome or Mullerian agenesis as the structures are derived from the embryological Mullerian ducts. As ovarian endocrine function is normal, development of secondary sexual characteristics appears at the normal pace. The diagnosis is often made at their presentation with primary amenorrhea when no functioning uterus can be found on imaging investigations (absent uterus), or when obstruction to menstrual flow as a result to agenesis of a segment of the upper vagina and/or the cervix (vaginal atresia) is detected. As surgical correction to produce a functioning cervix and uterus for reproduction is often incomplete or impossible in these cases, the future fertility potential of these women is compromised. The negative impact on the QOL of these women will thus be expected to come basically from their subfertililty or infertility.
2.2
Chromosomal Abnormalities – Turner Syndrome
Another common group of causes for primary amenorrhea is chromosomal abnormalities, of which Turner syndrome and its mosaic forms are the most common. Turner syndrome is a genetic condition which results from a total or partial absence of one of the X chromosomes. It has an incidence of around 1 in 2,500–3,000 live female births (Sybert and McCauley, 2004). The most prominent clinical features of Turner syndrome are short stature and gonadal dysgenesis which leads to primary or premature ovarian failure, absence of, or delayed or incomplete pubertal development, and infertility. A myriad of dysmorphic features and
Quality of Life in Amenorrhea and Oligomenorrhea
113
congenital abnormalities can also be observed amongst patients with Turner syndrome. They are at risk of congenital heart defects such as coartation of the aorta, bicuspid aortic valves, and may have progressive aortic root dilatation or dissection. They are also at risk of congenital lymphoedema, renal malformations, sensorineural hearing loss, osteoporosis, obesity, diabetes mellitus, and atherogenic lipid profile. Patients usually have normal intelligence but may have problems with nonverbal, social and psychomotor skills. Physical manifestations may include misshapen ears, a webbed neck, a broad chest and widely spaced nipples, and cubitus valgus (> Table 113-2). Turner syndrome should be suspected in girls presenting with primary amenorrhea and short stature (Morgan, 2007). Treatment for girls with Turner syndrome mainly aims at increasing their height and promoting sexual maturation. They would thus be treated with growth hormone therapy for short stature from early childhood, and supplementary estrogen is initiated by adolescence for pubertal development and for prevention of osteoporosis. Life-long follow-up for Turner syndrome patients are generally recommended for management of commonly occurring conditions such as thyroid diseases, hypertension and diabetes mellitus (Conway, 2002). Several studies to date have been performed to evaluate the effect of treatment on the QOL in Turner patients. These studies mainly focused on the impact of growth hormone therapy and puberty induction amongst these patients (Hull and Harvey, 2003). The health related quality of life of a large sample of young French women with Turner syndrome was evaluated after growth hormone treatment using the French version of the Short Form 36 questionnaire (Carel et al., 2005). The mean age of this cohort was young at only 22.6 years. Their overall SF-36 scores were similar to that of French women in the general population. Nevertheless, amongst the different characteristics of these Turner patients, cardiac involvement, otological
. Table 113-2 Turner syndrome – possible clinical features in Turner syndrome Short stature Epicanthus, ptosis strabismus Low set ears, malformed external ears, recurrent otitis media, hearing loss Alopecia, low posterior hairline High arched palate, micrognathia Webbed neck Broad chest, no breast development, wide spaced nipples Scoliosis, cubitus valgus, genu valgus Coartation of aorta, deformed aortic valves Horseshoe kidneys, double ureters/ double renal pelvis Hypertension Lymphoedema of hands and feet Dysplastic nails Multiple pigmented naevi Obesity and diabetes mellitus Streak gonads
1925
1926
113
Quality of Life in Amenorrhea and Oligomenorrhea
involvement and puberty induction after the age of 15 were associated with significantly lower general health perception scores in at least one of the SF-36 domains. On the other hand, factors such as their final height, which would be related to the direct effects of the growth hormone therapy, did not affect the perceived QOL score. The authors concluded that QOL is normal and unaffected by height in young adults with Turner syndrome treated with growth hormone. They also pointed out the need to attend to the general health and otological care rather than their stature alone. In another smaller Dutch cohort of Turner patients (Bannink et al., 2006) with a mean age of 19.6 years who reached normal height after receiving growth hormone therapy and had age-appropriate pubertal development, normal overall health related quality of life was reported. Again using SF-36 as a measure, this cohort of women with Turner syndrome had even higher related QOL in some domains, including bodily pain, social functioning and emotional role. The authors suggested that such results might be explained by the estrogen effects of the therapy they received, or by a possible response shift from their basic expectations, indicating a different internal reference in these women with Turner syndrome. In addition, these Dutch patients who were satisfied with their final height had a significantly better health related QOL on the physical performance scales. The height gain had a significant positive effect on the outcome of their physical role domain. The absolute height gain significantly influenced the domain of vitality. Satisfaction with breast development apparently also had a positive impact on several of these QOL domains. Moreover, the social economical status significantly and positively influenced the scores in the depressive domain. On the contrary, actual karyotype diagnosis and the number of medical problems in the past had no influence on the domains. However, in another recent study in which self-esteem, depressive symptoms and anxiety symptoms in girls with Turner syndrome were compared to those in girls with familial short stature and healthy controls using he Children’s Depression Inventory, State-Trait Anxiety Inventory for Children, and Piers-Harris Children’s Self Concept Scale, girls with Turner syndrome had lower self-esteem and higher state anxiety levels than both girls with familial short stature and normal controls (Kilic et al., 2005). Thus, it can be seen that even for young Turner syndrome women, they are at risk of psychological problems compared to the general population independent of their stature. It is obvious that Turner patients were less likely to be able to get married, and this issue becomes much more apparent when cohorts were surveyed in later life beyond their adolescence. In one Scandinavian study (Sylven et al., 1991), among a population of 49 women over the age of 37 years, only 31 (63%) were ever married. In a Japanese cohort, only 4/20 (20%) (Okada, 1994) of Turner women with a mean age of 25.7 were married. In another French cohort, only 17/105 (16%) of the Turner women between 18 and 53 years old in their survey were married. In addition, 58% of the women in this cohort did not have any sexual life whatsoever. Similarly, another cohort had shown that up to 45% of Turner patients never had sexual intercourse (Pavlidis et al., 1995). It is most probable that the strong cultural emphasis of fertility in many communities have affected the decision and the opportunities of these Turner syndrome women to get married once they realized their unlikely chance of fertility during the counseling they received about their diagnosis. This low frequency of marriage would also directly bear an impact on their QOL. The limited opportunities for sexual contact would be expected to further negatively influence their social and psychological well being. A French cohort has reported low esteem in Turner patients was related to otological involvement and limited sexual experience. Low social adjustment was associated with lower paternal socioeconomic class and an absence of sexual experience. Late age at first date was
Quality of Life in Amenorrhea and Oligomenorrhea
113
associated with cardiac involvement and a lack of pubertal development. Age at first intercourse was related to age at puberty and paternal social class, and delayed induction of puberty had a long lasting effect on sex life (Carel et al., 2006). In addition, mood and psychological problems were also found to be more common in these women. In a study of 100 women with Turner syndrome, assessment using the Structured Clinical Interview for DSM IV after a 2-week period during which their hormone replacement had been discontinued found that 52% of the them met criteria for a current or a past depressive or anxiety disorder. Eighteen percent met criteria for a current Axis I psychiatric disorder (major or minor depression, dysthymia, anxiety) and 46% met criteria for a past Axis I psychiatric illness (unipolar and bipolar depression, anxiety, eating disorders and substance dependence). Another 5% met criteria for an Axis II personality disorder. It was concluded that women with Turner syndrome reported a higher rate of lifetime depression compared with rates observed in communitybased studies but similar to those obtained from gynecologic clinic samples (Cardoso et al., 2004). It is apparent that women with Turner syndrome were at risk of psychological problems, which became more prominent in later life when they failed to establish a marriage or family. In many of these studies on Turner syndrome, these women reported significantly higher levels of shyness and social anxiety and reduced self-esteem compared to normally menstruating women, but the incidence was still marginally higher than karyotypically normal women with premature ovarian failure, suggesting that the experience of ovarian failure and infertility since puberty significantly added to the negative impact on their QOL. Indeed, women with this diagnosis reported in open-ended interviews that dealing with the infertility and ovarian failure was the most difficult part of having Turner syndrome (Sutton et al., 2005). In summary, it can be seen that women with Turner syndrome suffer from the dysmorphology that can be associated with the condition, the most obvious being short stature. They also suffer from other subtle psychosocial anomalies, and the usually poorly formed secondary sexual characteristics from the lack of pubertal development. Due to their infertility, they face long term social problems from being unable to easily form long term marital relationships. All these factors would be conducive to the negative psychosocial and quality of life performance in women with Turner syndrome. Therefore, in addition to medical treatment and monitoring, women with Turner syndrome should also be supported psychologically by social, educational and psychotherapeutic interventions which aim to address their self-esteem and emotional difficulties.
2.3
Other Chromosomal Abnormalities
Compared to Turner syndrome, testicular feminization syndrome is a rarer condition, in which the affected women will actually be genetically a male with 46 XY karyotype, but that they are phenotypically females because of androgen insensitivity in heir bodies due to the lack of the critical enzyme system. These women usually have excellent female secondary sexual characteristics development and the major psychosocial obstacle they will have to face will be that of infertility due to the absence of an anatomical uterus, similar to women with Mullerian agenesis. They often present late at puberty because of primary amenorrhea, and the diagnosis is suspected when a complete absence of the uterus is found. Additional identity confusion may also be introduced if in their counseling, they have been told, perhaps mistakenly, that they are actually males ‘‘genetically.’’ No established data on the QOL of these affected women are available in the literature currently.
1927
1928
113
Quality of Life in Amenorrhea and Oligomenorrhea
Many other heterogeneous causes of primary amenorrhea may also present as secondary amenorrhea. For instance, those with premature ovarian failure could experience a number of years of relatively normal menstruation before they have a cessation of menses. It can thus be realized that apart from anatomical causes that precludes any normal menstruation at all, the causes of primary and secondary amenorrhea will overlap substantially.
3
Oligomenorrhea and Secondary Amenorrhea and the Quality of Life
Menstrual dysfunction commonly occurs in young females in their late adolescent period and early twenties. Irregular periods with variable cycle lengths appear to be the most commonly encountered pattern, with irregular ovulation and variable amount of flow depending on whether ovulation has actually occurred in the particular cycle (WHO, 1986). Oligomenorrhea and amenorrhea is thus one of the most frequently encountered forms of dysfunctional menstrual pattern in this age group. One can readily realize that oligomenorrhea and secondary amenorrhea are different degrees of clinical manifestation of the same spectrum of diseases sharing the same etiological factors (> Table 113-3), and indeed the clinical picture may shift from one to the other over time. When faced with the clinical manifestation of secondary amenorrhea or oligomenorrhea, the concerns or worries of affected women about their menstrual problems may be different, depending on their knowledge of menstrual physiology, their interpretation of the probable etiology of the menstrual dysfunction, as well as their personal beliefs, acceptance, and cultural background. Additional psychological burden may also come from problems associated with the menstrual dysfunction, such as infertility. The most common pathophysiology underlying such menstrual dysfunction includes premature ovarian failure
. Table 113-3 Causes of oligomenorrhea and secondary amenorrhea – classification according to etiology Causes Hypothalamic pituitary axis
Conditions Kallman’s syndrome Intense physical exercises Anorexia nervosa and weight loss Hyperprolactinemia Ablation by irradiation, surgery, compression by tumors or vascular infarction (Sheehan’s syndrome)
Ovarian causes
Chromosomal abnormalities (such as mosaic Turner syndrome) Ablation by surgery, irradiation or chemotherapy Premature ovarian failure (including autoimmune causes) Polycystic ovarian syndrome Resistant ovary syndrome
Uterine causes
Ablation by surgery (hysterectomy or endometrial ablation) or pelvic irradiation (destroying functional endometrium) Ashermann syndrome and cervical stenosis Infections (such as tuberculosis endometritis)
Quality of Life in Amenorrhea and Oligomenorrhea
113
. Table 113-4 Most common etiology of secondary amenorrhea (modified from Balen et al., 1993) Causes
Percentage
Polycystic ovarian syndrome
36.9%
Premature ovarian failure
23.6%
Hyperprolactinemia
16.9%
Weight-related amenorrhea
9.8%
Hypogonadotrophic hypogonadism
5.9%
Hypopituitarism
4.4%
Exercise-related amenorrhea
2.5%
. Table 113-5 Causes of premature ovarian failure Causes
Conditions
Genetic
Familial
Gonadal dysgenesis
– Turner syndrome with deletion of X chromosome and other mosaic patterns – Pure gonadal dysgenesis with normal XX complement
Autoimmune diseases
e.g., systemic lupus
General metabolic disorders
e.g., galactosemia, 17-hydroxylase deficiency, Cooley’s anemia with iron overloading
Iatrogenic causes
Surgical ablation, chemotherapy, irradiation
Infections
Mumps, pelvic tuberculosis
Idiopathic
and polycystic ovarian syndrome, followed by hypothalamic pituitary axis dysfunction (> Table 113-4). The impact of these conditions on quality of life evaluations will be reviewed.
3.1
Premature Ovarian Failure
Premature ovarian failure (POF) is a disorder characterized by the cessation of ovary functioning and elevated gonadotrophin levels before the age of 40 years. The term premature ovarian failure or premature menopause is often used to refer to women who exhibit hypergonadotrophic secondary amenorrhea before the age of 40 (> Table 113-5). POF has significant repercussions leading to physical, psychosexual and social impairment in affected women (de Taraciuk et al., 2008). The incidence of POF ranges from 1 to 3% of the female population (Coulam et al., 1986) being around 1 in 100 at the age of 40 and 1 in 1,000 at the age of 30. The prevalence of POF in patients with primary amenorrhea is between 10 and 28%, and between 4 and 18% in patients with secondary amenorrhea. The pathophysiology of POF is explained by follicle depletion in the ovaries, which could begin as early as intrauterine life. The process of follicular depletion could be accelerated via idiopathic mechanisms, or as a result of genetic, immunological or inflammatory processes, as well as other exogenous toxins such as chemotherapy or radiation.
1929
1930
113
Quality of Life in Amenorrhea and Oligomenorrhea
When the psychological well-being of women with POF was compared with the general population, it was shown that these patients reported higher levels of depression and perceived stress, and low levels of self-esteem and life satisfaction (Liao et al., 2000). Self reported scores on dimensions in sexuality were significantly more negative than controls. The factors identified that could affect the degree of reported distress in women with POF include the age, age at diagnosis, time since diagnosis, already having children, being in a long term relationship, or having psychological treatment in the past or present. As POF could pose significant psychological morbidity for a substantial proportion of those with the condition, it was concluded that the provision of psychological support should be an integral part of the clinical management of POF. Similarly, a qualitative study that evaluated the psychiatric effects of idiopathic premature menopause showed that in many women, such an event was interpreted as a major disruption in their lives that confronted them with a multitude of issues related to the timing of the event and their embodied understanding of menopause (Boughton, 2002). Another psychological assessment of patients diagnosed with POF did not show high levels of depression, but did show high values of anxiety. Psychological stress was higher throughout the year before they became amenorrheic, than during the year before the psychological evaluation. Personality profiles identified as prominent in POF women include actively modifying, self-indulging, internally focused, realistic/sensing, feeling guided, dominant/controlling and dissatisfied/complaining. The study also identified difficulty for women to reach female identification and a conflictive bond with the mother figure (de Taraciuk et al., 2008). When the emotional response to learning the diagnosis of POF was also studied, it was shown that a high proportion of affected women were unsatisfied with the manner in which they were counseled of the diagnosis by their doctor, and reported moderate to severe emotional distress at the time of disclosure of the diagnosis. The authors concluded that learning the diagnosis of POF could be emotionally traumatic and difficult, and that the way the disclosure of the diagnosis was made could significantly impact on their level of distress (Groff et al., 2005). Several studies have evaluated the psychological impact of POF after hysterectomy. Hysterectomy for various indications has been shown to increase the risks of subsequent POF due to impairment of the normal blood supply to the ovaries after the surgical procedure, even if the ovaries are anatomically preserved during the operation. In one study, depressed mood before hysterectomy due to prolonged menstrual periods, chronic pelvic pain and severe premenstrual tension that warranted the surgical treatment was relieved as the cure for physical symptoms after hysterectomy resulted in an improvement of mood after the procedure (Khastgir and Studd, 1999). Nevertheless, the authors stated that in women with a preexisting psychiatric illness or previous personality problems, depressed mood might still persist or occur after the surgery with the stress of hysterectomy. In addition, estrogen deficiency following hysterectomy (with and without bilateral oophorectomy) may also be responsible for negative mood changes, so that regular hormonal monitoring should be practiced after hysterectomy to detect the need for hormonal replacement and to reduce the incidence of post-hysterectomy depression. In another study, a comparison of women with POF after hysterectomy and those who did not have the operation showed that the hysterectomy group had more negative symptoms. Logically, it was observed that those with their ovaries removed had more symptoms. However, the authors pointed out that the rate of ovarian failure symptoms and psychiatric symptoms could be reduced by adequate information before and after performing the hysterectomy (Habelt et al., 1996). Finally, when the relationship between spiritual well-being and functional well-being in women who have spontaneous premature ovarian failure was evaluated, an overall positive
Quality of Life in Amenorrhea and Oligomenorrhea
113
correlation could be demonstrated between the two. The Meaning/Peace subscale used in the study strongly correlated with functional well being (62% of variance), while the faith subscale was less strongly correlated with functional well being (7% of variance). It was suggested that strategies to improve spiritual well-being in the domains of meaning, purpose, and inner peace might provide a therapeutic approach to reduce the emotional suffering that accompanied the life-altering diagnosis of POF (Ventura et al., 2007). Thus, it can be seen from these studies that both idiopathic and surgically induced POF could be associated with significant psychological morbidity and negative impact on quality of life measures. Supportive counseling both at the disclosure of the diagnosis and at subsequent treatment are important aspects of the management of this group of women.
3.2
Polycystic Ovarian Syndrome
Polycystic ovarian syndrome (PCOS) is a heterogeneous condition which is defined by the presence of two out of the following three criteria (1) oligo-amenorrhea with oligo- and/or anovulation, (2) hyperandrogenism (clinical and/or biochemical), (3) polycystic ovaries (The Rotterdam ESHIRE PCOS Consensus Working Group, 2004). PCOS therefore encompasses symptoms of menstrual cycle dysfunction and as such comprises the most common cause of secondary amenorrhea (> Table 113-6). The medical problems that can be encountered by women with PCOS includes, in addition to menstrual dysfunction, possible infertility due to anovulation, obesity, increased tendency for insulin resistance and clinical diabetes mellitus, and > hirsutism and metabolic disturbances
. Table 113-6 Polycystic ovarian syndrome – clinical manifestations of polycystic ovarian syndrome Features Possible clinical manifestations
Conditions Obesity Oligomenorrhea and secondary amenorrhea Infertility Hirsutism Asymptomatic
Possible hormonal disturbances
Hyperandrogenism (raised testosterone and androstenedione levels) Raised luteinizing hormone levels Raised fasting insulin levels Mild hyperprolactinemia Low sex hormone-binding globulin Raised estradiol levels
Possible medical complications
Diabetes mellitus Hyperlipidaemia Hypertension and cardiovascular disorders Carcinoma of the breast Carcinoma of the endometrium
1931
1932
113
Quality of Life in Amenorrhea and Oligomenorrhea
associated with hyperandrogenism. Changes in physical appearance, particularly obesity and excessive body hair as well as the infertility have all been identified as important contributors to psychological problems in PCOS (McCook et al., 2005). Women with PCOS may thus fail to conform to societal norms because of their external appearance. Many PCOS women might feel stigmatized in the sense of a loss of ‘‘feminine identity.’’ In addition to somatic impairment, mood disturbances such as depression and limitations in emotional well-being, QOL and life satisfaction, as well as self-worth and sexual satisfaction could be negatively affected (Janssen et al., 2008). In a survey comparing PCOS women and controls using tools including the SF-36, women with PCOS revealed significantly lower scores for physical role function, bodily pain, vitality, social function, emotional role function, and mental health. Although the PCOS subjects and the controls had the same partner status and frequency of sexual intercourse, they were significantly less satisfied with their sex life and found themselves less attractive (Elsenbruch et al., 2003). In another questionnaire survey to women with PCOS, using various tools including the SF-36, the Quality of Life Questionnaire for Women with Polycystic Ovary Syndrome (PCOSQ) and the General health Questionnaire-28 (GHQ-28), it was found the PCOS women had SF-36 scores that were significantly lower than age-matched controls, and had higher psychological morbidity of 62.4% as compared to 26.4% in controls. The body mass index was negatively correlated with QOL. In addition, a positive association could be seen between the psychological domain of QOL and the subjective assessment of the quality of health information, hirsutism and menstrual irregularity (Ching et al., 2007). The findings clearly demonstrated that impaired QOL and increased prevalence of psychological morbidity in PCOS women as compared to controls. As the perception of inadequate information about the condition correlated with poorer QOL scores, it has been suggested that provision of detailed information to affected women might lead to an improvement in QOL scoring. Cultural and ethnic differences could also induce a major impact on the psychological burden from PCOS. In one study comparing health related QOL among Austrian women and Moslem immigrant women in Austria with infertility caused by PCOS, it was found that the scores from the latter group were affected to a greater extent, although no differences in the symptomatology could be shown between the two groups. This was true of all the five domains tested, including infertility, overweight, hirsutism, menstrual irregularities and emotional problems. This difference was explained by the fact that infertility was a very dramatic problem for immigrant Austrian Moslems who were under great social pressure to reproduce. Thus, health professionals should be sensitive to the ethnicity, religious and cultural background of the patients to provide the best medical support (Schmid et al., 2004). Recent developments in the use of insulin sensitizers such as metformin for weight control, together with psychological counseling and the participation in support groups, could further improve life satisfaction and coping in these women. In particular, metformin has been shown to improve biochemical, clinical and reproductive parameters in PCOS women. In a prospective study measuring health related QOL, emotional well being and sexuality were assessed using the SF-36 and other validated tools before, during and 6 months after treatment of PCOS by metformin. It was reported that during treatment, the psychological aspects of the SF-36 score, (including Vitality, Social function, Emotional Role Function, Mental Health and Psychological Sum scale) and emotional well being (as reflected by a lowering of the SCL-90-R scores) were significantly improved. These improvements in QOL scores significantly correlated with a reduction in body weight and significantly more pronounced in patients with normalized menstrual cycles. In addition, these patients were significantly more satisfied
Quality of Life in Amenorrhea and Oligomenorrhea
113
with their sex life and reported higher frequencies of sexual intercourse following treatment. It was concluded by the authors that metformin treatment can substantially improve the psychosocial, emotional and psychosexual situation of PCOS (Hahn et al., 2006). In summary, the literature had consistently shown the negative impact of the diagnosis of PCOS on health related QOL. Studies on the QOL of PCOS women employed both generic tools such as the SF-36 as well as specific questionnaires designed for evaluating co-morbidity in POCOS women. The availability of interdisciplinary treatment aimed at improving PCOS related symptoms, including the hirsutism, obesity, infertility and menstrual dysfunction, should help to reduce psychological distress and improve self-worth.
3.3
Hypothalamic Menstrual Dysfunction and Associated Disorders
> Hypothalamic amenorrhea and associated anovulation is the cessation of menses and ovulation in women with no identifiable organic causes for amenorrhea. This common and mostly reversible condition accounts for around 15–25% of cases of secondary amenorrhea. As the anovulation can occur intermittently at intervals in the women’s cycles, the clinical manifestation of oligomenorrhea often co-exists with amenorrhea. The pathophysiology of such functional hypothalamic amenorrhea has been ascribed to insufficient hypothalamic pulsatile gonadotrophin releasing hormone (GnRH) secretion, which in turn leads to reduced pituitary production of gonadotrophins (FSH and LH) with resulting anovulation. Numerous factors have been associated with hypothalamic amenorrhea, including environmental stressors, personality traits, psychological disorders, exercise, low body weight and weight loss. The combination of mild energy deprivation induced by nutritional or calorie restriction or exercise and psychosocial distress act synergistically to provoke a constellation of hypothalamic alterations that will disrupt the GnRH drive (Marcus et al., 2001). Hypothalamic menstrual dysfunction may be idiopathic, but can also be the result of eating disorders such as anorexia nervosa or bulimia, or can be the consequence of intense physical training and the stress of such training. These factors are often inter-related, as many sports that require intensive physical exercises also demands that the individual should remain aesthetic and slender in built. This scenario is most commonly seen in dancers, particularly ballerinas, and in gymnasts and athletes. For instance, in a study on a group of sports performers that emphasized thinness or muscularity, including ballet dancers, gymnasts and body builders, a high degree of body uneasiness and inappropriate eating attitudes and behaviors could be shown (Ravaldi et al., 2003). Indeed, body image disturbance was an essential element in almost all subjects suffering from eating disorders. Body image disturbance was common in these performers independent of their body mass index, suggesting that the body uneasiness was more related to their altered perception their body shape and not a consequence of the actual physical activity or demand. On the other hand, such body image disturbance was probably more severe in anorexic women as compared to exercising women with menstrual dysfunction. When elite gymnasts were compared to adolescents with anorexia nervosa, it was found that while elite gymnasts had significantly lower body mass index than the anorexic patients, they had only marginally distorted body image of their abdomen while the anorexic patients expressed a broad body image distortion (Salbach et al., 2007). Other studies on ballet dancers have found a high life time incidence for anorexia nervosa (6.9%), bulimia nervosa (10.3%) or a combination of the two (10.3%), and concluded that dancers frequently engage in binge eating and purging behaviors and that the pathology could be as severe as
1933
1934
113
Quality of Life in Amenorrhea and Oligomenorrhea
non-dancers with eating disorders (Ringham et al., 2006). Thus, the very high incidences of amenorrhea and oligomenorrhea in dancers could at least in part be explained by the underlying eating disorder, in addition to the impact of strenuous exercises. In a recent study comparing adolescent dancers and age-matched non-exercising women with or without oligomenorrhea and amenorrhea using the SF-36, it was found that oligo/ amenorrheic dancers did not score lower than eumenorrheic dancers. Compared to eumenorrheic controls or to oligo/amenorrheic dancers, oligo/amenorrheic non-exercising women had significantly lower QOL scores in the domains of physical functioning (PF) and general health (GH) and vitality (VT). It was concluded that QOL scores were lower in non-exercising adolescents with oligo/amenorrhea, but not those with exercise related oligo/amenorrhea. The negative effects of oligomenorrhea and amenorrhea on quality of life were apparently attenuated if the menstrual dysfunction was related to physical training (To and Wong, 2007). These results suggested that menstrual dysfunction was a more significant psychological issue for non-dancers as compared to dancers. As the incidence of menstrual dysfunction was often extremely high among dancers, there would be a tendency for these dancers to consider menstrual dysfunction as a relatively normal sequelae to their intensive physical training, and thus might not regard the problem as pathological at all. There appeared to be good physiological grounds to consider such menstrual dysfunction as normal (DeSouza and Williams, 2004) and other studies on dancers also showed a lack of significant differences in self-image scores in those that developed menstrual dysfunction as compared to those who . Table 113-7 Comparison of QOL scores by SF 36 in women with oligomenorrhea and other forms of menstrual abnormalities I
Domains
II
III
Other dysfunctional Dysmenorr-hea Oligo/amen bleeding N = 43(SD) orrhea N = 86(SD) N = 106(SD)
ANOVA
F
P-value
85.7(12)
6.45
0.002
Physical functioning
89.8(7.03)
81.6(15.2)
Role-physical
93.3(20.4)
82(23.9)
87.7(25.1)
3.39
0.035 0.01
Bodily pain
50(16.7)
59.5(22.1)
61.3(20.8)
4.72
General health
62.0(19.1)
57.9(20.3)
60.1(19.1)
0.65
Vitality
52.1(12.5)
42.1(15.6)
53.5(13.7)
Social functioning
80.1(18.2)
66.5(20.8)
71.0(20.8)
6.37
0.002
Role-emotional
65.1(32)
58.3(36.1)
62.6 (35.4)
0.62
0.53
Mental health
62.8(16)
59.5(17.1)
63.4(15.8)
1.47
0.23
Physical summary score
54.8(5.15)
51(8.6)
52(6.3)
4.05
0.019
Mental summary score
43.6(10.2)
39.6(10.4)
43.5(9.23)
4.29
0.015
Modified from Yang and To (2006)
16.2
0.52 0.001
Quality of Life in Amenorrhea and Oligomenorrhea
113
remained eumenorrheic during training (To et al., 2000). Qualitative psychological studies have shown that young women often have preset stereotypes of their menstrual experience (Choi and McKeown, 1997), and gross deviations from such experiences would be interpreted as pathological. This could explain the differential reactions of dancers and non-dancers to their oligomenorrhea and amenorrhea. Whether such an accepting perception would have any implications on the actual physical and psychological well being of young dancers suffering from menstrual dysfunction would require further evaluation. When the impact on QOL was measured in young women with various common menstrual problems, it was found that oligomenorrhea and amenorrhea has more significant negative impact on the quality of life scores than dysmenorrhea or other forms of dysfunctional uterine bleeding (Yang and To, 2006) (> Table 113-7). Thus, it can be seen that despite the common belief that pain symptoms (in the form of dysmenorrhea) should be of more significance to QOL than the relatively asymptomatic oligomenorrhea or amenorrhea, if affected women should accept the dysmenorrhea as part of their womanhood, while the amenorrhea or oligomenorrhea would be interpreted as a threat to their normal reproductive function, the impact of these symptoms on QOL could be reversed.
4
Conclusion
While amenorrhea and oligomenorrhea are closely related to reproductive function, little attention has been paid in the past to the psychological impact of these symptoms and their effect on the women’s QOL. Current available literature on the QOL associated with amenorrhea and oligomenorrhea has largely focused on specific conditions such as Turner syndrome, premature ovarian failure, or polycystic ovarian syndrome, while many other conditions have not yet been evaluated in detail. Demonstrable negative effects on the perceived QOL could be seen in women suffering from oligomenorrhea and amenorrhea with various conditions, indicating that clinical treatment should no longer be confined to medical therapy, but should extend to management of emotional and psychological well-being and social support.
Summary Points With reference to quality of life studies, it is apparent from a review of the relevant
literature that many causes of amenorrhea and oligomenorrhea have not been studied specifically in detail. The common causes of primary amenorrhea include Turner syndrome and anatomical abnormalities of the reproductive tract, such as congenital absence of the uterus. Women with Turner syndrome or absence of the uterus often face associated problems of infertility, inability to marry or establish a family, and suffer significant negative impact on their perceived quality of life. The common causes of secondary amenorrhea include premature ovarian failure, polycystic ovarian syndrome and hypothalamic amenorrhea, with specific associated problems of infertility, obesity, or eating disorders respectively. A negative impact on their perceived quality of life can frequently be demonstrated. The impact on quality of life in women with oligomenorrhea with/without secondary amenorrhea largely depends on their interpretation and acceptance of the symptoms.
1935
1936
113
Quality of Life in Amenorrhea and Oligomenorrhea
References Balen AH, Shoham Z, Jacobs HS. (1993). In: Asch RH and Studd JJW (eds.) Annual Progress in Reproductive Medicine. Parthenon Press, Carnforth, Lancashire, pp. 205–234. Bannink EM, Raat H, Mulder PG, de Muinck KeizerSchrama SM. (2006). J Pediatr. 48: 95–101. Boughton MA. (2002). J Adv Nurs. 37: 423–430. Cardoso G, Daly R, Haq NA, Hanton L, Rubinow DR, Bondy CA, Schmidt P. (2004). Gynecol Endocrinol 19: 313–319. Carel JC, Ecosse E, Bastie-Sigeac I, Cabrol S, Tauber M, Leger J, Nicolino M, Brauner R, Chaussain JL, Coste J. (2005). J Endocrinol Metab. 90: 1992–1997. Carel JC, Elie C, Ecosse E, Tauber M, Leger J, Cabrol S, Nicolino M, Brauner R, Chaussain JL, Coste J. (2006). J Clin Endocrinol Metab. 91: 2972–2979. Ching HL, Burke V, Stuckey BGA. (2007). Clin Endocrinol. 66: 373–379. Choi PY, McKeown S. (1997). J Psychoso Obstet Gynecol. 18: 259–265. Coffey S, Mason H. (2003). Gynecol Endocrinol. 17: 379. Conway GS. (2002). Baillieres Best Pract Res Clin Endocrinol Metab. 16: 243–261. Coulam CB, Adamson SC, Annegers JF. (1986). Obstet Gynecol. 67: 604–606. Critchley HOD. (2003). In: Shaw RW, Soutter WP, Stanton SL (ed.) Gynaecology, 3rd ed. Churchill Livingstone, London, pp. 229–244. DeSouza MJ, Williams NI. (2004). Hum Reprod Update. 10: 433–438. De Taraciuk MB, Nolting M, Fernandez G, Colela D, Onetto C, Straminsky V. (2008). Gynecol Endocrinol. 24: 44–53. Elsenbruch S, Hahn Kowalsky D, Offner AH, Schedlowski M, Mann K, Janssen OE. (2003). J Clin Endocrinol Metab. 88: 5801–5807. Groff AA, Covingtom SN, Halverson LR, Fitzgerald OR, Vanderhoof V, Calis K, Nelson LM. (2005). Fert Steril. 83: 1734–1741. Guyatt G, Weaver B, Cronin L, Dooley JA, Azziz R. (2004). J Clin Epidemiol. 51: 1279. Habelt K, Brosche T, Riedel HH. (1996). Zentralbl Gynakol. 118: 206–212. Hahn S, Benson S, Elsenbruch S, Pleger K, Tan S, Mann K, Schedlowski M, van Halteren WB, Kimmig R, Janssen OE. (2006). Human Reprod. 21: 1925–1934. Harlow SD, Ephross SA. (1995). Epidemiologic Rev. 17: 256–286. Hull KL, Harvey S. (2003). J Endocrinol. 179: 311–333. Janssen OE, Hahn S, Tan S, Benson S, Elsenbruch S. (2008). Semin Reprod Med. 26: 45–52. Kadir RA, Sabin CA, Pollard D, Lee CA, Economides DL. (1998). Haemophilia. 4: 836–41.
Khastgir G, Studd J. (1999). Menopause. 6: 180–181. Kilic BG, Ergur AT, Ocal G. (2005). J Pediatr Endocrinol. 18: 1111–1117. Liao KL, Wood N, Conway GS. (2000). J Psychosom Obstet Gynaecol. 21: 167–174. Manocchia M, Bayliss MS, Connor J, Keller SD, Shiely JC, Tsai C, Voris RA, Ware JE. (1998). SF-36 Health Survey Annotated Bibliography: Second Edition (1988–1996). The Health Assessment Lab, New England Medical Centre, Boston, MA. Marcus MD, Loucks TL, Berga SL. (2001). Fert Steril. 76: 310–316. McCook JG, Reame NE, Thatcher SS. (2005). J Obstet Gynecol Neonatal Nurs. 34: 12–10. Morgan T. (2007). Am Fam Physician. 76: 405–410. Okada Y. (1994). J Endocrinol. 41: 345–354. Orley J, Saxena S, Herman H. (1998). Br J Psychiat. 172: 291–293. Pavlidis K, McCauley E, Sybert VP. (1995). Clin Genet. 47: 85–89. Ravaldi C, Vannacci A, Zucchi T, Mannucci E, Cabras PL, Boldrini M, Murciano L, Rotella CM, Ricca V. (2003). Psychopathology. 36: 247–254. Ringham R, Klump K, Kaye W, Stone D, Libman S, Stowe S, Marcus M. (2006). Int J Eat Disord. 39: 503–508. Rotterdam ESHERE/ASRM-sponsored PCOS consensus workshop group. (2004). Human Reprod. 19: 41–43. Salbach H, Klinkowski N, Pfeiffer E, Lehmkuhl U, Korte A. (2007). Psychopath. 40: 388–393. Schmid J, Kirchengast S, Vytiska-Binstorfer E, Huber J. (2004). Human Reprod. 19: 2251–2257. Sutton EJ, McInerney-Leo A, Bondy CA, Gollust SE, King D, Biesecker B. (2005). Am J Med Genet A. 139: 57–66. Sybert VP, McCauley E. (2004). N Engl J Med. 35: 1227–1238. Sylven L, Hagenfeldt K, Brondum-Nielsen K, vonSchoultz B. (1991). Acta Endocrinol. (Copenh) 125: 359–365. To WWK, Wong MWN. (2007). J Pediatr Adolesc Gynecol. 20: 83–88. To WWK, Wong MWN, Lam IVY. (2000). Acta Obstet Gynecol. 79: 1117–1123. Ventura JL, Fitzgerald OR, Koziol DE, Covington SN, Vanderhoof VH, Calis KA, Nelson LM. (2007). Fert Steril. 87: 584–590. Ware JE, Sherbourne CD. (1992). Med Care. 30: 473–480. World Health Organization multicenter study on menstrual and ovulatory patterns in adolescent girls. (1986). J Adole Health Care. 7: 236–244. Yang TM, To WWK. (2006). HK J Gynaecol Obstet Midwif. 6: 24–31.
114 Quality of Life Among Japanese Oral Contraceptive Users Y. Matsumoto . S. Yamabe . K. Ideta 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1938 2 Whoqol-Bref . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1939 3 Risks and Benefits of OC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1939 4 Current Situation of OC Use in Japan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1940 5 QOL Among Japanese OC Users . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1941 6 QOL Differences Among Contraceptive Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1941 7 Ideal OC for Better QOL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1944 8 OC for Sexual and Reproductive Health Well-Being . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1945 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1946 Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1946 WHOQOL-BREF . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1946 Do you have any comments about the assessment? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1949
#
Springer Science+Business Media LLC 2010 (USA)
1938
114
Quality of Life Among Japanese Oral Contraceptive Users
Abstract: It is difficult to judge the well-being in sexual and reproductive health as it is very subjective matter. QOL is a possible indicator to measure sexual and reproductive health. > Oral contraceptives (OC) can contribute greatly to a woman’s well being by providing a choice of family planning and also as a tool to improve menstruation and hormonal-dependent symptoms such as > dysmenorrhea, irregular menstrual cycles, > menorrhagia, > premenstrual syndrome (PMS), and acne. WHO has developed an assessment instrument for QOL, which has proven systematically relevance in cross-cultural research. WHOQOLBref is comprised of 5 domains; physical health, psychological, social relationships, environment, and overall. We performed a questionnaire study to measure QOL among Japanese OC users. QOL measurement using WHOQOL-Bref showed the effectiveness of OC in providing side benefits, however, OC can worsen the QOL of users if it is taken only for contraceptive purpose. Studies have proven that women are more satisfied with permanent contraceptive methods such as vasectomy and tubal ligation than nonpermanent methods such as intrauterine device, injection, and OC. However, compliance has improved among OC users after counseling. Information, education, and counseling before staring OC will improve the practice of family planning and promote QOL among OC users. List of Abbreviations: DSFI, derogatis sexual functioning inventory; ICPD, international conference on population and development; OC, oral contraceptives; PMS, premenstrual syndrome; STD, sexually transmitted diseases; WHOQOL, World Health Organization quality of life
1
Introduction
Reproductive health is defined as ‘‘a state of complete physical, mental, and social well-being in all matters related to the reproductive system and to its functions and processes’’ at the International Conference on Population and Development (ICPD) in Cairo 1994 (UNFPA, 2008). After the ICPD conference in Cairo and the Fourth World Conference on Women in Beijing, the concept of ‘‘sexual and reproductive health rights’’ has gained some global acceptance. Sexual health is included in reproductive health and the purpose of sexual health is stated as ‘‘the enhancement of life and personal relations (UNFPA, 2008).’’ If the purpose of sexual health is ‘‘the enhancement of life and personal relations,’’ how can it be measured? How is an ill condition defined in terms of sexual and reproductive health? One endpoint of that can be sexually transmitted diseases (STD); therefore STD prevention and treatment are essential to promote good sexual and reproductive health. However, there are few clear endpoints to judge the condition of sexual and reproductive health. If so, how we can measure good or bad ‘‘relations of equality and mutual respect between genders’’ and ‘‘enhancement of life’’? Measuring sexual and reproductive health is difficult since it is considered private and there may also be inhibitions about discussing sexual issues in some culture. Although it has been recognized that sexual health, as a part of reproductive health, is essential as one of the basic human rights, improvement of sexual health is rarely a health focus. The use of OC is closely related to sexual and reproductive health. Talking about OC is taboo to some extent since it is regarded as a private matter, thus it is difficult to clarify its real impact on users’ lives. Despite the difficulty in measuring impact of OC, it is necessary to evaluate the effects of OC on individual subjects using certain indicators of QOL.
Quality of Life Among Japanese Oral Contraceptive Users
114
The mechanism of OC is to prevent ovulation by inhibiting gonadotropin secretion via an effect on both the pituitary and hypothalamic center. The progestational agent of OC suppresses luteinizing hormone surge in order to prevent ovulation, while the estrogenic agent promotes expression of progesterone receptors and suppresses follicle stimulating hormone secretion to prevent the emergence of a dominant follicle (Speroff and Fritz, 2005). Using universal QOL measuring scale such as World Health Organization Quality of Life (WHOQOL) questionnaire, we could overcome to some extent the difficulties in finding adequate indicators of sexual and reproductive health. For OC users, QOL measurement can be a tool to judge the effectiveness of OC in improving their lives.
2
Whoqol-Bref
QOL measurements has been used more frequently in clinical research to measure improvement in perceived well-being and a growing number of generic as well as disease-specific QOL assessment instruments are becoming available. However, QOL measurements are sometimes inadequately understood and wrongly targeted in the reported research. Therefore, there was an urgent need for research into the best ways of measuring and assessing QOL, giving particular attention to patients’ viewpoints. One of the fundamental issues in the area of assessment of QOL is to determine what is important to the individuals’ QOL. This is even more crucial when the instrument is to be used in diverse cultural settings (Saxena et al., 2001; Guyatt and Cook, 1994). WHO has been coordinating a multi-centered collaborative project on the development of an assessment instrument for QOL since the early 1990s. WHOQOL-Bref is comprised of 5 domains; physical health, psychological, social relationships, environment, and overall. It has shown proven relevance in cross-cultural research in a WHOQOL pilot field trial involving 4,804 respondents from 15 centers in 14 developed and developing countries using 12 languages (Saxena et al., 2001).
3
Risks and Benefits of OC
There is no medicine that does not have any side effects and OC is not an exception. First of all, OC increase the risk of venous thromboembolism since pharmacologic > estrogen in OC increases the clotting factor whereas > progestin has no significant impact (Speroff and Fritz, 2005). Estrogen has an additive effect on the risk of arterial thrombosis due to smoking, causing myocardial infarction and strokes. However, OC alone do not increase the risk in healthy, nonsmoking women, regardless of age (Speroff and Fritz, 2005). It has been reported that there is a small increase in breast cancer risk among current and recent combined OC users which disappears 10 years after cessation of use (Vessey and Painter, 2006). However, a recent British cohort study suggested that the breast cancer incidence was not related to the duration of OC use, although there is a strong positive relationship between cervical cancer incidence and the duration of OC use (Cogliano et al., 2005). Minor clinical problems such as nausea, edema, headache, breast tenderness, > breakthrough bleeding, spotting, amenorrhea, and missing pills are very common and have led users to cease taking OC. OC has several benefits besides contraceptive purpose. A recent randomized trial showed that OC also relieved pain associated with menstruation cycle such as ovulatory pain and
1939
1940
114
Quality of Life Among Japanese Oral Contraceptive Users
menorrhegia more effectively than placebo (Davis et al., 2005). OC is efficient in regulating menses and preventing bone loss in women with > secondary amenorrhea who are susceptible to the development of osteoporosis (The Practice Committee of the American Society for Reproductive Medicine, 2004). OC can improve acne regardless of the type of progestogen contant (Rosen et al., 2003). Fraser and McCarron reported that menstrual flow showed 43% reduction after using OC for more than 2 cycles (Fraser and McCarron, 1991). Moreover, uterine body cancer and ovarian cancer were strongly negatively associated with OC use and the rate of these cancers declined steadily with increasing duration of use, suggesting that minimum period of use may be necessary to obtain a beneficial effect (Cogliano et al., 2005). Although the influence of OC on mode and affect remain in conflict, Joffe et al. suggested that OC had a positive effect on women with PMS (Joffe et al., 2003).
4
Current Situation of OC Use in Japan
It has been declared that all men and women have the right to be informed and to have access to safe, effective, affordable, and acceptable methods of family planning, as well as other legally available methods of their choice for regulation of fertility, and the right of access to healthcare services that will enable women to go safely through pregnancy and childbirth (UNFPA, 2008). Furthermore, women’s autonomy and socially supported capacity for reproductive health and choice should be promoted (Miller, 2000). However, in Japan, there has been very limited choice among contraceptive methods as shown in > Table 114-1. And low-dose oral contraceptive pills were just approved by the government in 1999. As a contraceptive method, OC is as effective as 0.3% (perfect use) and 8% (typical use) of women experiencing an unintended pregnancy within the first year of use, which is obviously lower than that using a male condom (2%: perfect use, 15%: typical use) or periodic abstinence (9%: perfect use, 25%: typical use) (World Health Organization, 2004). Periodic abstinence (known as the calendar method) and male condoms have been major contraceptive methods in Japan and the common point is that these two methods are totally male dependent. OC is a method of contraception that females can chose and perform totally by themselves.
. Table 114-1 Prevalence of modern contraceptive methods (Percentage among women aged 15–49, married or in union, 2005) Sterilization Any method Female Male OC
a
Japan
55.9
←
World
60.5
20.5
3.6
Injectable or implant
IUD Condom
!
2.3a
b
a
3.4
7.5
3.2
13.6
Vaginal barrier methods
42.1
15.4
4.8
0.4
2.3% include OC and IUD prevalence in Japan Both injectable and implant are not available in Japan Adapted and modified from: United Nations Population Division Department of Economics and Social Affairs World Contraceptive Use 2005 (United Nations Population Division Department of Economics and Social Affairs, 2005)
b
Quality of Life Among Japanese Oral Contraceptive Users
114
Even after the ICPD conference in Cairo and the Fourth World Conference on Women in Beijing, OC was not approved by the Japanese government due to a fear of spreading STDs, especially HIV/AIDS. OC was approved finally in 1999; therefore we have only a short history of OC use in Japan. Only less than 2.3% of Japanese women of reproductive age (15–49 years) use OC for contraception (United Nations Population Division Department of Economics and Social Affairs, 2005), however, usage of OC has gradually expanded due to side benefits such as relaxation of dysmenorrhea, regulation of menstrual cycles, improvement of acne, remission of menorrhagia, and improvement of PMS since OC has come commercially available. These side benefits of OC are well known especially among internet users in Japan (Matsumoto et al., 2006). And also it has become common for internet users to review their symptoms and choice of treatments by using internet search engines before consulting doctors (Sullivan and Wyatt, 2005). Electronic health tools provide access to many resources that may satisfy patients’ need in this internet era. Therefore, as responsible health providers, we should be sensitive to what patients really want for their well-being.
5
QOL Among Japanese OC Users
We performed a questionnaire study using the Japanese version of WHOQOL-Bref questionnaires in Japan in 2005. There were 217 new OC users participating in the study and followed these subjects for 3 months after taking OC. All participants were divided into six groups according to their purposes for starting to take OC. These reasons were contraception, relaxation of dysmenorrhea, regulation of menstrual cycles, improvement of acne, remission of menorrhagia, and improvement of PMS. Participants often showed overlapping reasons because these propose were presented as multiple choices and some of subjects had more than two purposes for beginning to take OC. WHOQOL scores were significantly higher in all domains in the dysmenorrhea group. Scores were also increased in the social and overall domains of the irregular cycle group. In the acne group, physical, environmental, and overall domain were better after taking OC for 3 months, and the PMS group showed improvement in psychological and overall domain. In the social domain of the contraception group, the WHOQOL score was decreased after 3 months OC use (> Table 114-2) (Matsumoto et al., 2007). Our findings suggested marvelous effects on QOL among Japanese women using OC for reasons related to its side benefits. OC is originally a contraceptive method and in Japan, that is the only indication approved by the government. Nevertheless, the QOL score of those who took OC for contraceptive purposes was decreased in the social domain. To improve clinical practice, we need to determine why QOL worsened among OC users seeking a contraceptive purpose.
6
QOL Differences Among Contraceptive Methods
Women’s contraceptive use and choice are strongly influenced by perceptions of the physical and psychological effects of contraceptive methods and also by their satisfaction with each method (Oddens, 1999; Rosenfield et al., 1993). The study in Hong Kong, also proved that OC did not improve the QOL of women any more than other contraceptive methods such as
1941
1942
114
Quality of Life Among Japanese Oral Contraceptive Users
. Table 114-2 Quality of life score [mean(SD)] at recruitment and at 3-month follow-up At recruitment
At 3-month follow-up
P-value
1.Physical domain
23.61(3.59)
24.03(3.72)
0.047
2.Psychological domain
19.57(3.49)
20.04(3.79)
0.017
3.Social domain
10.97(1.52)
10.91(1.88)
0.001
4.Environmental domain
27.43(3.38)
27.92(3.95)
0.002
6.03(1.42)
6.55(1.47)
0.001
1.Physical domain
24.33(3.25)
24.02(4.05)
0.143
2.Psychological domain
19.76(3.67)
20.00(4.34)
0.247
3.Social domain
11.09(1.63)
10.87(2.30)
0.011
4.Environmental domain
27.60(3.35)
27.54(3.83)
0.285
6.53(1.40)
6.83(1.62)
0.007
1.Physical domain
20.11(2.08)
20.29(3.67)
0.002
2.Psychological domain
20.06(2.94)
19.76(2.75)
0.001
3.Social domain
10.83(1.72)
10.71(2.20)
0.001
4.Environmental domain
28.17(3.20)
27.94(4.10)
0.001
6.78(1.44)
6.94(1.64)
0.013
1.Physical domain
22.36(3.86)
23.71(3.95)
0.019
2.Psychological domain
19.10(3.41)
20.62(3.87)
0.002
3.Social domain
10.92(1.56)
11.03(2.10)
0.045
4.Environmental domain
27.15(3.83)
28.79(4.82)
0.0005
5.54(1.45)
6.44(1.50)
0.001
1.Physical domain
21.08(3.40)
22.84(3.34)
0.110
2.Psychological domain
18.69(3.50)
20.31(3.86)
0.196
3.Social domain
10.62(1.61)
10.54(81.56)
0.285
4.Environmental domain
26.46(2.99)
28.15(4.58)
0.114
5.46(1.66)
6.15(1.68)
0.051
1.Physical domain
23.35(3.65)
23.10(3.73)
0.254
2.Psychological domain
19.16(3.22)
19.32(3.74)
0.202
3.Social domain
10.71(1.44)
10.84(1.49)
0.002
4.Environmental domain
26.65(3.30)
26.90(3.30)
0.118
5.61(1.28)
6.10(1.37)
0.008
23.32(3.70)
24.65(3.33)
0.030
Overall (n = 110)
5.Ovarall domain Contraception total (n = 46)
5.Ovarall domain Contraception only (n = 17)
5.Ovarall domain Menalgia (n = 39)
5.Ovarall domain Menorrhagia (n = 13)
5.Ovarall domain Irregular menses (n = 31)
5.Ovarall domain Acne (n = 31) 1.Physical domain
Quality of Life Among Japanese Oral Contraceptive Users
. Table 114-2 (continued)
114
At recruitment
At 3-month follow-up
P-value
2.Psychological domain
18.68(4.11)
19.94(4.40)
0.075
3.Social domain
10.77(1.52)
10.87(1.61)
0.079
4.Environmental domain
27.48(3.91)
28.61(4.01)
0.012
6.00(1.39)
6.77(1.41)
0.001
1.Physical domain
22.67(3.03)
24.25(4.27)
0.131
2.Psychological domain
18.75(3.41)
20.17(3.86)
0.033
3.Social domain
10.58(1.62)
11.08(1.51)
0.086
4.Environmental domain
27.17(2.74)
28.25(4.09)
0.286
5.75(1.29)
6.67(1.72)
0.050
5.Ovarall domain PMS (n = 12)
5.Ovarall domain
Scores were significantly higher in all domains in the dysmenorrhea group and also increased in the social and overall domains of the irregular cycle group. In the acne group, physical, environmental, and overall domain were improved and the PMS group showed better result in psychological and overall domain. However, the score was decreased in the social domain of the contraception group after 3 months OC use
intrauterine contraceptive device or injectable contraceptives. Whereas the social domain of the WHOQOL score and sexual satisfaction subscale of the Derogatis Sexual Functioning Inventory (DSFI) were both significantly different, which suggested only female sterilization could improve QOL and satisfaction as a contraceptive method (Li et al., 2004). The DSFI assesses a variety of aspects of sexuality that have been theoretically linked to sexual function and the individual’s current state of sexual functioning can be measured by primary domains such as sexual cognition/fantasy, sexual arosal, sexual behavior/experience, orgasm, and sexual drive/relationship (Derogatis, 2008). Other studies in the USA showed that women were significantly more likely to be satisfied with permanent methods of contraception such as vasectomy, tubal ligation, and hysterectomy than nonpermanent methods, either coital-independent methods, such as oral contraceptives, or coital-dependent methods, such as condoms, foams, and gels (Rosenfield et al., 1993). Furthermore, all women were satisfied when their husband or partner had a vasectomy, whereas some women were dissatisfied by tubal ligations and hysterectomy because of their permanence and menstrual problems (Rosenfield et al., 1993). According to the survey among German women, satisfaction was greatest with female sterilization (92% of users), followed by OC (68% of those who had ever used), intrauterine device (59%), natural family planning (43%), and condoms (30%) (Oddens, 1999). The results of these studies from three different countries were affected by family planning practice and socio-economic background in each country. With all these differences, the results indicated that women were more likely to be satisfied with permanent methods than nonpermanent methods. As the Hong Kong and German studies did not include vasectomy, it is difficult to draw conclusions whether male sterilization offers higher satisfaction as a whole. When contraception is needed, discussion and education by health providers might improve the satisfaction with contraceptive methods. Misconception and ignorance is likely to be observed at a higher frequency among women who have either chosen not to use a pill or to use no contraception at all (van de Weijer, 2005). In countries with a low prevalence of OC usage like Japan, potential users may not be aware of the major and minor risks and benefits of OC.
1943
1944
114
Quality of Life Among Japanese Oral Contraceptive Users
There might be less information especially about minor, but common side effects such as spotting, nausea, and breast tenderness. If potential OC users are given more adequate information, education, and counseling about OC including the minor side effects, compliance with OC use would likely improve leading to better preventive health care. Counseling will also resolve misunderstandings due to persistent myths regarding OC such as weight gain and decreased fertility. The important role of a health care provider is to give suitable advice to women in search of a personal contraceptive. A prerequisite is the health provider’s understanding of the woman’s information level and interpretation, her somatic and psychosexual needs, her perception of product-related benefits and risks and her need to maintain control by understanding and making her own decision (Gaudet et al., 2004).
7
Ideal OC for Better QOL
OC has many side benefits as indicated previously, but there are also several annoying side effects besides major risks of fatality such as thrombosis and cancer. Japanese perceptions of the merits and demerits of taking OC identified by our study are shown in > Table 114-3 (Matsumoto et al., 2007). As indicated in > Table 114-3, subjects first mentioned the fear of missing a pill as a demerit of OC. Especially for reproductive-aged women, daily pill intake is a bothering task even though it is for their better health and control of their future life. Along with women’s contraceptive needs and lifestyles, the efficacy and reasonability of OC should be considered before making a decision. . Table 114-3 Benefits/Demerits of taking OC 1) What do you think is the benefit of OC? Relaxation of menalgia
34.5%
Regulation of menses
34.5
Reliable contraception
10.9
Remission of menorrhagia
5.5
Improvement of acne
5.5
Improvement of PMS
3.6
Others
2.7
No answer
2.7
2) What do you think is the demerit of OC? Fear of missing a pill
26.4%
Nausea
7.3
Cost
4.5
Deterioration of depressive mood
3.6
Edema
1.8
Nothing particular
32.7
Others
23.6
Quality of Life Among Japanese Oral Contraceptive Users
114
Once-weekly transdermal patch has consistently better compliance at all age groups; and the contraceptive efficacy of the patch is comparable to OC. Furthermore, the patch shows higher satisfaction with increasing age, which may be attributed to improvements in emotional and physical well-being as well as reduction of premenstrual symptoms (Urdl et al., 2005). The transdermal patch has not yet been introduced in Japan, however, if it is approved, Japanese women may have another choice. Rosenberg et al. reported that reasons for discontinuing OC were mainly side effects such as bleeding irregularities, nausea, weight gain, mood changes, breast tenderness, headaches, and also clinician-recommended discontinuation (Rosenberg and Waugh, 1998). Irregular bleeding in the first few months after starting OC can be observed 10–30% of the users and seen more often among the lowest dose OC. The onset of bleeding is not associated with decreased efficacy and usually disappears by the third cycle in the majority of women (Speroff and Fritz, 2005). As it is the leading reason for discontinuing OC (Rosenberg and Waugh, 1998), counseling with encouragement and reassurance will improve user’s satisfaction. There is no evidence that OC causes weight gain, however young women, especially adolescents, are very preoccupied with body image and fear weight gain with use of OC (Urdl et al., 2005; Gupta (2000)). Weight gain is obviously a problem of perception and should be dispelled so that compliance with OC use will improve. Nausea, breast tenderness, and headache are symptoms that are commonly seen as excess estrogen symptoms. If there are any of these estrogen dependent side effects, OC with a lower estrogen dose may relieve the symptoms. However, these symptoms are most intense in the first few months of use and, in most cases, gradually disappear (Speroff and Fritz, 2005), moreover there is not a significant difference in the incidence of nausea, headache, and weight gain between the OC group and the placebo group (Redmond et al., 1999). Thus, in this case, informed choice and counseling are also essential for better results. Mood change including decreased libido is also a common complaint of OC users (Redmond et al., 1999). There was a report to show that more women reported a decline in the frequencies of sexual intercourse, sexual thoughts, and psychosexual arousability, thus they discontinued or switched OC (Sanders et al., 2001). The reasons why these side effects occur have been studied pharmacologically and psychosocially. There was no detectable association between hormone levels and emotional functioning in females. Psychiatric evaluations among OC users and other non hormonal contraceptive methods did not demonstrate any significant differences. Moreover, women who were given a placebo experienced a side effect profile similar to that of OC users. These findings indicated that most of the side effects of hormonal contraception are a result of a psychological response to the practice of contraception (Robinson et al., 2004). Many of the side effects have a chance to be reduced or disappear following information, education, and counseling as indicated above. For women to achieve a better QOL when using OC, health care providers should bear in mind the need to give well informed choices to OC users.
8
OC for Sexual and Reproductive Health Well-Being
OC is a valuable tool for women to achieve better sexual and reproductive health. It is obviously important as a contraceptive method that supports a woman’s choice to decide
1945
1946
114
Quality of Life Among Japanese Oral Contraceptive Users
when and how often to become pregnant. The need to improve the quality of family planning practice by promoting clients’ informed and voluntary choice should be more emphasized. Moreover, the extended use of OC to achieve side benefits such as relaxation of dysmenorrhea, regulation of menstrual cycles, improvement of acne, remission of menorrhagia, and improvement of PMS has should become more widely known among health care providers as well as users to improve women’s QOL.
Summary Points
QOL measurement is an important indicator to judge sexual and reproductive well-being OC contributes to women’s QOL by providing side benefits When used for contraceptive purpose only, OC can worsen the QOL of users Information, education, and counseling before staring OC will improve the practice of family planning as well as users’ QOL
Appendix WHO Quality of Life-BREF (WHOQOL-BREF)
WHOQOL-BREF The following questions ask how you feel about your quality of life, health, or other areas of your life. I will read out each question to you, along with the response options. Please choose the answer that appears most appropriate. If you are unsure about which response to give to a question, the first response you think of is often the best one. Please keep in mind your standards, hopes, pleasures, and concerns. We ask that you think about your life in the last four weeks.
1. How would you rate your quality of life?
Very poor
Poor
Neither poor nor good
Good
Very good
1
2
3
4
5
Very dissatisfied Dissatisfied 2. How satisfied are you with your health?
1
2
Neither satisfied nor dissatisfied 3
Very Satisfied satisfied 4
5
114
Quality of Life Among Japanese Oral Contraceptive Users
The following questions ask about how much you have experienced certain things in the last four weeks. Not at A A moderate Very all little amount much
An extreme amount
3. To what extent do you feel that physical pain prevents you from doing what you need to do?
5
4
3
2
1
4. How much do you need any medical treatment to function in your daily life?
5
4
3
2
1
5. How much do you enjoy life?
1
2
3
4
5
6. To what extent do you feel your life to be meaningful?
1
2
3
4
5
Not at all
A little
A moderate amount
Very much
Extremely
7. How well are you able to concentrate?
1
2
3
4
5
8. How safe do you feel in your daily life?
1
2
3
4
5
9. How healthy is your physical environment?
1
2
3
4
5
The following question refers to how often you have felt or experienced certain things in the last four weeks.
Not at all A little Moderately Mostly Completely 10. Do you have enough energy for everyday life?
1
2
3
4
5
11. Are you able to accept your bodily appearance?
1
2
3
4
5
12. Have you enough money to meet your needs?
1
2
3
4
5
13. How available to you is the information that you need in your day-to-day life?
1
2
3
4
5
14. To what extent do you have the opportunity for leisure activities?
1
2
3
4
5
1947
1948
114
Quality of Life Among Japanese Oral Contraceptive Users
15. How well are you able to get around?
Very poor
Poor
Neither poor nor good
Good
Very good
1
2
3
4
5
Neither Very satisfied nor Very dissatisfied Dissatisfied dissatisfied Satisfied satisfied 16. How satisfied are you with your sleep?
1
2
3
4
5
17. How satisfied are you with your ability to perform your daily living activities?
1
2
3
4
5
18. How satisfied are you with your capacity for work?
1
2
3
4
5
19. How satisfied are you with yourself?
1
2
3
4
5
20. How satisfied are you with your personal relationships?
1
2
3
4
5
21. How satisfied are you with your sex life?
1
2
3
4
5
22. How satisfied are you with the support you get from your friends?
1
2
3
4
5
23. How satisfied are you with the conditions of your living place?
1
2
3
4
5
24. How satisfied are you with your access to health services?
1
2
3
4
5
25. How satisfied are you with your transport?
1
2
3
4
5
The following questions ask about how completely you experience or were able to do certain things in the last four weeks.
Quite Very Never Seldom often often Always 26. How often do you have negative feelings such as blue mood, despair, anxiety, depression?
5
4
3
2
1
Quality of Life Among Japanese Oral Contraceptive Users
114
Do you have any comments about the assessment? [The following table should be completed after the interview is finished] Transformed scoresa Equations for computing domain scores 27. Domain 1 (6-Q3) + (6-Q4) + Q10 + Q15 + Q16 + Q17 + Q18
Raw score
4–20 0–100
a. =
b:
c:
a. =
b:
c:
a. =
b:
c:
a. =
b:
c:
□ + □ + □ + □ + □ + □ + □ 28. Domain 2 Q5 + Q6 + Q7 + Q11 + Q19 + (6-Q26) □ + □ + □ + □ + □ + □ 29. Domain 3 Q20 + Q21 + Q22 □ + □ + □ 30. Domain 4 Q8 + Q9 + Q12 + Q13 + Q14 + Q23 + Q24 + Q25 □ + □ + □ + □ + □ + □ + □ + □ a
See Procedures Manual, pages 13–15
References Cogliano V, Grosse Y, Baan R, Straif K, Secretan B, Ghissassi FE, WHO International Agency for Research on Cancer. (2005). Lancet Oncol. 6(8): 552–553. Davis AR, Westhoff C, O’Connell K, Gallagher N. (2005). Obstet Gynecol. 106: 97–104. Derogatis LR. (2008). Int J Impot Res. 20(1): 35–44. Fraser IS, McCarron G. (1991). J Obstet Gynaecol. 31: 66–70. Gaudet LM, Kives S, Hahn PM, Reid RL. (2004). Contraception. 69(1): 31–36. Gupta S. (2000). Hum Reprod Update. 6(5): 427–431. Guyatt GH, Cook DJ. (1994). JAMA. 272(8): 630–631. Joffe H, Cohen LS, Harlow BL. (2003). Am J Obstet Gynecol. 189: 1523–1530. Li RHW, Lo SST, Teh DKG, Tong NC, Tsui MHY, Cheung KB, Chung TKH. (2004). Contraception. 70: 474–482. Matsumoto Y, Yamabe S, Asahara S, Yokota H, Mandai K, Ideta K. (2006). Adv Obstet Gynecol. 58(2): 130–135. Matsumoto Y, Yamabe S, Ideta K, Kawabata M. (2007). J Obstet Gynaecol Res. 33(4): 529–535. Miller AM. (2000). Health Hum Rights. 4(2): 68–109. Oddens BJ. (1999). Contraception. 59: 277–286.
Redmond G, Godwin AJ, Olson W, Lippman JS. (1999). Contraception. 60(2): 81–85. Robinson SA, Dowell M, Pedulla D, Mc Cauley. (2004). Med Hypotheses. 63(2): 268–273. Rosen MP, Breitkopf DM, Nagamani M. (2003). Am J Obstet Gynecol. 188: 1158–1160. Rosenberg MJ, Waugh MS. (1998). Am J Obstet Gynecol. 179: 577–582. Rosenfield JA, Zahorik PM, Saint W, Murphy G. (1993). J Fam Pract. 36: 169–173. Sanders SA, Graham CA, Bass JL, Bancroft J. (2001). Contraception. 64(1): 51–58. Saxena S, Carlson D, Billington R. (2001). Qual Life Res. 10(8): 711–721. Speroff L, Fritz MA. (2005). Oral contraception: Clinical gynecologic endocrinology and infertility, 7th ed. Lippincott Williams & Wilkins, Philadelphia, pp. 862–942. Sullivan F, Wyatt JC. (2005). BMJ. 331: 625–627. The Practice Committee of the American Society for Reproductive Medicine. (2004). Fertil Steril. 82: 266–272. UNFPA: ICPD & MDG Follow up; Summary of the ICPD Programme of Action http://www.unfpa.org/ icpd/summary.htm#chapter7 (Accessed on January 5, 2008).
1949
1950
114
Quality of Life Among Japanese Oral Contraceptive Users
United Nations Population Division Department of Economics and Social Affairs World Contraceptive Use. (2005). http://www.un.org/esa/population/ publications/contraceptive2005/2005_World_ Contraceptive_files/WallChart_WCU2005.pdf (Accessed on January 5, 2008). Urdl W, Apter D, Alperstein A, Koll P, Scho¨nian S, Bringer J, Fisher AC, Preik M. (2005). Eur J Obstet Gynecol Reprod Biol. 121(2): 202–210.
van de Weijer P. (2005). Eur J Contracept Reprod Health Care. 10 Suppl 1: 2–6. Vessey M, Painter R. (2006). Br J Cancer. 95(3): 385–389. World Health Organization. (2004). Medical eligibility criteria for contraceptive use http://www.who.int/ reproductive-health/publications/mec/mec.pdf (Accessed on January 5, 2008).
115 Premenstrual Syndrome and Premenstrual Dysphoric Disorder: Issues of Quality of Life, Stress and Exercise M. Kathleen B. Lustyk . W. G. Gerrish 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1952 2 PMS and PMDD Diagnostic Criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1953 3 Epidemiology/Etiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1955 4 Quality of Life . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1960 5 Symptomatology and Stress . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1962 6 Treatments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1967 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1971
#
Springer Science+Business Media LLC 2010 (USA)
1952
115
Premenstrual Syndrome and Premenstrual Dysphoric Disorder
Abstract: The > symptomatology associated with the menstrual cycle in women ranges broadly in severity. > Molimina is the subclinical symptomatology affecting up to 90% of all women. > Premenstrual Dysphoric Disorder (PMDD) is the most severe form of > premenstrual syndrome (PMS). PMDD is debilitating and consists mainly of affective symptomatology that interferes with quality of life (QOL). While the etiologies of PMS/ PMDD remain unknown, symptoms are both physiological and psychological and as such an interdisciplinary biopsychosocial approach is needed to investigate the burden and decreased QOL in sufferers. This burden is considerable as up to 30% of women suffer from PMS and 5–6% have PMDD with nearly 4 years of projected disability for the latter. Published treatment guidelines recommend behavioral modifications as first-line therapeutic interventions for PMS with effective pharmacological options approved for PMDD. However, the efficacies for behavioral interventions are not well established, in part due to weaknesses in the research methods used to test a treatment effect, and resultant inconsistencies in findings. In addition, some strategies involving daily effort (e.g., > Cognitive-Behavioral Therapy) may be impractical in the face of the unique characteristics of cyclic symptoms. Other strategies such as aerobic exercise may be effective, but require motivation to perform during a period of time when sufferers feel particularly poor. As such, aerobic exercise by itself may be an unrealistic treatment option. Treatments that can reduce and/or manage > stress, elevate mood, and curb physical discomforts are needed. However, it may be impracticable to expect therapeutic success in all of these areas from a single intervention. Current research is therefore investigating complementary combinations of pharmacological and behavioral treatments as possible management strategies for PMS/PMDD. List of Abbreviations: AAFP, American Academy of Family Physicians; ACOG, American College of Obstetricians and Gynecologists; CAM, Complementary and Alternative Medicine; CBT, Cognitive Behavioral Therapy; DSM, > Diagnostic and Statistical Manual; FDA, Food and Drug Administration; HPA, Hypothalmic-Pituitary-Adrenal; HR-QOL, Health Related Quality of Life; ICD, > International Classification of Diseases; NIMH, National Institute of Mental Health; PMDD, Premenstrual Dysphoric Disorder; PMS, Premenstrual Syndrome; QOL, Quality of Life; SSRIs, > Serotonin Specific Reuptake Inhibitors
1
Introduction
It is common for women of child-bearing age to experience discomfort during the days prior to menstruation. For some women, these premenstrual symptoms are severe enough to > affect their quality of life (QOL) by negatively affecting behavior and interfering with daily activities. According to Campagne and Campagne (2007), ‘‘More women and their families are affected by the physical and psychological irregularities due to premenstrual symptoms than by any other condition’’ (p. 4). Still, others seem to remain nearly symptom free or have the ability to cope with their discomforts. Attempts at understanding the nature of these extremes has led to the adoption of terms such as molimina, which describes the typical subclinical symptomatology affecting up to 90% of all women, Premenstrual Syndrome (PMS), which is the diagnosis given when symptomatology is severe enough to interfere with daily activities and negatively affect well-being, and Premenstrual Dysphoric Disorder (PMDD), the diagnosis for severe PMS with a specific focus on affective symptomatology.
Premenstrual Syndrome and Premenstrual Dysphoric Disorder
115
Unlike other psychophysical conditions that affect women on a daily basis, the burden of PMS/PMDD may be misperceived as less severe because it affects only a subset of women during their > luteal cycle phase. Yet, as Stoddard et al. (2007) point out ‘‘. . . [women] have between 400 and 500 menstrual cycles over their reproductive years, and since premenstrual distress symptoms peak during the 4–7 days prior to menses, consistently symptomatic women may spend from 4 to 10 years of their lives in a state of compromised physical functioning and/or psychological well-being’’ (2007, p. 28). Therefore, in this chapter we address the burden of PMS/PMDD as a primary women’s health concern. We begin by providing the research/diagnostic criteria for PMS and PMDD, briefly address epidemiology and etiology, and continue with a discussion of the effects of PMS/PMDD on QOL including issues of stress. We conclude with an overview of treatment options with particular focus on behavioral medicine strategies such as exercise.
2
PMS and PMDD Diagnostic Criteria
The evolution of diagnostic criteria for PMS and PMDD has a confusing and controversial history that has led to frustration among scholars and caregivers who are unclear of what symptoms constitute either disorder. > Figure 115-1 provides a time line starting with initial clinical observations and moving through the establishment of research guidelines. Today, the tenth revision of the International Classification of Diseases (ICD-10) places PMS under ‘‘Diseases of the genitourinary system: Pain and other conditions associated with female genital organs and menstrual cycle’’ and labels it as Premenstrual Tension Syndrome (N94.3) (WHO: World Health Organization, 2004). Given that the ICD does not provide a minimum number of symptoms or functional impairment criteria required for a diagnosis of PMS, the American College of Obstetricians and Gynecologists (ACOG) published diagnostic guidelines in 2000 for PMS combining both the National Institute of Mental Health (NIMH) criteria and supportive research evidence (> Figure 115-2 summarizes these guidelines). Accordingly, a diagnosis of PMS may be made if symptoms include at least one of the somatic and affective symptoms listed, with a calculated 30% increase in symptom reports during the 6 days preceding menses compared to days 5–10 post-menses. These symptom pattern/severity changes need to be documented in a daily diary for 2–3 cycles for diagnosis. In addition, the severity of change must result in some life impairment. In other words, the magnitude of change has to be clinically meaningful and not simply represent a mathematical change which may be virtually imperceptible to the patient and therefore not be a hindrance to them. These guidelines also serve to distinguish PMS from premenstrual magnification of other disorders. Numerous symptom assessments exist and are summarized in > Table 115-1 and several examples are provided in the Appendix. PMS is a distinct diagnosis from PMDD which is identified by the ICD-10 as ‘‘Other mood [affective] disorders (F38)’’ (WHO, 2004). Diagnostic criteria for PMDD as they appear in the current Diagnostic and Statistical Manual for Mental Disorders (DSM-IV-TR) are provided in > Figure 115-3. Overlap in the symptoms listed for PMS and PMDD exist; however, with PMDD emphasis is placed on the first four symptoms listed, which are affective symptoms. Symptoms of PMDD are disabling in that they interfere with normal functioning and often lead women to seek treatment. In general, PMDD is seen as the most severe form of PMS inflicting the greatest amount of impairment on women’s functioning and perceived life quality.
1953
1954
115
Premenstrual Syndrome and Premenstrual Dysphoric Disorder
. Figure 115-1 Time line depicting the evolution of terminology and diagnostic criteria for PMS/PMDD. From the early writings of Hippocrates to the current diagnostic standards of the American College of Obstetricians and Gynecologists and the American Psychiatric Association, this timeline provides an overview of noteworthy figures and events in the history of what we now call PMS and PMDD. PMS, premenstrual syndrome; PMDD, premenstrual dysphoric disorder; OB/Gyn, medical doctor of obstetrics and gynecology; QOL, quality of life
Premenstrual Syndrome and Premenstrual Dysphoric Disorder
115
. Figure 115-1 (continued)
3
Epidemiology/Etiology
Prevalence estimates for PMS vary widely among reports. Factors contributing to this range are how the condition is defined and assessed (retrospective vs. prospective measures) and the different study populations investigated. According to recent epidemiological investigations using the current diagnostic criteria for PMS published by the ACOG, the prevalence of PMS among women in the United States (US) ranges from 19% (Strine et al., 2005) to up to 30% (Dean et al., 2006) with women in their late twenties and early thirties most likely to seek health care for their symptoms (Dell, 2004). The prevalence for PMDD is considerably less. Using prospective assessments and DSM-IV-TR diagnostic criteria in women of reproductive
1955
1956
115
Premenstrual Syndrome and Premenstrual Dysphoric Disorder
. Figure 115-2 Diagnostic criteria for premenstrual syndrome (PMS). These criteria allow healthcare providers to make a diagnosis of PMS by assessing the presence of an affective and somatic symptom and their cyclical nature. aSymptoms appear alphabetically and are not in order of importance or prevalence. PMS, premenstrual syndrome
age, four different studies report very similar findings. In a community sample in Munich, Germany, Wittchen et al. (2002) identified 6% with PMDD. The prevalence in US women is between 5% (Sternfeld et al., 2002) and 6% (Cohen et al., 2002), and 5% among Canadian women (Steiner and Born, 2000). It is worth noting that the prevalence difference between PMS and PMDD may be related to the self-report measures used to diagnose the two conditions. When the ACOG guidelines are used to diagnose PMS, women have the opportunity to report each symptom, including physical symptoms separately (> Figure 115-2). Conversely, the DSM-IV-TR criteria for PMDD tethers all physiological symptoms other than fatigue and appetite changes together as a single item (see item 11 in > Figure 115-3). Also, the presence of five symptoms is required for diagnosis. It may be that a woman suffers from all five of the physical symptoms listed in item 11 along with only one affective symptom. In this situation, she would meet the criteria for diagnosis of PMS but not for PMDD. On one hand, these strict criteria may prevent over diagnosing or pathologizing women. On the other hand, it may cause women who just miss the cut-off criteria to go without beneficial treatments. These difficulties surrounding diagnoses are further compounded by the fact that the etiologies of PMS/PMDD are unknown. Research which takes a biomedical approach has
Calendar of premenstrual experiences
Daily record of severity of problems
Daily symptom rating scale
Menstrual distress questionnaire
Menstrual symptom severity list
Premenstrual experience assessment
Premenstrual daily symptom diary
Premenstrual symptom diary
Premenstrual symptom tracker
DRSPa
DSR
MDQb
MSSL
PEAb
PDSD
PMSD
PMST
Instrument Full Name
COPE
Instrument Abbreviation
. Table 115-1 PMS/PMDD symptomatology assessments
NWHIC; available for use at http:// www.4woman.gov/faq/pms.htm
Thys-Jacobs et al., 1995; sample appears in the Appendix
Diary appears in its entirety in: Dickerson et al., 2003
Futterman et al., 1988; sample appears in the Appendix
Mitchell et al., 1991; sample appears in the Appendix
Moos, 1968a; available for purchase at: http://portal.wpspublish.com/portal/ page?_pageid = 53,112689&_dad = portal&_schema = PORTAL
Freeman et al. 1996; sample appears in the Appendix
Endicott et al. 2004; sample appears in the Appendix; available for download at: http://pmdd.factsforhealth.org/have/ dailyrecord.asp
University of California, San Diego; Department of Reproductive Medicine, H-813; Division of Reproductive Endocrinology; psychometrics available in: Mortola et al. 1990
Source
RM
RM
RM
SA
RM
RM/SA
RM
RM
RM
Repeated Measure (RM) vs. Single Assessment (SA)
Premenstrual Syndrome and Premenstrual Dysphoric Disorder
115 1957
Premenstrual symptoms screening tool
Premenstrual profile
Shortened premenstrual assessment form
PSST
PMP
SPAFb,c
Allen et al., 1991; sample appears in the Appendix
Campagne and Campagne 2007; sample appears in the Appendix
Tool appears in its entirety in: Steiner et al., 2003
Reid 1985; sample appears in the Appendix
Source
SA
RM
SA
RM
Repeated Measure (RM) vs. Single Assessment (SA)
This table lists many of the tools used to assess a woman’s premenstrual symptoms. Some of these tools are designed to be a one-time measure of symptoms as indicated by the abbreviation SA for single assessment. The others are daily diary methods of assessment (i.e., repeated measures abbreviated as RM). We have provided the actual tools (that are not copyrighted or that we received permission to reproduce) in the Appendix along with scoring information. For those that are available on-line, the web address is provided in the table a This is the updated Daily Rating Form by Endicott et al. (1986) b Multiple versions exist c This is the shortened version of the original 95-item Premenstrual Assessment form of Halbreich et al., 1982
Prospective record of the impact and severity of menstrual symptoms
PRISM
Instrument Full Name
115
Instrument Abbreviation
. Table 115-1 (continued)
1958 Premenstrual Syndrome and Premenstrual Dysphoric Disorder
Premenstrual Syndrome and Premenstrual Dysphoric Disorder
115
. Figure 115-3 Research criteria for premenstrual dysphoric disorder. To diagnose a woman with PMDD, a healthcare provider would begin by assessing the symptomatology listed in A and proceed with applying the instructions detailed in B through D. PMDD, Premenstrual Dysphoric Disorder
1959
1960
115
Premenstrual Syndrome and Premenstrual Dysphoric Disorder
postulated that while reproductive hormone release patterns are normal in women with PMS/PMDD, they may have heightened sensitivity to these hormonal changes (Halbreich, 2003; Halbreich and Monacelli, 2004). This biomedical model is partly supported by evidence from twin studies demonstrating a higher concordance of symptoms in monozygotic versus dizygotic pairs (Treloar et al., 2002). Further support comes from treatment success with pharmacological interventions such as serotonin specific reuptake inhibitors (SSRIs) and oral contraceptives (see Treatment section below). This therapeutic efficacy argues for the influence of putative neuroendocrine factors. Collectively, these findings have some scholars positing that women with a genetic predisposition for PMS/PMDD may experience endocrine/neuroendocrine responses in a more extreme manner that contribute to symptom expression (Mishell, 2005). Even still, the absence of a unified biological explanation has led to the development of multivariable models that include biopsychosocial factors.
4
Quality of Life
One such variable that relates to both diagnostic criteria and overall disease burden is QOL. QOL is a multidimensional construct that includes a person’s subjective judgment of their overall life experience. As a central measure in > Positive Psychology, QOL assessments direct attention away from the negative aspects of life and offer an alternative to traditional objective measures of success, social functioning, and/or life-satisfaction. When QOL assessments combine these subjective judgments of satisfaction with ratings of importance they allow for determination of congruence between desired and achieved life experiences. When the assessment of QOL is framed by a particular illness, it is termed health related quality of life (HR-QOL). As a research construct, no single definition of QOL exists. Instead, it is operationally defined by the tools used to measure it. The scientific study of QOL is a young science. In studying human behavior, the science of psychology has historically been concerned with dysfunction rather than health, positive functioning, or well-being. Relatedly, the subjectiveness of QOL reports has historically been unwelcome through the doors of medicine. Fortunately, the aims of modern medicine include fostering preventive mindsets and health maintenance behaviors in patients rather than providing only reactive and acute care. Even the most hard-lined biomedical approach, in light of these aims, must recognize the importance of assessing patients from a more global perspective. This includes assessing cognitive and affective burdens of a disease or disorder. This approach, along with the desire to assess patients’ self-reported/subjective impressions of their health, has led to a scholarly boom in the literature addressing QOL. Studies investigating HR-QOL in women with PMS/PMDD have typically used symptom checklists such as those designed for diagnosis. This is in part due to the fact that the diagnostic criteria for PMS/PMDD requires the use of daily checklists over consecutive cycles, and as such the burden of the illness is inherent in the diagnosis. For a diagnosis of PMDD, impairment in social or occupational functioning must be attributable to reported symptoms. These disease-specific HR-QOL measures therefore differ from non HR-QOL in two important ways. First, the use of symptom checklists leaves out many QOL domains affected by symptoms indirectly. For example, symptom checklists do not ask about the financial burden of PMS/PMDD manifested by increased doctor visits, increased drug costs, lost work days, and/or childcare costs. Second, by adding importance ratings, non HR-QOL measures allow for assessment of congruence between acquired and desired outcomes. These may be
Premenstrual Syndrome and Premenstrual Dysphoric Disorder
115
distinguishable from the disease or disorder in question by the person doing the reporting. For example, a women with PMS may find that her affective changes interfere with her relationships at work; however, she may not value those relationships and therefore she is unaffected by cyclically intermittent interference with them. Assessing these subjective aspects of disturbance without a measure of satisfaction and importance, as is done with a HR-QOL symptom checklist, would not allow for determination of such incongruence. Recently several researchers have investigated the financial burden of PMS/PMDD. Dean and Borenstein (2004) found that women with PMS reported more days missed at work and less work productivity than women without PMS, a pattern also found in women with PMDD (Robinson and Swindle, 2000). Chawla et al. (2002) also found that the economic burden of PMDD was due to reported decreases in productivity at work. Furthermore, these researchers found that as symptom severity increased so too did healthcare utilization including emergency room visits and visits to primary care physicians. Similarly, Borenstein et al. (2003) reported a financial burden of $500 US dollars over a 2 year period attributed to physician costs for PMS care. More recently, Borenstein et al. (2005) found the annual financial burden of absenteeism and decreased productivity while at work resulted in $4333 US dollars lost per patient, whereas the cost for healthcare increased only $59 US dollars. This implicates an indirect path (workplace difficulties and the resultant financial losses) as a more significant burden to women than the direct encumbrance of increased medical expenses. Another striking finding by Halbreich et al. (2003) is their reported estimate that the burden of PMDD over the reproductive years of those diagnosed is 3.8 years of disability. Thus, the cumulative effects of PMS/PMDD on a woman’s QOL are not restricted to what is covered in a physical examination or assessed by symptom checklists. Missed workdays and affected relationships continue to add stress long past the end of the symptomatic period. Thus, a true measure of burden should assess the affects of PMS/PMDD across various domains of life that are relevant to women in the 20–40 year old age range when PMS symptoms are most prevalent. The Frisch Quality of Life Inventory (QOLI: NCS Pearson Inc, 1994) is a 32-item scale that offers an overall QOL score as well as a weighted life satisfaction profile across 16 domains and has been used in women’s health research (Lustyk et al., 2004a). The QOLI was originally developed to provide a measure of positive mental health, and it is designed to take into account both cognitive and affective components of wellbeing. With a heuristic approach, the QOLI measures how individuals feel in terms of negative or positive affect, as well as how satisfied individuals are in terms of their cognitive appraisal of how well their needs are being met. As this tool is not public domain, we will briefly describe it here and refer interested readers to–http://www.pearsonassessments.com/tests/qoli.htm for ordering information. In completing this inventory, participants are instructed to rate the importance of, and their satisfaction with, 16 specific life domains including: Health, Self-esteem, Goals-andValues, Money, Work, Play, Learning, Creativity, Helping, Love, Friends, Children, Relatives, Home, Neighborhood, and Community. The resultant scores are the products of importance and satisfaction ratings for each domain. Domains that are rated as not important by the participants are not given a domain score, nor are they included in the overall QOL score. The QOLI is appropriate for use with adults (18-years and older). It is written at a sixthgrade reading level, and can be used in an interview format for those who cannot read or who are visually impaired. The measure was separately normed in racially and ethically diverse nonclinical samples. Psychometric tests found the QOLI to have adequate internal
1961
1962
115
Premenstrual Syndrome and Premenstrual Dysphoric Disorder
consistency, temporal stability, and convergent validity with other life satisfaction/QOL measures. Weighted satisfaction ratings for each domain have not been individually psychometrically validated. However, because these domains highlight areas relevant to women’s lives (e.g., career, family, and social networks), assessing them may aid in treatment planning and targeting specific treatment goals. The QOLI has also been used successfully as a > repeated measures tool in both longitudinal research designs and for charting clinical changes in patients (NCS Pearson Inc, 1994). The QOLI was used in a women’s health study investigating interrelationships among QOL, perceived stress, premenstrual symptomatology, and exercise (Lustyk et al., 2004a). Symptom severity was used to separate women into high and low symptomatology groups for comparisons. Not only did the more symptomatic women report more stress, they reported poorer QOL. Calculated values for each of the QOLI domains are provided in > Table 115-2. Of particular interest are the significant differences in the self-esteem, goals and values, and money domains as none of these aspects of QOL are assessed by symptom checklists. Such findings support a theoretical argument against using only symptom checklists to assess QOL in women with PMS/PMDD as they leave important variables unassessed. Perhaps the best strategy is to use a symptom specific assessment along with a global, non-health related QOL assessment. In 2006, Sarah Gehlert and her colleagues developed the Women’s Quality of Life Scale designed specifically for healthy women of reproductive age (Gehlert et al., 2006). We provide this tool in its entirety along with scoring instructions in > Figure 115-4. This groundbreaking tool, which provides a gender specific non HR-QOL assessment, may actually serve to measure QOL in women with PMS/PMDD. In the development of the questionnaire, respondents were queried on their perceived importance of each item. Furthermore, all items were evaluated by experts in women’s health research further bolstering their semantic validity. > Factor analyses used to identify items of importance revealed four 10-item domains assessing physical, mental, social, and spiritual health. This work is of particular note because it was based on a large multi-ethnic sample of 1,207 women from both rural and urban dwellings that represent a broad socioeconomic status range. The resultant tool is particularly easy to read, use, and score. Time will tell if its psychometric properties hold for the investigation of QOL in women with PMS/PMDD.
5
Symptomatology and Stress
When the NIMH provided research criteria for PMS in 1983, (> Figure 115-1) an era of investigations into the biopsychosocial concomitants of the condition began. The research acumen of Nancy Fugate-Woods and colleagues contributed much to our understanding of the interplay among perceived/psychological stress, stressful life events, and physiological stress with premenstrual symptomatology (Woods et al., 1982, 1997, 1998). While this work was performed before the new diagnostic criteria for PMS were published (ACOG, 2000), it remains noteworthy given its superior methodological and analytic approach. Furthermore, Mitchell, Woods, and Lentz (1991) are credited with bringing to the fore the importance of criterion-based symptom severity assessments. Such assessments address the clinical relevance of symptom changes across the cycle in women which may ultimately impair their QOL. Yet, to underscore the importance of understanding the interrelationships among stress and symptomatology in women diagnosed under current criteria, studies reported here for
115
Premenstrual Syndrome and Premenstrual Dysphoric Disorder
. Table 115-2 Quality of life scores for women with low and high PMS Quality of Life Value Low Mean PMS (n = 38) SD High Mean PMS (n = 44) SD P value1 Overall raw score
3
.92
2
1.5
.01
Subscales: Healtha b
3
2.7
2
3.1
.45
3
2.8
1
3.6
.03
Goals & Valuesc
3
2.5
3
3.1
.94
d
Money
1
1.6
1
1.8
.29
Worke
2
2.2
1
2.6
.29
Playf
4
2.0
3
2.8
.05
Learning
3
2.1
2
2.3
.32
Creativityh
2
2.1
2
2.6
.92
Helpi
1
2.1
2
2.6
.54
j
Love
2
3.5
2
3.4
.96
Friendsk
4
2.4
4
2.7
.68
Kidsl
2
3.0
2
2.8
.61
Home
3
2.3
2
2.8
.50
Neighborhoodn
2
2.4
2
2.3
.93
Community
2
2.2
2
2.6
.18
Relativesp
4
2.0
4
2.2
.56
Self Esteem
g
m
o
This table shows the mean QOL scores for women with high or low levels of premenstrual symptoms. The means reflect the overall QOL score and the scores for the individual 16 domains. Scores are a product of ratings of importance and overall satisfaction. This generates weighted scores that range from 6 (extremely important and very dissatisfied) to +6 (extremely important and very satisfied). Scores above zero are in the satisfied range and indicate increasing importance and satisfaction as they approach +6 a Physical fitness, free of illness, disability or pain b Self approval c Desired accomplishments and matters of importance d Adequate earnings and goods at present and future projections e Activities in and out of home or school where one spends most of their time f Leisure time activity g Knowledge acquisition h Using imagination to come up with solutions to problems or engaging in a hobby i Assisting those in need j Intimate romantic relationship k Non-relative, close relationships l Importance of having/not having a child or one’s happiness and relationship with children m Importance and satisfaction with one’s dwelling n Importance and satisfaction with area surrounding one’s dwelling o Importance and satisfaction with the city of one’s dwelling p Relationships with those one is related to: SD, standard deviation; n, sample size; PMS, premenstrual syndrome; QOL, Quality of Life 1
Group differences assessed by T-test. Source: Lustyk et al. (2004) Women & Health 39: 35–44. Reprinted with permission from Haworth Press, Inc., http://www.haworthpress.com/web/WH. Article copies available from the Haworth Document Delivery Service: 1–800-HAWORTH. E-mail address: [email protected]
1963
1964
115
Premenstrual Syndrome and Premenstrual Dysphoric Disorder
. Figure 115-4 The women’s quality of life questionnaire. This questionnaire allows for the assessment of QOL in women. To score, a Items are reverse scored. [P] items contributes to the Physical domain, [Y] items contribute to the Psychological domain, [C] are Social domain items, and [S] are Spiritual domain items. Domain scores reflect endorsement totals weighted one point each. After reverse scoring, larger values are indicative of higher QOL
Premenstrual Syndrome and Premenstrual Dysphoric Disorder
115
PMDD will be dated 1994 or later and 2000 or later for PMS (see > Figure 115-1 for overview of this history). The morbidity of PMS or PMDD may be augmented by stress, and stress-related illnesses may be co-morbid with PMS and PMDD. Stress is a multifaceted construct comprising biopsychosocial and spiritual components. > Stressors are those stimuli perceived as threatening that lead to the stress response. These range from uncontrollable environmental challenges or threats such as war and natural disasters to self-generated mental anguish such as ruminative thinking. The stress response involves activation of the sympathetic nervous system with subsequent increased release of epinephrine (i.e., adrenaline) from the adrenal medulla in an adaptive attempt to deal with the stressor. Additionally, whether acute or sustained, stress activates the > hypothalamic-pituitary-adrenal (HPA) axis with resultant release of the glucocorticoid > cortisol. This latter response affects the hypothalamic and > pituitary influences over the menstrual cycle. The interactions among the HPA axis and the menstrual cycle are depicted and summarized in > Figure 115-5. Co-morbidity with stress related illnesses is evidenced by reports that women with menstrual problems are more likely to suffer from anxiety, nervousness, and restlessness than asymptomatic women (Strine et al., 2005). In an epidemiological analysis of PMDD in a large community sample, Perkonigg et al. (2004) found increased rates of posttraumatic stress disorder among women with PMDD. This provides further evidence for co-morbid stress related illnesses in women with menstrual problems. Stress may augment PMS/PMDD symptomatology. In their investigation of 114 women divided into sub-samples with high and low-symptom reports, Lustyk et al. (2004a) found significantly more perceived stress in the high symptom group compared to the low symptom group. Yet, as this study was not longitudinal in design, we can not be sure if stress preceded the symptoms reported, or vice versa. While findings such as these may suggest a bi-directional relationship among perceived stress and symptomatology, more recent studies assessing multivariable models in which perceived stress serves as a mediator argue for its influential effects on symptomatology. In two separate studies investigating the mediating role of perceived stress, Lustyk et al. found that perceived stress partially mediated the relationships of abuse history (2007) and spiritual well-being (2006) with premenstrual symptomatology. While it seems intuitive that stress can affect premenstrual symptoms, the counter argument remains—premenstrual symptoms may serve as stressors with the potential of creating a negative feedback loop that further exacerbates symptoms. Support for the latter idea comes from laboratory studies where cardiovascular responses to cognitive and/or physical stressors are assessed in women with varying degrees of symptomatology. Current investigations using up-to-date diagnostic criteria are few (i.e., diagnostic criteria from 1994 for PMDD and 2000 for PMS). In one study of this type, Girdler et al. (1998) demonstrated that women with PMDD had significantly greater peripheral resistance and norepinephrine reactivity in response to a mental stressor (i.e., serial addition test) compared to control women. These researchers noted similar response patterns during both > follicular cycle phase and luteal cycle phases. Conversely, Epperson et al. (2007) recently demonstrated a significantly greater acoustic startle response during the luteal phase compared to the follicular phase in women diagnosed with PMDD using 2 months of daily diary reporting. These cycle-related inconsistencies may be due to an unassessed neuroendocrine relationship, specifically the role of estrogens in the stress response. In a more naturalistic study, Pollard et al. (2007) assessed cardiovascular responses to real life stressors by having women journal about their perceived stress, take their own heart rate and blood pressure, and provide a urine
1965
1966
115
Premenstrual Syndrome and Premenstrual Dysphoric Disorder
. Figure 115-5 Stress, HPA-axis, HPO-axis interactions. The HPO axis regulates the menstrual cycle. GnRH is released in pulsatile fashion from the hypothalamus causing the production and release of LH/FSH from the pituitary. In turn the ovary (ies) release(s) E & P in varying quantities affecting target tissue throughout the body (e.g., uterus). Cortisol, released during stress, can inhibit GnRH, LH, and E release. It can also decrease sensitivity of target tissue to E. ACTH, adrenocorticotropic hormone; CRH, corticotropin-releasing hormone; E, estrogen; FSH, follicle stimulating hormone; GnRH, gonadotropin-releasing hormone; HPA, hypothalamic-pituitary-adrenal; HPO, hypothalamic-pituitary-ovary; LH, leutinizing hormone; P, Progesterone
sample for estrogen analyses. These researchers sought to determine the specific role of estrogen in the stress responses of premenopausal women. Their analyses revealed a positive association between heart rate responses and perceived stress on days when estrogen levels were high (days 11–13 of the cycle) as opposed to the negative association between these variables when estrogen was low (days 4–6 of the cycle). Collectively considered with findings from women with PMS/ PMDD, there is no clear and reliable cyclical pattern and as such additional research is needed to elucidate these neurendocrine responses in women with PMS/PMDD.
Premenstrual Syndrome and Premenstrual Dysphoric Disorder
6
115
Treatments
The compounding effects of stress and the unknown etiologies for PMS/PMDD have created roadblocks in the development of validated therapies effective for all symptom patterns and all symptomatic women. This subsequently has created enormous treatment challenges for health care providers. Both PMS and PMDD are diagnoses of exclusion since no consistent biological markers or laboratory tests exist that verify their onset or presence. Currently, the ACOG and American Academy of Family Physicians (AAFP) recommend lifestyle change as the first-line treatment option for PMS (> Table 115-3). However, the most frequent course of treatment is oral contraceptives and/or SSRIs. While it appears that the ACOG recommendation is generally being ignored, it may simply be that meeting this treatment standard proves too difficult given the dearth of information supporting evidence-based behavioral interventions that are successful in producing long lasting lifestyle changes. Also worth noting are latency to treatment effects, which are consistently shorter with pharmacological interventions. According to Steiner et al. (2003), most women with PMDD require SSRIs to successfully manage their symptoms. While side effects to SSRIs exist (e.g., insomnia, gastrointestinal disturbances, decreased libido), research suggests that the improved QOL with managed symptoms in women with PMDD outweighs these concerns (Freeman, 2005). Therefore, while our focus here is on behavioral management, we provide a summary of current pharmacological treatment options in > Table 115-4. We have included in this table information on nonprescription diet aids as some have been shown to have pharmacological effects. In a recent report, Freeman (2005) writes: ‘‘Results of a survey that examined the diagnosis and treatment of PMS in the US indicated that only one in four physicians provided adequate treatment for PMS’’ (p. 440). This inadequacy is likely influenced by the lack of validated treatment modalities specific for PMS. If the physician’s goal is to use evidenced-based treatments, than it is not surprising that the pharmaceutical route is chosen over the firstline recommendation of the ACOG since consistent supportive evidence from well-controlled clinical trials for behavioral treatments is lacking. Even still, this trend is interesting since drug options that are available are Food and Drug Administration (FDA) approved for PMDD not PMS (see > Table 115-4). It may be that the combination of unclear diagnostic criteria and the weak evidence base from research are leading physicians to prescribe medications off-label. Additionally, the evidence in support of dietary changes and/or supplement implementation is mixed, with efficacy reports coming from less methodologically rigorous studies than the clinical trials seeking FDA approval for pharmaceuticals. Poor study quality likely contributes to the general lack of understanding of the pharmacological effects of many ‘‘natural’’ diet aids, yet the fact that such effects exist makes one wonder if they belong under the non-pharmacological first-tier of treatment. Behavioral interventions also lie in this first tier of treatments. Of those behavioral medicine options empirically studied, PMS/PMDD treatment efficacy has been demonstrated with aerobic exercise, cognitive-behavioral therapy (CBT), and systematic relaxation with considerably more support for aerobic exercise. While much of this research was based on the now dated diagnostic criteria, a few factors contributing to the therapeutic efficacy are worth noting. For example, there are different latency to treatment effects for these behavioral medicine modalities. Aerobic exercise seems to exert some positive affects while the person actually engages in the activity. CBT on the other hand can take months for improvements to occur. Given that the treatment recommendation guidelines are to proceed to the next tier of
1967
1968
115
Premenstrual Syndrome and Premenstrual Dysphoric Disorder
. Table 115-3 Hierarchical treatment guidelines for PMS and PMDD Tier 1: Nonpharmacologic treatments Lifestyle change such as aerobic exercise, diet modification and education (e.g., reduce salt and caffeine intake or add OTC supplements such as magnesium), rest, therapy If physical symptoms predominate, treat the specific symptoms Bloating–spironolactone Breast tenderness–danazol, evening primrose oil, spironolactone luteally, vitamin E, chaste berry fruit Fatigue/Insomnia: education on sleep hygiene, alter caffeine consumption Headaches: OTC pain reducers such as NSAIDs, acetaminophen. Tier 2: If patient is diagnosed with PMDD, psychotropic therapy with FDA approved SSRIs (continuous or intermittent therapy). Tier 3: If the above approaches fail, manipulate the cycle with hormone therapy (e.g., OCs, GnRH agonist therapy). This table summarizes the current treatment recommendations of the American College of Obstetricians and Gynecologists. These are meant to guide healthcare providers (HP) in making evidence-based treatment choices. HPs are advised to start at Tier 1 and proceed to the next tier if improvements are not observed within 2–4 cycles. If a patient is diagnosed with PMDD, however, HPs can start at Tier 2 as FDA approved drug interventions now exist Key: PMS, Premenstrual Syndrome; PMDD, Premenstrual Dysphoric Disorder; OTC, over the counter; NSAIDs, nonsteroidal anti-inflammatory drugs; FDA, Food and Drug Administration; SSRIs, serotonin specific reuptake inhibitors; OCs, Oral contraceptives; GnRH, Gonadotropin Releasing Hormone Note: Compiled in part from these sources: ACOG Practice Bulletin #15 April 2000; Campagne and Campagne, 2007; Johnson, 2004
treatment if symptom improvements are not noted within 2–4 cycles, the long latency to treatment effect for CBT may reduce its therapeutic applicability. Additional theoretical support for the use of aerobic exercise as a therapeutic modality comes from the ease with which a regimen can be prescribed. Without specialized training or certification as an exercise instructor/trainer, healthcare providers can safely recommend an aerobic program (e.g., brisk walking) that adequately elevates heart rate for a sufficient period of time with a low risk of iatrogenic harm and subsequent injury. This is not the case for CBT, which requires administration by a trained therapist and as such requires referral from the diagnosing physician. Even though PMDD is a mental disorder, the diagnosis is predominately made by primary care physicians (e.g., doctors of Obstetrics & Gynecology) and as such CBT for PMS/PMDD requires additional healthcare coverage or, if healthcare insurance limitations exist, the assumption of costs by the patient. Correlational evidence demonstrates a positive interrelationship between maintaining an aerobic exercise program and QOL reports (e.g., Lustyk et al., 2004a, b). Yet, the absence of controlled clinical trials employing aerobic exercise as a sole behavioral medicine intervention for women diagnosed with PMS/PMDD under current criteria is a serious limitation in establishing evidenced based therapeutic regimens. Exercise intervention studies that
Uncontrolled studies and medical opinions suggest efficacy is possible
Some randomized trials, and retrospective studies indicating efficacy is possible
Some randomized trials, and retrospective studies indicating efficacy is likely
Some randomized trials, and retrospective studies indicating efficacy is likely
Some randomized trials, and retrospective studies indicating efficacy is likely
Some randomized trials, and retrospective studies indicating efficacy is likely
Randomized, double-blind, placebo controlled trials with affirmed efficacy
Not approved for PMS, used off-label for PMS
Clinical trial research for PMDD leading to FDA approval
Not approved for PMS, used off-label for PMS
Clinical trial research for PMDD leading to FDA approval
Clinical trial research for PMDD leading to FDA approval
Type of Supporting Research
This table lists the drugs that are FDA approved for PMDD and the supplements or diet aids that have a treatment effect. Along with the dose range, we provide a general overview of the kind of research that was performed to test the treatment. Randomized, controlled, clinical trials are the most rigorous and well-designed studies for investigating the efficacy of a treatment. They involve having participants assigned to different treatment conditions in a non-systematic way to impose control over various person factors, like motivation to be in the study for example. Those performed over time with prospective measures are considered more rigorous in design than those that use retrospective measures where participants are asked to recall how they felt on some prior date. Uncontrolled studies lack rigor and carry much less weight than those previously described. When a treatment receives FDA approval it is deemed safe and effective having undergone the most rigorous research Key: Schering/Bayer trade name for Drospirenone & Ethinyl Estradiol (EE). SSRIs, Serotonin Specific Reuptake Inhibitors; Mg, milligrams; FDA, Food and Drug Administration; PMS, Premenstrual Syndrome; PMDD, Premenstrual Dysphoric Disorder Source: Dell, DL (2004). Premenstrual Syndrome, Premenstrual Dysphoric Disorder, and premenstrual exacerbation of another disorder. Clinical Obstetrics and Gynecology, 47: 568–575 Adapted with permission from Lippincott, Williams, & Wilkins
40 mg twice daily
4–20 mg/day
Chasteberry
Black Cohosh
50–100 mg/day
Vitamin B6
80 mg/twice daily
400–800 mg/day
Magnesium
300 mg three times/day
1200–1600 mg/day
Calcium
St. John’s wort
25–50 mg/day
Sertraline
Ginkgo
10–20 mg/day
3 mg Drospirenone;.02 mg EE
Dose
Fluoxetine
SSRIs
Oral Contraceptive (Yaz)
Medication/Supplement
. Table 115-4 Current pharmacological treatments for PMS/PMDD including supplements recommended as therapeutically effective
Premenstrual Syndrome and Premenstrual Dysphoric Disorder
115 1969
1970
115
Premenstrual Syndrome and Premenstrual Dysphoric Disorder
employed prospective controlled designs are now dated (e.g., Prior et al., 1987, Steege and Blumenthal, 1993). Most recently Stoddard et al. (2007) performed a cross-sectional study comparing symptom reports in sedentary women and regular exercisers. Unfortunately, PMS/ PMDD diagnoses were not made but rather the Moos Menstrual Distress Questionnaire was used to assess subclincial levels of premenstrual distress. This tool provides a broad list of negative symptoms without allowing women to indicate change in severity from whatever is normal for them, a strong criticism of this type of measure (Halbreich et al., 1982). Additionally, the cross-sectional nature of this study makes it questionable whether exercise can serve as a successful behavioral treatment for improving and/or preventing symptoms in women diagnosed with PMS/PMDD under current criteria. It may be that exercise will prove to be therapeutic for women with PMS/PMDD by decreasing stress and/or improving QOL. While data exists supporting the beneficial role of exercise on stress and QOL (e.g., Lustyk et al., 2004b), the specific effects in women diagnosed with PMS/PMDD under current criteria are not known. Given that decreased HR-QOL is inherent in the diagnosis of these conditions, assessing improvements in this measure seems important for determining treatment efficacy. Yet, as pointed out earlier, it would also be beneficial to assess more QOL domains even if they are only secondarily related to symptoms. It is curious that exercise is recommended for PMS/PMDD with so little supportive evidence. The 2000 ACOG Practice Bulletin states: ‘‘Although the evidence base is modest at this time, aerobic exercise can be recommended to all women with PMS because of its numerous other health benefits (p. 4).’’ How these other health benefits are expected to help these women is not offered. Furthermore, it is unknown what role, if any, aerobic exercise has on worsening symptoms. For women who may already feel poor, aerobic exercise may exacerbate those feelings. It seems intuitive that premenstrual women with swollen tender breasts, abdominal bloating, headaches, and/or negative affect will have low motivation for aerobic exercise. Furthermore, if their symptoms include fatigue and/or stress, a woman may sense having too little energy for aerobic exercise. One commonly reported barrier to exercise is lack of time. This may be particularly noteworthy for women. In the US 46% of the total workforce are women with 38% holding jobs that are professional in nature (USDLWB, 2007). Furthermore, PMS/PMDD occurs during the reproductive years when many women are adding the roles of wife and mother to their existing jobs outside of the home. These time limitations for self-care along with the stress imposed from the fast paced daily demands known to the working mother likely complicate symptomatology. Adding an exercise schedule to the mix may well increase overall stress rather than reduce it. As stress is associated with increased symptomatology, the addition of exercise may produce a negative cycle of further symptom complaints. Growing interest in complementary and alternative medical (CAM) treatments for physical and mental illnesses has led to the exploration of options historically ignored by mainline medicine. As of late, treatments such as acupuncture, reflexology, and massage have gained empirical support for their efficacy in reducing premenstrual symptom severity. Yet again, as with exercise, time limitations may apply since each of these remedies requires treatment from a trained professional. An unfortunate limitation to some CAM therapies is that they are not covered or are only minimally covered by medical insurance plans, which potentially add financial stress as well.
Premenstrual Syndrome and Premenstrual Dysphoric Disorder
115
These issues provide a theoretical argument for the investigation of sedentary self-care for women with PMS/PMDD. Since new diagnostic criteria for these conditions were published, this area of behavioral medicine is a neglected field of study. For example, what effects, if any, rest, mental imagery training or meditation have on symptoms, perceived stress, and/or QOL in women diagnosed with PMS/PMDD under current criteria is/are unclear. As with other areas of behavioral medicine previously discussed, many sedentary self-care options empirically studied to date have investigated effects using now dated diagnostic criteria and/or involved methodologies that lack the control and rigor needed to appropriately assess treatment effects. One exception is a study by Hernandez-Reif et al. (2000) in which the effects of massage therapy on PMDD symptomatology were compared to systematic relaxation effects. After 5 weeks of 30-min sessions performed twice per week by both groups of women, those that received massage demonstrated reductions in anxiety and somatic symptoms. These improvements were not observed with the sedentary self-care option (i.e., relaxation). As a measure of HR-QOL, these researchers assessed the behavioral change domain of the Menstrual Distress Questionnaire (Moos, 1968) revealing no change in reported decreases in school or job performance, and the avoidance of social activities in either treatment group. It is important to note that only women with the most severe form of PMS (i.e., PMDD) were included in this study and as such additional research is needed to investigate the potential benefits of systematic relaxation on symptoms in women with PMS diagnosed under the current criteria. Time will tell if such sedentary self-care options will have therapeutic efficacy including improved QOL for these women.
Summary Points Up to 90% of all women report negative symptomatology during the premenstrual phase
of the menstrual cycle. In 1994, PMDD research criteria were added to the DSM. In 2000, the ACOG published diagnostic and treatment guidelines for PMS. PMS and PMDD are associated with poor QOL in women. PMS and PMDD are associated with increased perceived stress. The burden of PMS/PMDD includes missed days at work, increased health care costs, and negative affects on relationships. FDA approved treatments for PMDD include SSRI’s and the oral contraceptive Yaz. Behavioral changes, diet adjustment and/or supplementation are the recommended firstline treatment options for PMS. FDA approved pharmacological treatment options for PMDD are reportedly used ‘‘off label’’ to treat PMS. Exercise may improve premenstrual symptomatology by reducing stress and improving QOL. Prospective randomized controlled-clinical trials assessing the therapeutic efficacy of exercise for PMS/PMDD under current diagnostic guidelines are lacking. Investigations of sedentary self-care options for PMS/PMDD are needed.
1971
1972
115
Premenstrual Syndrome and Premenstrual Dysphoric Disorder
Premenstrual Syndrome and Premenstrual Dysphoric Disorder
115
1973
1974
115
Premenstrual Syndrome and Premenstrual Dysphoric Disorder
Premenstrual Syndrome and Premenstrual Dysphoric Disorder
115
References ACOG. (2000). Practice Bulletin. Obstet Gynec. 15: 1–9. Allen SS, McBride CM, Pirie PL. (1991). J Reprod Med. 36: 769–772. Borenstein JE, Dean BB, Endicott J, Wong J, Brown C, Dickerson V, Yonkers KA. (2003). J Reprod Med. 48: 515–524. Borenstein JE, Chiun-Fang C, Dean BB, Wong J, Wade S. (2005). J Occup Environ Med. 47: 26–33. Campagne DM, Campagne G. (2007). Eur J Obstet Gynec Reprod Biol. 130: 4–17. Chawla A, Swindle R, Long S, Kennedy S, Sternfeld B. (2002). Med Care. 40: 1101–1112. Cohen LS, Soares CN, Otto MW, Sweeney BH, Liberman RF, Harlow BL. (2002). J Affect Disord. 70: 125–132. Dean BB, Borenstein JE. (2004). J Occup Environ Med. 46: 649–656. Dean BB, Borenstein JE, Knight K, Yonkers K. (2006). J Womens Health. 15: 546–555. Dell DL. (2004). Clin Obstet Gynec. 47: 568–575. Dickerson LM, Maxyck PJ, Hunter MH. (2003). Am Family Physician. 67: 1743–1752. Endiccott J, Nee J, Cohen, J, Halbreich U. (1986). J Affect Disord. 10: 127–135. Endiccott J, Nee J, Harrison W. (2005). Arch Womens Ment Health. 9: 41–49. Epperson CN, Pittman B, Czarkowski KA, Stiklus S, Krystal JH, Grillon C. (2007). Neuropsychopharmacology. 32: 2190–2198. Freeman EW. (2005). Pharmacoeconomics. 23: 433–444. Freeman EW, DeRubeis RJ, Rickels K. (1996). Psychiatry Res. 65: 97–106. Frisch MB. (1994). Quality of Life Inventory: Manual and Treatment Guide. NCS Pearson, Minneapolis, MN. Futterman LA, Jones MF, Miccio-Fonseca LC, Quigley T. (1988). Psychol Rep. 63: 19–34. Gehlert S, Chang C-H, Bock RD, Hartlage SA. (2006). J Clin Epidemiology. 59: 525–533. Girdler SS, Pedersen CA, Straneva PA, Leserman J, Stanwyck CL, Benjamin S, Light KC. (1998). Psychiatry Res. 81: 163–178. Halbreich U, Endicott J, Schacht S, Nee J. (1982). Acta psychiat scand. 65: 46–65. Halbreich U. (2003). Psychoneuroendocrinology. 28: S55–99. Halbreich U, Borenstein J, Pearlstein T, Kahn LS. (2003). Psychoneuroendocrinology. 28: S1–23. Halbreich U, Monacelli E. (2004). Primary Psychiatry. 11: 33–40. Hernandez-Reif M, Martinez A, Field T, Quintero O, Hart S, Burman I. (2000). J Psychosomatic Obstet Gynec. 21: 9–15. Johnson SR. (2004). Obstet Gynec. 104: 845–859. Lustyk MKB, Widman L, Paschane A, Ecker E. (2004a). Women Health. 39: 35–44.
Lustyk MKB, Widman L, Paschane A, Olson KC. (2004b). Behav Med. 30: 124–131. Lustyk MKB, Beam C, Miller A, Olson KC. (2006). J Psychol Theology. 34: 311–317. Lustyk MKB, Widman L, de Laveaga L. (2007). Women Health. 46: 61–80. Mishell DR. (2005). Am J Managed Care. 11: S473–479. Mitchell ES, Woods NF, Lentz MJ. (1991). In: Taylor DL, Woods NF (ed.) Menstruation, Health, and Illness: Series in health care for women. Taylor & Francis, New York, pp. 89–102. Moos RH. (1968). Psychosom Med. 30: 853–867. Mortola JF, Girton L, Beck L, Yen SSC. (1990). Obstet Gynaec. 36: 302–307. Perkonigg A, Yonkers KA, Pfister H, Lieb R, Wittchen H-U. (2004). J Clin Psychiatry. 65: 1314–1322. Pollard TM, Pearce KL, Rousham EK, Schwartz JE. (2007). Am J phys Anthrop. 132: 151–157. Prior JC, Vigna Y, Sciarretta D, Alojado N, Schulzer M. (1987). Fert Steril. 47: 402–408. Reid RL. (1985). Curr Problems Obstet Gynec Fertil. 8: 1–57. Robinson RL, Swindle RW. (2000). J Womens Health. 9: 757–768. Steege JF, Blumenthal JA. (1993). J Psychosom Res. 37: 127–133. Steiner M, Born L. (2000). CNS Drugs. 13: 287–304. Steiner M, Macdougall M, Brown E. (2003). Arch Womens Mental Health. 6: 203–209. Sternfeld B, Swindle R, Chawla A, Long S, Kennedy S. (2002). Obstet Gynec. 99: 1014–1024. Stoddard JL, Dent CW, Shames L, Bernstein L. (2007). Eur J Appl Physiol. 99: 27–37. Strine TW, Chapman DP, Ahluwalia IB. (2005). J Womens Health. 14: 316–323. Thys-Jacobs S, Alvir JMJ, Fratarcangelo P. (1995). Psychopharmacology Bull. 31: 389–396. Treloar SA, Health AC, Martin NG. (2002). Psychol Med. 32: 25–38. United States Department of Labor, Women’s Bureau (USDLWB). Available at http://www.dol.gov/wb/ factsheets/Qf-laborforce-04.htm. Retrieved December 8, 2007. Wittchen HU, Becker E, Lieb R, Krause P. (2002). Psychol Med. 32: 119–132. Woods NF, Dery GK, Most A. (1982). J Human Stress. 8: 23–31. Woods NF, Lentz M, Mitchell ES, Heitkemper M, Shaver J. (1997). Res Nursing Health. 20: 329–340. Woods NF, Lentz MJ, Mitchell ES, Shaver J, Heitkemper M. (1998). Res Nursing Health. 21: 129–142. WHO. (2004). The ICD-1- Classification of Mental, Behavioral and Developmental Disorders. 10th revision [2nd Ed.] Geneva, Switzerland.
1975
116 Quality of Life and Infertility A. Montazeri 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1978
2 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8
Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1978 Findings from the Netherlands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1979 Findings from United Arab Emirates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1980 Findings from the USA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1980 Findings from Canada . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1980 Findings from Austria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1980 Findings from Italy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1980 Findings from Brazil . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1981 Findings from Iran . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1981
3
Instruments used to Measure Quality of Life . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1982
4
Irrational Parenthood Cognitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1982
5
Male, Female Factors and Quality of Life . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1982
6
Concluding Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1983 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1985 Appendix: Assisted Reproductive Therapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1985
#
Springer Science+Business Media LLC 2010 (USA)
1978
116
Quality of Life and Infertility
Abstract: Health related quality of life is an important issue for infertile couples. A review was conducted to examine the literature. The review covered all full publications that have appeared in the English language biomedical journals. The search strategy included a combination of key words, ‘‘quality of life’’, ‘‘> infertility’’, and ‘‘infertile’’ in title. A total of ten citations were identified and examined in this review. The major findings are summarized and presented in several headings including findings from each individual study and instruments used to measure quality of life in people with infertility problems. There were quite a few studies that reported on health related quality of life in infertile couples. It seems that there is need to conduct more studies on the topic using valid and standard measures. List of Abbreviations: ICSI, Intracyto-plasmic Sperm Injection; IPC, scale: The Irrational Parenthood Cognitions scale; IVF, > in-vitro fertilization; PCOS, polycystic ovarian syndrome; PCOSQ, the Polycystic Ovary Syndrome Questionnaire
1
Introduction
Health related quality of life is now considered to be an important outcome measure in many clinical settings. Quality of life can indicate the directions that are needed to treat different patient populations more efficiently. It is argued that for many people infertility is a major life crisis (Leiblum and Geenfiled, 1997). In addition, since infertility and its treatment have several psychosocial effects on infertile couples, studying health related quality of life in this group of people is very crucial (Greil 1997). Infertility can cause depression, anxiety, social isolation and sexual dysfunction (Chen et al., 2004; Fassino et al., 2002; Irvine, 1996). Undoubtedly the clinical efforts and technology have improved outcomes in infertile couples. However, the nature of the disease and its treatment remind us that, although the treatment of infertility is important, the patients’ quality of life is equally important. This mini-review of the literature has aimed to investigate the extent to which quality of life studies have contributed to the reproductive health and fertility literature. It was hoped that this review would contribute to the existing knowledge, indicate achievements and discrepancies, help both researchers and clinicians to have a better insight into the topic, and consequently aid in improving quality of life in infertile couples. A review of the literature on quality of life and infertility was carried out using the MEDLINE search engine. The intention was to review all full publications that have appeared in the English language biomedical journals during 1982–2008. The year 1982 was chosen because the first study on quality of life and infertility was published in 1982. The search strategy included a combination of key words ‘‘quality of life,’’, infertility, and infertile in title. It was thought that this might help to carry out a more focused investigation. This provided the initial database for the review. The initial search was carried out in early 2007 and updated twice in October and November 2007 and once for the final check in March 2008.
2
Findings
A total of 9 citations were identified, 5 citations containing the key words ‘‘quality of life and infertility’’ and 4 citations containing quality of life and infertile in their title. Of these, one study (Leiberman and Insler, 1982) was excluded since neither the paper nor its abstract was
Quality of Life and Infertility
116
available for the review. However, instead studies by Fekkes et al., (2003) and Rashidi et al. (2008) were included in this review, in all ten studies. The results are shown in > Table 116-1. Here, the major findings are summarized and presented in the following sections.
. Table 116-1 A list of studies on quality of life and infertility Author(s)
Year of publication
Country
Study sample
Main focus
Fekkes et al.
2003
The 447 women + 425 Netherlands men
Quality of life in couples planning IVF treatment
Khayata et al.
2003
United Arab 269 infertile women Factor influencing quality of Emirates life
Monga et al.
2004
USA
18 infertile couples + 12 couples seeking sterilization
Impact of infertility on quality of life, marital adjustment, and sexual function
El-Messidi et al.
2004
Canada
3 groups of 50 couples
Effects of repeated treatment failure on quality of life
Schmid et al.
2004
Austria
49 women
Infertility caused by polycystic ovarian syndrome and quality of life
Ragni et al.
2005
Italy
1,000 couples
IVF and quality of life
Chachamovich et al.
2007
Brazil
177 women
Factors predicting quality of life
Ghasemzad and Farzadi
2007
Iran
192 women
Quality of life and its correlates
Shindel et al.
2008
USA
121 infertile couples Sexual function and quality of life in male partner
Rashidi et al.
2008
Iran
514 women and 514 men
Comparing quality of life in women and men and causes of infertility
Studies on quality of life and infertility in different countries IVF. in-vitro fertilization
2.1
Findings from the Netherlands
The study finding showed that young men and women (aged 21–30 years) planning in-vitro fertilization (IVF) had more short-term social and emotional problems than people of the same age group in the general population. No substantial differences were found in cognitive and physical functioning for all age groups of men or women planning IVF compared with the general population. A high level of irrational parenthood cognition accounted for a less optimal score on all the different domains of quality of life (Fekkes et al., 2003).
1979
1980
116 2.2
Quality of Life and Infertility
Findings from United Arab Emirates
Parameters that most affected quality of life in women attending the assisted reproduction clinic were mood related issues mainly in women above 30 years of age, with primary and female factor infertility (Khayata et al., 2003).
2.3
Findings from the USA
There were two studies from the USA. A study of 18 infertile couples and 12 couples seeking elective sterilization indicated that 83% of infertile couples reported feeling societal pressures to conceive. The marital adjustment score compared to control was lower for infertile women but this was not the case for men. In addition lower quality of life score was noted for women but not in the man (Monga et al., 2004). In another study assessing sexual function and quality of life in the male partner of infertile couples, Shindel et al. (2008) reported that depression, erectile dysfunction and sexual relationship problems were prevalent among male partners of infertile couples and male partners reported significantly lower standardized scores on mental health compared to normative data.
2.4
Findings from Canada
The study found that quality of life score for control group (a couple with least one child and no history of infertility) was higher than the scores reported by the couples with repeated failure infertility treatment and the couples who never attempted any medical treatment. In addition the study results showed that there were no significant differences in quality of life scores between the couples with repeated failure treatment and the couples who never attempted any medical treatment or between male and female partners (El-Messidi et al., 2004).
2.5
Findings from Austria
Studying a group of native Austrian and Moslem immigrant women the study showed that in term of polycystic ovarian syndrome (PCOS) the typical heterogeneity of PCOS could be found in both subgroups with no differences. However, health-related quality of life of women with Islamic background was affected to a greater degree than that of Austrian women, although no differences in symptoms were observed (Schmid et al., 2004).
2.6
Findings from Italy
Male quality of life scores as measured by the SF-36 was higher than female. Duration of infertility and previous IVF treatment significantly affected health related quality of life. However, performing a robust statistical analysis the study findings showed that there were no significant differences between quality of life of infertile couples and the normal population both for men and women (Ragni et al., 2005).
Quality of Life and Infertility
2.7
116
Findings from Brazil
The study found that some dependent variables could predict quality of life in infertile women. The analysis showed that some independent variables such as age (for better general health and physical functioning), previous reproductive tract surgery (for worse general health but higher environmental scores), advanced education (for higher vitality, mental health, and environmental health but worse for social relationships) and perception of worse sexual life (for lower overall quality of life score) were predicting factors of quality of life in Brazilian women experiencing infertility (Chachamovich et al., 2007).
2.8
Findings from Iran
A study of 192 Iranian infertile women showed that 12% of the study sample had poor and nearly half had good quality of life. The study findings did not show any significant correlations between age, length of marital life, type of infertility, or duration of treatment time. However the study indicated that there was an inverse correlation between irrational parenthood cognitions and quality of life scores (Ghasemzad and Farzadi, 2007). Another study from Iran carried out by Rashidi et al. (2008) with a relatively large sample of 514 and 514 infertile women and men and found that there were significant differences between women and men indicating that male patients scored higher on the SF-36 (> Table 116-2). Quality of life was found to be better in older infertile couples, those with higher level of education and economic status, in working women and in infertility due to male factor. They concluded that infertile couples are at risk of a sub-optimal quality of life and recommended that psychological counseling, especially psychotherapy support must be provided for patients in order to improve their quality of life.
. Table 116-2 The SF-36 scores by gender (scores ranging from 0 to 100 with higher scores indicating a better condition) Sub-scales
Male (n = 514) Mean (SD)
Female (n = 514) Mean (SD)
P < 0.0001
PF
86.7 (20.9)
80.6 (21.8)
RP
78.3 (31.0)
72.0 (33.3)
0.002
BP
77.4 (16.8)
69.1 (20.2)
< 0.0001
GH
71.0 (17.1)
66.1 (18.1)
< 0.0001
SF
76.4 (21.9)
72.6 (22.8)
0.007
RE
73.0 (35.6)
65.4 (38.6)
0.001
VT
66.2 (17.6)
59.2 (18.5)
< 0.0001
MH
67.2 (17.8)
61.6 (19.6)
< 0.0001
Source: Rashidi et al., Vali-e-Asr Reproductive Health Research Center, Tehran University of Medical Sciences, Tehran, Iran The figures show that men had a better condition on all quality of life subscales as measured by the SF-36 PF physical functioning; RP role physical; BP bodily pain; GH general health; SF social functioning; RE role emotional; VT vitality; MH mental health
1981
1982
116 3
Quality of Life and Infertility
Instruments used to Measure Quality of Life
Several instruments were used to measure quality of life in infertility research. One study (Fekkes et al., 2003) used the Hopkins Symptom Checklist, the Sickness Impact Profile, the Irrational Beliefs Inventory, and the Irrational Parenthood Cognitions scale. The Italian study (Ragni et al., 2005) measured quality of life using the Short Form Health Survey (SF-36). The Austrian investigators (Schmid et al., 2004) used the Polycystic Ovary Syndrome Questionnaire (PCOSQ), a questionnaire developed to measure the health-related quality of life of women with polycystic ovary syndrome (Cronin et al., 1998; Jones et al., 2004). This is a 26-items questionnaire measuring five areas of health related quality of life: emotions (8 items), body hair (5 items), weight (5 items), infertility problems (4 items) and menstrual problems (4 items). The Brazilian study (Chachamovich et al., 2007) used the SF-36 and the shorter version of the World Health Organization Quality of Life Questionnaire (WHOQOLBREF). Monga et al. (2004) reported that they used the Quality of Well-being Scale (selfadministered version), the Locke Wallace Marital Adjustment Test, and the Brief Index of Sexual Functioning for Women and the International Index of Erectile Function for men in their study in the USA. Shindel et al. (2008) indicated that they used the SF-36, the Center for Epidemiological Studies Depression Inventory, and for male partners International Index of Erectile Function and Self-Esteem and Relationship Quality scale and for female partners the Female Sexual Function Index and a version of the Self-Esteem and Relationship Quality Scale modified for women. However, others (Khayata et al., 2003; El-Messidi et al., 2004; Ghasemzad and Farzadi, 2007) used ad hoc instruments to measure quality of life in this population.
4
Irrational Parenthood Cognitions
One important aspects of studying quality of life in infertile couples relates to irrational parenthood cognitions. As suggested patients with high levels of irrational parenthood cognitions are at risk of less optimal quality of life and thus as suggested a short cognitive counseling therapy is essential for patients with high levels of irrational ideas (Fekkes et al., 2003). It is argued that irrational ideas can lead to the development of psychological and emotional distress. The Irrational Parenthood Cognitions (IPC) scale was developed by Fekkes et al. (2003). It is a short instrument that contains 14 statements about parenthood ideas and respondents rate their ideas on a five-point scale raging from ‘‘I agree totally’’ to ‘‘I disagree totally’’; for example respondents indicate to what extent they agree or disagree with the statements such as ‘‘A life without children is useless and empty’’ or ‘‘Not having children causes lifelong suffering.’’ The measure gives a score ranging from 0 to 56; with higher scores indicating a stronger need to have children in order to live a happy life.
5
Male, Female Factors and Quality of Life
There are several reasons for infertility. Broadly speaking, the causes of infertility could be due to male factor, female factor, both or unexplained. Studies have shown that infertile women report more psychological distress and lower quality of life than infertile men (Newton et al., 1999; Oddens et al., 1999). One reason for such findings is due to the fact that in general
Quality of Life and Infertility
116
women usually rate their quality of life lower than male gender. Another explanation is that women are blamed (or some times they take the blame) more frequently for the couples’ infertility and thus, the stigma associated with such blaming (regardless of the diagnosis) causes more distress and deteriorations in quality of life in female partners (Peronace et al., 2007). A study showed that infertility was associated with a higher than expected incidence of erectile dysfunction, depressive symptoms and dysfunctional sexual relationship in infertile men (Shindel et al., 2008). It has been suggested that men diagnosed with male factor infertility experience more suffering than men with infertility due to other causes (Nachtigall et al., 1992). However, a study investigated physical and mental health in a sample of 265 infertile men prior and after 12 months of unsuccessful treatment according to their diagnosis: unexplained, female factor, male factor or mixed and showed that when treatment was not successful, all men regardless of diagnosis demonstrated similar increased suffering in the form of decreased mental health, increased physical stress reactions, decreased social support and increased negative social stress over time. As suggested these findings indicate that involuntary childlessness is difficult for all men and is not dependent on with whom the cause lies (Peronace et al., 2007; > Table 116-3). . Table 116-3 The SF-36 scores by infertility causes in a sample of infertile couples (514 women and 514 men, scores ranging from 0 to 100 with higher scores indicating a better condition) Female factor Sub-scales (n = 233) Mean (SD)
Male factor (n = 578) Mean (SD)
Both (n = 82) Mean (SD)
Unexplained (n = 135) Mean (SD)
P
PF
80.5 (23.8)
85.0 (20.5)
86.6 (17.1)
81.3 (23.5)
0.01
RP
71.4 (33.8)
77.0 (31.2)
78.3 (31.3)
71.8 (34.6)
0.06
BP
72.6 (18.9)
72.9 (19.6)
73.2 (18.4)
76.0 (17.4)
0.36
GH
66.1 (17.9)
69.5 (17.5)
68.3 (20.9)
69.0 (16.3)
0.09
SF
69.4 (23.8)
76.4 (21.6)
77.2 (22.7)
73.8 (22.3)
0.001
RE
63.9 (39.3)
70.8 (36.6)
73.5 (35.4)
69.3 (37.3)
0.07
VT
60.5 (19.2)
64.1 (17.8)
61.7 (18.7)
61.4 (18.7)
0.05
MH
60.7 (20.7)
65.8 (18.1)
66.2 (20.2)
63.9 (17.7)
0.005
Source: Rashidi et al., Vali-e-Asr Reproductive Health Research Center, Tehran University of Medical Sciences, Tehran, Iran Quality of life scores as measured by the SF-36 in infertile couples. The table indicates that according to the infertility causes there are significant differences among infertile couples in physical functioning, social functioning and mental health subscales PF physical functioning; RP role physical; BP bodily pain; GH general health; SF social functioning; RE role emotional; VT vitality; MH mental health
6
Concluding Remarks
This review of the literature indicated that there is need to conduct more studies in quality of life in infertile women or infertile men or for both. This, however, indicates a requisite condition that emerges from having standard instruments to measure quality of life in these
1983
1984
116
Quality of Life and Infertility
groups of population. This review indicated that only one study (Schmid et al., 2004) used a specific measure and the remaining papers reported that either used general measures such as the SF-36 and the WHOQOL-BREF or they synthesized an ad hoc instrument tailored to their study objectives. At present the PCSQ for infertile women (Jones et al., 2004) and the TLMK [an instrument for measuring quality of life in male patients with involuntary childlessness] (Schanz et al., 2005) are among the recent existing instruments for measuring quality of life in infertile couples. The TLMK is a 35-item questionnaire with 4 subscales: desire for a child (13 items), sexual relationship (7 items), gender identity (8 items) and psychological wellbeing (7 items). Response categories of this instrument range from 1 to 5 on a Likert scale with the score of 1 representing a low and 5 a high level of agreement with the item. The higher overall scores indicate a lower quality of life (Schanz et al., 2005). Factors that predicting quality of life may vary in different infertile populations, different gender and different ethnic backgrounds. Thus, the identification of factors associated with better or worse quality of life in different domains is vital in order to propose and test scientifically based interventions for infertile populations (Chachamovich et al., 2007). It is argued health professionals should be sensitive to the ethnicity, religious and cultural background of their patients to provide the best possible medical support when treating infertility (Schmid et al., 2004). In addition, it should be noted that different treatment modalities might result in different quality of life outcome. Thus the results from different treatments should not be generalized to all other treatment types. For example in the case of IVF treatment the treatment itself might not affect the quality of life but duration of infertility and failure to achieve pregnancy through IVF might have a negative impact on patients’ quality of life (Ragni et al., 2005). This was a review of the literature with a limited objective. It is quite possible to suggest that the other search strategies would have been resulted in more citations. This review only focused on quality of life, infertility and infertile keywords in title. In addition only the MEDLINE search engine was examined. Thus still there is a potential to carry out a similar review with much more extensive search strategy using different search engines. Finally, despite the existence of an extensive body of the literature on psychological aspects of the infertility, this mini-review, although limited in many aspects, indicates that there are quite a few studies on quality of life in reproductive health including infertility literature. Indeed, there is need to conduct more studies on the topic using valid and standard measures. Worldwide, couples view infertility as a tragedy which carries social, economic and psychological consequences. In 1965, the 18th World Health Assembly recognized that under the auspices of family planning, building a family should be the free choice of the individual couple In 2006, the UN General Assembly adopted the Secretary-General’s report recommending the inclusion of the target to achieve universal access to reproductive Health under the Millennium Development Goal 5, Improve maternal health The Department of Reproductive Health and Research recognizes that infertility is an unmet need in family planning in both the developed and developing world. As published in our joint WHODHS Comparative Report in 2004, based on data evaluated up to mid-2002, one in four evermarried women of reproductive age in most developing countries are infertile because of primary or secondary infertility. The Department works to develop, support and establish protocols for infertility prevention, diagnosis and management. For more information and resources visit the following website Source: World Health Organization, Reproductive Health http://www.who.int/reproductive-health/publications/infertility.html
Quality of Life and Infertility
116
Summary Points Generally, the infertility might be due to male factor, female factor, mixed factor or unexplained factor.
Infertility and its treatment have a considerable effect on couples’ quality of life. Infertile women usually report more deterioration in quality of life compared to infertile men.
Involuntary childlessness is difficult for all men and is not dependent on the causes of infertility.
Factors that predicting quality of life may vary in different infertile populations, different gender and different ethnic backgrounds.
The treatment of infertility itself might not affect quality of life. The treatment duration and failure to have children are important factors that negatively associated with poorer quality of life.
Appendix: Assisted Reproductive Therapy Indication
Procedure
In Vitro Fertilization (IVF)
Tubal factor severe endometriosis unexplained infertility male factor
Involves controlled ovarian hyperstimulation, which is aimed at producing multiple oocytes. Once the oocytes are mature, hCG (human chorionic gonadotropin) is administered and 34–36 h later, they are retrieved under ultrasound guidance with the patient under light general anesthesia. The oocytes are then combined with sperm in a Petri dish to allow for fertilization. The embryos are incubated in growth medium and then transferred back into the female partner’s uterus 3–5 days later.
Cryoembryo Transfer
Indicated for patients who have undergone a cycle of IVF in which excess eggs were cryopreserved.
In this procedure, the excess cryopreserved fertilized embryos from the previous IVF may be transferred at a later time. The advantages of this procedure are that a repeat ovarian stimulation can be avoided. In addition, this procedure allows a woman with advanced maternal age to use embryos that were fertilized with oocytes from when she was younger.
1985
1986
116
Quality of Life and Infertility
Appendix (continued) Indication
Procedure
IVF with Donor Premature ovarian failure Oocytes perimenopause, or menopause failed IVF due to oocyte factors (comprises 50% of cases)
The donor may either be anonymous or selected by the couple. A legal contract is needed between the donor and the recipient couple prior to initiation of the procedure. Insurers do not cover the payment to the donor and screening of the donor.
Intracytoplasmic Sperm Injection (ICSI)
Congenital absence of the vas deferens obstructive and non-obstructive azoospermia or men with less than one million total motile sperm previous vasectomy
ICSI involves direct injection of a single sperm into the cytoplasm of an oocyte. Success has been reported even with immotile and immature sperm. Success rates are the same as those reported for IVF, approximately 35% per embryo transfer.
Gestational Carrier
Women without a uterus women with a medical condition that preclude carrying a pregnancy to term male homosexual couples
Involves IVF (see above) with transfer of the embryos to a gestational carrier, which is a woman with a uterus who will carry the pregnancy to term. To avoid custody lawsuits, when oocytes are needed, use of a separate oocyte donor, (i.e., an individual who is different from the gestational carrier) is recommended. This is particularly important for male homosexuals or for women who lack functional ovaries or uterus, or who have a medical contraindication to pregnancy.
Source: Brigham and Women’s Hospital (a teaching affiliate of Harvard Medical School) Web address: http://www. brighamandwomens.org/reproductivemedicine/med_info/crm_treatment-mi.aspx#top
References Chachamovich JR, Chachamovich E, Zachia S, Knauth D, Passos EP. (2007). Hum Reprod. 22: 1946–1952. Chen TH, Chang SP, Tsai CF, Juang KD. (2004). Hum Reprod. 19: 2313–2318. Cronin L, Guyatt G, Griffith L, Wong E, Azziz R, Futterweit W, Cook D, Dunaif A. (1998). J Clin Endocrinol Metab. 83: 1976–1987. El-Messidi A, Al-Fozan H, Lin Tan S, Farag R, Tulandi T. (2004). J Obstet Gynaecol Can. 26: 333–336. Fassino S, Piero A, Boggio S, Piccioni V, Garzaro L. (2002). Hum Reprod. 17: 2986–2994. Fekkes M, Buitendijk SE, Verrips GHW, Braa DDM, Brewaeys AMA, Dolfing JG, Krotman M, Leerentveld RA, Macklon NS. (2003). Hum Reprod. 18: 1536–1543.
Ghasemzad A, Farzadi L. (2007). Med Sci Monit. 13: CR313–CR317. Greil AL. (1997). Soc Sci Med. 45: 1679–1704. Irvine SCE. (1996). Sexual Marital Ther. 11: 273–280. Jones GL, benes K, Clark TL, Denham R, Holder MG, Haynes TJ, Mulgrew NC, Shepherd KE, Wilkinson VH, Singh M, Balen A, Lashen H, Ledger WLL. (2004). Hum Reprod. 19: 371–377. Khayata GM, Rizk DE, Hasan MY, Ghazal-Awad S, Asaad MA. (2003). Int J Gynaecol Obstet. 80: 183–188. Leiberman JR, Insler V. (1982). Med Law. 1: 33–38. Leiblum SR, Greenfield DA. (1997). In: Leiblum SR (ed.) Infertility: psychological issues and counseling strategies. Wiley, New York, USA, pp. 83–102. Monga M, Alexandrescu B, Katz SE, Stein M, Ganiats T. (2004). Urology 63: 126–130.
Quality of Life and Infertility Nachtigall RD, Becker G, Wozny M. (1992). Fertil Steril. 57: 113–121. Newton CR, Sherrard W, Glavac I. (1999). Fertil Steril 72: 54–62. Oddens BJ, den Tonkelaar I, Nieuwenhuyse H. (1999). Hum Reprod. 14: 255–261. Peronace LA, Boivin J, Schmidt L. (2007). J Psychosom Obstet Gynecol. 28: 105–114. Ragni G, Moscom P, Baldini MP, Somigliana E, Vegetti W, Caliari H, Nicolosi AE. (2005). Hum Reprod. 20: 1286–1291. Rashidi B, Montazeri A, Ramezanzadeh F, Shariat M, Abedinia N, Ashrafi M. (2008). Health-related
116
Quality of Life in Iranian Couples Receiving IVF/ ICSI Treatment. Vali-e-Asr Reproductive Health Research Center, Tehran University of Medical Sciences, Tehran, Iran. Schanz S, Baeckert-Sifeddine IT, Braeunlich C, Collins SE, Batra A, Gebert S, Hautzinger M, Fierlbeck G. (2005). Hum Reprod. 20: 2858–2865. Schmid J, Kirchengast S, Vytiska-Binstofer E, Huber J. (2004). Hum Reprod. 19: 2251–2257. Shindel AW, Nelson CJ, Naughton CK, Ohebshalom M, Mulhall JP. (2008). J Urol 179: 1056–1059.
1987
117 Speech Determines Quality of Life Following Total Laryngectomy: The Emperors New Voice? P. Farrand . R. Endacott 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1990
2
Total Laryngectomy: Etiology, Epidemiology and Clinical Outcomes . . . . . . . . . . . . 1991
3
Treatment Following Total Laryngectomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1992
4 4.1 4.2 4.3
Methods of Voice Restoration Following Total Laryngectomy . . . . . . . . . . . . . . . . . . . 1992 Electrolarynx . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1993 Esophageal Speech . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1993 Tracheoesophageal Speech (TEP) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1994
5
Voice Quality Determines Quality of Life: The Emperors New Voice? . . . . . . . . . . . 1994
6
Research Examining Voice Quality and Quality of Life . . . . . . . . . . . . . . . . . . . . . . . . . 1995
7
Methods of Voice Restoration and Quality of Life . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1996
8
Considerations for Practice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1999 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1999
#
Springer Science+Business Media LLC 2010 (USA)
1990
117
Speech Determines Quality of Life Following Total Laryngectomy
Abstract: > Total laryngectomy, performed for advanced laryngeal carcinoma, results in loss of the vocal chords and a permanent > neck stoma. To help restore speech various voice restoration methods – > tracheoesophageal speech (TEP), > electrolarynx and > esophageal speech – are commonly employed. Whilst different methods exist however, TEP produces the best voice quality, is assumed to maximize patient quality of life and unsurprisingly is most commonly adopted by medical and health professionals. Although voice quality is clearly of importance however, each method is associated with varied costs and benefits that impact upon other areas related to quality of life in total laryngectomy patients. Following a discussion concerning the etiology, epidemiology and treatment for laryngeal carcinoma, this chapter reviews the literature related to health-related quality of life amongst total laryngectomy patients employing different voice restoration methods. It argues that, although important, voice quality alone does not determine quality of life following total laryngectomy. Consideration should therefore be given to the impact that each voice restoration method has upon all areas of quality of life, rather than focus upon voice quality alone. Implications of these conclusions for clinical practice are discussed at the end of the chapter. List of Abbreviations: HNQOLQ, University of Michigan head and neck quality of life questionnaire; > SF-36, health specific quality of life measure; TEP, tracheoesphageal speech
1
Introduction
Total laryngectomy is a surgical procedure that involves removal of the entire larynx and is most commonly used to treat patients with advanced-stage laryngeal carcinoma (see > Table 117‐1 for the key features of total laryngectomy). Although necessary to extend life, the procedure places many demands upon patients and is associated with lower quality of life compared to techniques which preserve the larynx (LoTempio et al., 2005; Terrell et al., 1998). Unfortunately laryngeal preservation is not always possible, and as a consequence, attention needs to be directed towards maximizing the quality of life of patients undergoing total laryngectomy. This chapter examines the evidence related to patient quality of life following total laryngectomy. A particular focus is to challenge an often held assumption that speech following total laryngectomy actually dictates quality of life and as such, quality of life is maximized solely by employing a voice restoration method that delivers the best voice quality. . Table 117‐1 Key features of a laryngectomy Laryngectomy – key features
The term laryngectomy is used to describe removal of the larynx
In most cases, laryngectomy is performed to remove cancer of the larynx
The most common initial symptom for patients with laryngeal cancer is a sore throat and change in voice quality
Prior to removal of the larynx, biopsies are taken to confirm the presence of cancer
For some patients laryngectomy follows other treatments, most commonly radiotherapy
As part of the laryngectomy procedure, the neck is re-molded to allow separation of the airway and esophagus (gullet)
Speech Determines Quality of Life Following Total Laryngectomy
2
117
Total Laryngectomy: Etiology, Epidemiology and Clinical Outcomes
Laryngeal carcinoma is the eleventh most prevalent form of cancer in men world-wide (Marioni et al., 2006) and is most commonly caused by smoking or alcohol use (Devlin and Langer, 2007). As can be seen in > Figure 117‐1 within the UK incidence increases with age up to 80 years old for both males and females, but to a much greater extent in men (Office for National Statistics, 2007). Early stage tumors have an 80–90% survival rate which drops to 60% with more advanced forms (Marioni et al., 2006). There is debate about which types of surgery and/or treatment should be used (Devlin and Langer, 2007; Marioni et al., 2006), with areas such as molecular biology showing promise for future treatment directions (Forastiere et al., 2006). A retrospective, > longitudinal study of 158,000 cases in the US showed that survival rates decreased in the mid 1990s, with the authors suggesting that the impact of changes in treatment (with reduced surgical management) require further investigation (Hoffman et al., 2006). The main emphasis in patients with laryngeal carcinoma is a multi-disciplinary approach from the outset to ensure all options are carefully considered. . Figure 117‐1 The Rate Per 100,000 of the UK Population of Larygeal Cancer for Males and Females, 2004 (Office for National Statistics, 2007). National Statistics website: ww.statistics.gov.uk. Crown copyright material is reproduced with the permission of the Controller Office of Public Sector Information. The incidence of larygeal cancer in UK for males and females in 2004
1991
1992
117
Speech Determines Quality of Life Following Total Laryngectomy
The end result for all patients undergoing total laryngectomy is division of the respiratory and digestive pathways and formation of a permanent neck stoma. Other outcomes from the procedure fall into two groups: firstly, complications that can arise during or after the procedure, for example, hemorrhage, fistula formation or infection (Hall et al., 2003; Herranz et al., 2000). Whilst these can be prolonged, depending on the co-morbid state of the patient (Herranz et al., 2000), they are mostly reversible. Second, long term effects such as > dysphagia (Ward et al., 2002), olfactory nerve damage (Risberg-Berlin et al., 2007; Sinkiewicz et al., 2006), voice loss and communication problems (Armstrong et al, 2001) which can persist and may be permanent. In addition general health (SF-36; see Jenkinson et al., 1993) scores have been noted to be lower for patients following total laryngectomy than for those with other serious medical conditions (Armstrong et al., 2001). In addition to the effects from surgery, patients often suffer reactions to concomitant treatments such as radiation or chemotherapy (Pillon et al., 2004). These treatments, in turn, have also been noted to increase the risk of complications. For example, a meta-analysis of observational studies examining fistula formation found that pre-operative radiotherapy was associated with increased severity and duration of pharyngocutaneous fistula (Relative Risk 2.28 (1.59–3.25) p < 0.001) (Paydarfar and Birkmeyer, 2006). Total laryngectomy is sometimes performed following failure of other treatments such as radiotherapy with outcomes no worse in this group of patients (Hall et al., 2003). However, the extent of surgery performed during total laryngectomy (e.g., inclusion of neck dissection) has been linked to increased complication rate (Herranz et al., 2000). Different types of surgery (partial, subtotal or total laryngectomy) result in variation in the extent of voice loss; for the purposes of this chapter, we focus on total laryngectomy only.
3
Treatment Following Total Laryngectomy
Following total laryngectomy, treatments focus on preventing/managing any recurrence of the malignancy and relieving the long terms effects identified above. The impact of loss of smell on quality of life (Sinkiewicz et al., 2006) is reflected in the efforts expended to optimize smell rehabilitation (Risberg-Berlin et al., 2007; Sinkiewicz et al., 2006). Similarly, the impact of dysphagia following laryngectomy on levels of disability, handicap and well-being has been examined (Ward et al., 2002) with authors emphasizing the need for management and further investigation of this effect (Armstrong et al., 2001; Pillon et al., 2004; Ward et al., 2002). Of particular note is the finding in Pillon and colleagues (2004) that the second commonest response by patients to dysphagia was to reduce food intake. Despite the potential for total laryngectomy to have a negative impact on patients across a number of quality of life dimensions however, the main focus in relieving the long term effects of total laryngectomy is on restoring speech through the employment of a voice restoration method.
4
Methods of Voice Restoration Following Total Laryngectomy
Ability to communicate following total laryngectomy has come a long way since patients were dependent upon writing words down onto a tablet or simply mouthing words for others to
Speech Determines Quality of Life Following Total Laryngectomy
117
interpret. Considerable progress has been made with the development of three main methods of voice restoration – electrolarynx, esophageal speech and tracheoesphageal speech (TEP) for patients to regain some ability to reproduce speech. Each method is associated with a specific range of costs and benefits discussed below (for a review see Blom, 2000; Finizia and Bergman, 2001; Morris et al., 1992), which most likely determine patient quality of life. A summary of each method is provided in > Table 117‐2.
. Table 117‐2 Advantages and disadvantages of methods of voice restoration Advantages Electrolarynx
Esophageal speech
Immediate voice restoration
Disadvantages An intraoral device may be needed
Easy to learn
Poor quality of speech
Minimal ongoing maintenance
Difficult to comprehend over the phone
No artificial appliance required
Quality of speech can be husky, low pitched and rough
No ongoing maintenance
Speech production may not be fluent Very difficult to learn Tracheoesophageal speech (TEP)
Highest quality of voice Natural sounding speech within 2 weeks of surgery
Requires valve replacement every 6 months Risk of valve failure Risk of discomfort from complications
4.1
Electrolarynx
Speech is performed with the aid of a handheld or artificial larynx with patients utilizing an electrolarynx method of voice restoration. With the neck form of electrolarynx, the device is placed against the side of the neck, under the chin or on the cheek. Sound is conducted into the device and speech is articulated. However not all patients are able to generate enough sound conduction through the skin, and on these occasions an intraoral device is recommended. With this form of electrolarynx a small tube is placed in the oral cavity and the generated sound is articulated. The electrolarynx form of voice restoration supports immediate voice restoration following surgery, is easy to learn and needs little ongoing maintenance. However on the negative side the quality of speech produced is poor, producing a mechanical and genderless sounding voice often difficult to comprehend face to face and particularly over the phone.
4.2
Esophageal Speech
Esophageal speech is performed by patients forcibly injecting air into their esophagus through a form of gulping. This air is then immediately expelled from the esophagus amounting to
1993
1994
117
Speech Determines Quality of Life Following Total Laryngectomy
something akin to a forceful and controlled belch with the consequent vibration articulated into speech. The sound quality of the voice is variable but often reported to be husky, rough and low pitched. Additionally speech is also characterized by sounds of short duration with speech discontinuous due to small amounts of air being available within the esophagus to power voice production. Whilst the method initially requires the surgical development of a > ‘‘pseudoglottis’’ following laryngectomy, little ongoing medical treatment is required. Patients who master this method often acquire an effective method of voice and speech requiring no artificial appliances or ongoing maintenance. Unfortunately however it is very difficult to learn, with at best only 60% of patients successful in acquiring speech (Blom, 2000).
4.3
Tracheoesophageal Speech (TEP)
Within this form of voice restoration, speech is dependent upon a > tracheoesophageal puncture which creates a tract in the wall separating the trachea and esophagus. A catheter is then inserted into this tract, which acts as a stent. The length of the puncture tract is measured and an appropriate sized one-way valved prosthesis is selected and placed into the tract several days following surgery. The one-way valve allows air to pass into the esophagus whilst preventing food and liquid entering the trachea. The prosthesis therefore allows air from the lung to pass into the esophagus, with the lung air vibrating the pharyngoesophageal segment to produce sound. This technique is widely felt to produce more natural sounding speech within 2 weeks of surgery (Pou, 2004) with a significant consensus in the literature that it also produces the highest voice quality of all voice restoration methods (Drummond et al., 1996; Farrand and Duncan, 2007; Robbins et al., 1984). Improved voice quality however is only achieved at the cost of increased inconvenience to the patient. The method requires that the patient meet regularly with health professionals to replace the flap valve every 6 months, alongside the possibility of valve failure due to fungal infection, salivary leakage and the prosthesis being dislodged. All of these medical complications can cause severe discomfort and require further medical treatment (Blom, 2000; Ferrer Ramı´rez et al., 2001).
5
Voice Quality Determines Quality of Life: The Emperors New Voice?
The prevailing discussion has highlighted that each voice restoration method is associated with a range of specific costs and benefits. It has also been noted that no specific method had an advantage across all the criteria related to clinical acceptability associated with methods of voice restoration identified by Blom (1998) (see > Table 117‐3). Whilst some of these criteria are clearly related to medical considerations alone, others such as swallowing, voice quality and use of a prosthetic valve that prevents stenosis and aspiration are likely to impact on quality of life. Unfortunately there is no single voice restoration method that can deliver the best quality of life in all of these areas. Across voice restoration methods therefore, there seems to be a trade off between voice quality and the need for continued ongoing medical management. TEP produces a much improved voice quality but at the expense of regular hospital visits and the possibility of the need for future treatment (Blom, 2000). Electrolarynx is associated with a reduced need for ongoing medical care, but at the expense of a significantly reduced voice quality and, in the case of esophageal
Speech Determines Quality of Life Following Total Laryngectomy
117
. Table 117‐3 Criteria related to clinical acceptability associated with methods of voice restoration for voice restoration methods (adapted from Blom, 1998) 1. Normal swallowing without aspiration 2. Reliable good quality voice 3. No oncologic compromise 4. Surgical simplicity 5. Universally consistent reproducibility 6. Inclusion of an uncomplicated, cost-effective prosthetic valve to prevent stenosis and aspiration 7. Viability in irradiated tissues used to assess the impact of each method These criteria are used by clinicians to determine whether a voice restoration method is appropriate for an individual patient
speech, a difficulty in learning to use the method (Blom, 2000). In the absence of medical justification for employing a specific voice restoration method therefore, it might be expected that choice should be based upon a consideration of the extent to which patients prioritize an improved voice quality over continued hospital treatment or visa-versa. This however does not seem to be the case. Regardless of the specific demands each voice restoration method places upon patients, there seems to be an inherent assumption that TEP should be considered the procedure of choice for restoration of speech after laryngectomy (Clements et al., 1997). Such an assumption has potentially helped to establish it as the method most commonly adopted by medical staff (Blom, 2000; Ferrer Ramı´rez et al., 2001; Finizia et al., 1998b; Schuster et al., 2003). This situation is likely to be primarily based upon the significant advantages associated with TEP with respect to several areas related to voice quality (Drummond et al., 1996; Robbins et al., 1984) leading to the assumption that improved quality of life is associated with better voice quality. On occasions such an assumption also seems to have led some authors to reach conclusions that go beyond the actual data reported. For example, quality of life differences between laryngectomy patients using TEP and a healthy population were restricted to physical functioning and physical and emotional role dimensions of the SF-36 (Schuster et al., 2003). This led the authors to conclude that the improved voice quality afforded by TEP prevented more widespread differences arising on the other SF-36 dimensions. However on the basis of the actual data alone it is premature to reach such a conclusion. Rather than believing the failure to find widespread quality of life differences as being determined solely by improved voice quality afforded by TEP, potentially other factors associated with TEP, or indeed voice restoration methods in general, accounted for the data.
6
Research Examining Voice Quality and Quality of Life
Overall there seems to be much in the literature to highlight the inherent assumption amongst medical staff that improved voice quality may not just be associated with patient quality of life, but may actually determine it. Indeed the conclusion that verbal communication may actually predict quality of life was drawn over two decades ago (e.g., see McNeil et al., 1981). Before the development of the TEP voice restoration method this assumption even led to the proposal that life without a voice would have such negative consequences upon a patients quality of life,
1995
1996
117
Speech Determines Quality of Life Following Total Laryngectomy
that even advanced disease should be treated with irradiation at the expense of prolonged survival (Harwood and Rawlinson, 1983). However, from the research literature there seems to be very little evidence to support this proposition. This is probably best highlighted by findings reported across a number of studies that have compared quality of life in head and neck cancer patients receiving various forms of treatment or surgical intervention (each having a different impact upon voice quality). For example, although patients report higher speech intelligibility, voice quality and speech acceptability following radiotherapy (Finizia et al., 1998a,b), studies have consistently failed to report a quality of life advantage for radiotherapy compared to laryngectomy (Deleyiannis et al., 1999; DeSanto et al., 1995; List et al., 1996b; Terrell et al., 1998) or compared to those having total laryngectomy and employing TEP (Finizia and Bergman, 2001). Furthermore, although quality of life differences have been reported between patients undergoing total, near-total and partial laryngectomy, differences were explained by the varied physical limitations imposed by the stoma rather than by voice quality. Specifically, patients having a permanent stoma were significantly less adjusted across all quality of life domains than those who had no stoma (DeSanto et al., 1995). Eadie and Doyle (2004) took a different approach to examining the contribution of voice quality to quality of life. In this study the voice quality of total laryngectomy patients employing TEP was rated by non-clinicians on a number of factors related to its auditoryperceptual quality. These ratings were then correlated with patient quality of life as measured by one of the several head and neck specific measures, called the University of Michigan Head and Neck Quality of Life instrument (HNQOLQ; see Taylor et al., 2004). Commonly used head and neck specific quality of life measures can be seen in > Table 117‐4 (for a systematic review see Pusic et al., 2007). Rather than a high correlation being reported between the HNQOLQ and voice quality, as would be expected were voice quality to determine quality of life, correlations were reported to be moderate at best. This led the authors to conclude that measures of auditory-perceptual voice quality and quality of life questionnaires evaluate different aspects of function after laryngectomy.
7
Methods of Voice Restoration and Quality of Life
There have been very few studies that have compared quality of life between patients adopting TEP, electrolarynx and esophageal speech following total laryngectomy. In part this situation potentially arises as a consequence of the significant clinical preference for TEP, making it very different to recruit participants employing esophageal speech or electrolarynx for statistical comparison. In the few studies that have been done, patients employing TEP were reported to have fewer problems associated with everyday living and lower levels of anxiety and depression than those using electrolarynx (Finizia and Bergman, 2001). Conclusions reached on the basis of this study however need to be treated with caution due to the small number of patients recruited into each voice restoration group, which included merely five electrolarynx users. Attempts to overcome limitations associated with low sample sizes were addressed by a study utilizing the SF-36 (Jenkinson et al., 1993), a general health specific quality of life, to compare quality of life between all three methods of voice restoration and a healthy age related control group (Farrand and Duncan, 2007). As would be predicted, and consistent with previous research (e.g., Clements et al., 1997; Robbins et al., 1984; Williams and Watson, 1987), there was a clear advantage for TEP over other voice restoration methods with respect
Speech Determines Quality of Life Following Total Laryngectomy
117
. Table 117‐4 Commonly adopted specific health and neck quality of life measures Quality of life measure
Details
Auckland quality of life questionnaire (Morton and Witterick, 1995).
Initially developed for a head and neck population in Auckland, New Zealand, 12 items examine quality of life across the following dimensions: appearance, eating, pain and speech
European Organization for Research and Treatment of Cancer: head and neck module (Bjordal et al., 1994).
A 35 item questionnaire addresses seven major quality of life dimensions: pain, senses, sexuality, social contact, social eating, speech, swallowing and 10 single item questions. Used in conjunction with the generic cancer quality of life measure
Functional assessment of cancer therapy: head and neck (List et al., 1996a).
An 11 item subscale covering five dimensions: emotional and functional well-being, relationship with doctor, physical, social and additional head and neck concerns
Head and neck cancer inventory (Funk et al., 2003).
Quality of life examined using 30 items to explore five dimensions: aesthetics, eating, pain/ discomfort, social activity and speech with one question examining overall quality of life
University of Michigan head and neck quality of Twenty items examine quality of life across four life questionnaire (Terrell et al., 1997). dimensions: communication, eating, emotion and pain University of Washington quality of life instrument (Rogers et al., 1999).
Twelve items examine quality of life across the following domains: activity, anxiety, appearance, chewing, mood, pain, recreation, shoulder, speech, swallowing and taste in addition to three questions that examine global health related quality of life
This table provides details of the quality of life dimensions measured by different quality of life instruments. It is interesting to note therefore that, using a range of methodological approaches, the literature has failed to consistently support the assumption that voice quality determines quality of life. A question arises therefore as to why there remains a strong preference by medical staff for employing TEP over other methods of voice restoration (Blom, 2000; Finizia et al., 1998a,b) without empirical evidence in support of such an assumption? One possibility is that whilst improved voice quality afforded by TEP does not determine quality of life, the method is associated with other benefits that still result in quality of life advantages in comparison to esophageal speech and electrolarynx
to voice quality and perception. As can be seen in > Table 117‐5 in quiet and noisy surroundings and over the telephone, TEP users rated the intelligibility of their voice more positively than users of electrolarynx or esophageal speech. Whilst resulting in improved voice perception however, TEP did not result in widespread quality of life advantages (see > Table 117‐6). Only with respect to the SF-36 quality of life dimensions of pain and role limitation – physical problems were there any quality of life advantages associated with TEP. On both dimensions however significant differences arose
1997
1998
117
Speech Determines Quality of Life Following Total Laryngectomy
. Table 117‐5 Mean rating of intelligibility for voice restoration methods in different surroundings TEP
Esophageal
Electrolarynx
Quiet
3.7 (3.6, 3.9)
3.4 (3.1, 3.7)
3.3 (2.9, 3.7)
Noisy
2.4 (2.2, 2.5)
1.8 (1.6, 2.0)
1.8 (1.5, 2.1)
Telephone
3.2 (3.6, 3.9)
2.8 (3.6, 3.9)
2.8 (3.6, 3.9)
Average voice perception score
3.1 (2.8, 3.0)
2.7 (2.4, 2.8)
2.8 (2.5, 3.0)
This table provides average ratings (95% confidence interval) between voice restoration group for voice intelligibility in quiet and noisy surroundings, on the telephone and an average voice perception score where 1 = ‘‘Very Poor’’ and 5 = ‘‘Very Good’’ (data from Farrand and Duncan, 2007)
. Table 117‐6 Comparisons on SF-36 dimensions between voice restoration groups SF-36 dimension Energy/vitality
TEP
Esophageal
Electrolarynx
55.05 (51.1, 59.0)
51.04 (43.8, 58.3)
51.30 (43.2, 59.2)
General health perception
58.11 (53.2, 62.9)
52.22 (43.4, 61.0)
50.43 (40.1, 60.3)
Mental health
72.14 (68.7, 75.6)
67.36 (61.1, 73.6)
77.14 (70.1, 84.2)
Paina
70.13 (65.2, 75.1)
56.07 (46.8, 65.3)
75.44 (65.5, 85.4)
Physical functioning
53.89 (48.7, 59.1)
56.02 (46.4, 65.5)
58.40 (47.8, 68.9)
Role limitation emotional problems
61.61 (54.0, 69.2)
56.98 (43.4, 70.6)
65.66 (50.8, 80.5)
Role limitation physical problemsb
46.03 (38.7, 53.3)
28.93 (15.1, 42.8)
48.09 (33.7, 62.4)
Social functioning
72.07 (67.2, 76.9)
68.97 (59.9, 78.1)
76.86 (67.3, 86.4)
Quality of life scores (95% confidence interval) for different dimensions of the SF-36 across different voice restoration groups. Higher scores reflect higher quality of life, maximum = 100 (data from Farrand and Duncan, 2007) a p < 0.01 level b p < 0.0001
because of a significantly lower quality of life experienced by users of esophageal speech rather than a specific advantage experienced by TEP users. Across all of the SF-36 quality of life dimensions differences between TEP and electrolarynx were non-significant. That improved voice perception did not dictate better quality of life reaffirms previous findings demonstrating no clear relationship between quality of life and voice perception in patients undergoing total laryngectomy (e.g., see Finizia et al., 1998a,b; Stewart et al., 1998). Although not dictating quality of life as has been proposed (Clements et al., 1997; Harwood and Rawlinson, 1983; McNeil et al., 1981), it remains likely that voice quality contributes to quality of life but does not determine it on its own. Other factors may therefore make just as important a contribution to patient quality of life, such as physical consequences and interference with social activities (Mohide et al., 1992) and are therefore worthy of equal consideration by medical staff. With respect to physical symptoms, a series of symptoms associated with respiratory function, such as daily sputum production, coughing, need for forced expectoration to clear airway and frequent stoma cleaning were identified as the principal complaints by patients following total laryngectomy (Hilgers et al., 1990). Given
Speech Determines Quality of Life Following Total Laryngectomy
117
the importance of physical symptoms in contributing to quality of life following total laryngectomy therefore, parsimony would suggest that the impact of voice restoration method upon physical symptoms is also taken into account when choosing the most appropriate method of voice restoration for the patient. Obviously one potential limitation of the TEP method in this respect is the greater likelihood of respiratory difficulties due to problems associated with fungal infection, salivary leakage or the prosthesis becoming dislodged (Blom, 2000; Ferrer Ramı´rez et al., 2001). So whereas TEP produces best voice quality, for some patients the increased need for ongoing medical treatment compared to esophageal speech and electrolarynx may result in a lower quality of life.
8
Considerations for Practice
A review of the literature reinforces the need for medical and health staff to take into account quality of life considerations from the perspective of the patient rather than to assume that patients will share their beliefs. It is now generally recognized across many areas of medical and health care that there is disagreement between doctors and their patients as to what factors will help to maximize quality of life (Janse et al., 2004). Indeed this has been explicitly addressed with respect to quality of life following total laryngectomy. Whilst medical staff ranked communication impairment, along with self image and self esteem as the most important factors affecting laryngectomy patients’ quality of life, patients ranked communication only as the third most important factor, after physical consequences and interference with social activities (Mohide et al., 1992). Additional research has confirmed the difference between medical and health staff and patients concerning the importance of voice quality to quality of life. Using a ‘‘> time-trade off ’’ quality of life methodology whilst 46% of the medical and health professionals felt that their patients would compromise reduced survival for improved voice and speech, only 20% of the patients themselves expressed this preference (Otto et al., 1997). Given that medical and health staff are poor at appreciating the factors that contribute to patient quality of life, we suggest that greater attention is paid towards the preferences of the patients themselves before recommending a particular voice restoration method. One way in which this could be easily achieved would be to include quality of life measures as a routine part of the patient assessment to inform treatment selection (Janse et al., 2004; Schuster et al., 2003). Furthermore, the need to consider other aspects of rehabilitation, such as swallowing and smell (Armstrong et al., 2001; Pillon et al., 2004; Risberg-Berlin et al., 2007; Sinkiewicz et al., 2006; Ward et al., 2002), places emphasis on a > multi-disciplinary approach to rehabilitation following total laryngectomy. Patient’s preferences would be central to the rehabilitation plan to maximize quality of life.
Summary Points Voice quality alone does not determine quality of life following total laryngectomy. Although TEP results in better voice quality than esophageal speech or electrolarynx, it does not necessarily lead to better overall quality of life.
Quality of life measures should be included as a routine part of the assessment of patients undergoing laryngectomy to inform treatment selection.
1999
2000
117
Speech Determines Quality of Life Following Total Laryngectomy
Before recommending a voice restoration method, medical and health staff alongside patients should consider the impact of method upon quality of life in areas such as swallowing, smell and ongoing need for maintenance, in addition to voice quality. The complex interplay of treatment regimes, complications and long term effects for patients undergoing total laryngectomy mitigates against attributing quality of life to voice loss alone. A multidisciplinary approach should be adopted towards the rehabilitation of patients undergoing total laryngectomy. Medical and health staff should take into account quality of life considerations from the perspective of the patient rather than assume that patients share their beliefs.
References Armstrong E, Isman K, Dooley P, Brine D, Riley N, Dentice R, King S, Khanhbai F. (2001). Head Neck. 23: 16–24. Bjordal K, Ahlner-Elmqvist M, Tollesson E, Jensen AB, Razavi D, Maher EJ, Kaasa S. (1994). EORTC Quality of Life Study Group. 33: 879–885. Blom ED. (1998). In: Blom ED, Singer M, Hamaker RC. (eds.) Trachesophageal Voice Restoration Following Total Laryngectomy. Singular, San Diego, pp. 1–8. Blom ED. (2000). Oncology. 14: 915–922. Clements KS, Rassekh CH, Seikaly H, Hokanson JA, Calhoun KH. (1997). Arch Otol Head Neck Surg. 123: 493–496. Deleyiannis FW, Weymuller EA, Coltrera MD. Furtran N. (1999). Head Neck. 21: 319–324. DeSanto LW, Olsen KD, Perry WC, Rohe DE, Keith RL. (1995). Ann Oto Rhinol Laryn. 104: 763–769. Devlin JG, Langer CJ. (2007). Expert Rev Anticanc. 7: 331–350. Drummond S, Dancer J, Krueger K, Spring G. (1996). Percept Motor Skill. 83: 801–802. Eadie TL, Doyle PC. (2004). Laryngoscope. 114: 753–759. Farrand P, Duncan F. (2007). Psychol Health Med. 12: 255–265. Ferrer Ramı´rez MJ, Guallart Dome´nech F, Brotons Durba´n S, Carrasco Llatas M, Estelle´s Ferriol E, Lo´pez Martı´nez R. (2001). Eur Arch Oto-Rhino-L. 258: 463–466. Finizia MD, Bergman B. (2001). Laryngoscope. 111: 918–923. Finizia C, Hammerlid E, Westin T, Lindstro¨m J. (1998a). Laryngoscope. 108: 1566–1573. Finizia C, Lindstro¨m J, Dotevall H. (1998b). Laryngoscope. 108: 138–143. Forastiere AA, Trotti A, Pfister DG, Grandis JR. (2006). J Clin Oncol. 24: 2603–2605.
Funk GF, Karnell LH, Christensen AJ, Moran PJ, Ricks J. (2003). Head Neck. 25: 561–575. Hall FT, O’Brien CJ, Clifford AR, McNeil EB, Bron L, Jackson MA. (2003). ANZ J Surg. 73: 300–305. Harwood AR, Rawlinson E. (1983). Int J Radiat Oncol. 9: 335–338. Herranz J, Sarandeses A, Ferna´ndez MF, Barro CV, Vidal JM, Gavila´n J. (2000). Otolaryng Head Neck. 122: 892–898. Hilgers FJM, Ackerstaff AH, Aaronson NK, Schouwenburg PF, Van Zandwijk N. (1990). Clin Otolaryngol. 15: 421–425. Hoffman HT, Porter K, Karnell LH, Cooper JS, Weber RS, Langer CJ, Ang KK, Gay G, Stewart A, Robinson R. (2006). Laryngoscope. 116 (Suppl. 2): 1–13. Janse AJ, Gemke RJ, Uiterwaal CSPM, van der Tweel I, Kimpen JL, Sinnema G. (2004). J Clin Epidemiol. 57: 653–661. Jenkinson C, Coulter A, Wright L. (1993). British Med J. 306: 1437–1440. List MA, D’Antonio LL, Cella DF, Siston A, Mumby P, Haraf D, Vokes E. (1996a). Cancer. 77: 2294–2301. List MA, Ritter-Sterr CA, Baker TM, Colangelo LA, Matz G, Roa Pauloski B, Logemann JA. (1996b). Head Neck. 18: 1–10. LoTempio MM, Wang KH, Sadeghi A, Delacure MD, Juillard GF, Wang MB. (2005). Otolaryng Head Neck. 132: 948–953. Marioni G, Marchese-Ragona R, Giuseppe Cartei G, Fortunata Marchese F, Staffieri A. (2006). Cancer Treat Rev. 32: 504–515. McNeil BJ, Weichselbaum R, Pauker SG. (1981). New Engl J Med. 305: 982–987. Mohide EA, Archibald SD, Tew M, Young JE. (1992). Am J Surg. 14: 619–622. Morris HL, Smith AE, Van Demark DR, Maves MD. (1992). Ann Oto Rhinol Laryn. 101: 503–510.
Speech Determines Quality of Life Following Total Laryngectomy Morton RP, Witterick IJ. (1995). Am J Otolaryngol. 16: 284–293. Office for National Statistics. (2007). Cancer Registration Statistics 2004. The Stationary Office, London. Otto RA, Dobie RA, Lawrence V, Sakai C. (1997). Ann Oto Rhinol Laryn. 106: 693–699. Paydarfar JA, Birkmeyer NJ. (2006). Arch Otolaryngol. 132: 67–72. Pillon J, Gonc¸alves MIR, De Biase NG. (2004). Sao Paulo Med J. 122: 195–199. Pou AM. (2004). Otolaryng Clin N Am. 37: 507–517. Pusic A, Liu JC, Chen CM, Cano S, Davidge K, Klassen A, Branski R, Patel S, Kraus D, Cordeiro PG. (2007). Otolaryng Head Neck.136: 525–535. Risberg-Berlin B, Mo¨ller RY, Finizia C. (2007). Arch Otolaryngol. 133: 650–654. Robbins J, Fisher HB, Blom EC, Singer MI. (1984). J Speech Hear Disord. 49: 202–210. Rogers SN, Lowe D, Brown JS. (1999). Head Neck. 21: 394–401. Schuster M, Toy H, Lohscheller J, Eysholdt U, Rosanowski F. (2003). Folia Phoniatr Logo. 55: 211–219.
117
Sinkiewicz A, Winiarski P, Mackiewicz H, Owczarzak H, Janicka-Beutch L, Betlejewski S. (2006). Otolaryngol Pol. 60: 33–36. Stewart MG, Chen AY, Stach CB. (1998). Arch Otolaryngol. 124: 143–148. Taylor JC, Terrell JE, Ronis DL, Fowler KE, Bishop C, Lambert MT, Myers LL, Duffy SA. (2004). The University of Michigan Head and Neck Cancer Team. Arch Otolaryngol. 130: 764–769. Terrell JE, Fisher SG, Wolf MD. (1998). Arch Otolaryngol. 124: 964–971. Terrell JE, Nanavati KA, Esclamado RM, Bishop JK, Bradford CR, Wolf GT. (1997). Arch Otolaryngol Head Neck Surg. 123: 1125–1132. Ward EC, Bishop B, Frisby J, Stevens M. (2002). Arch Otolaryngol. 128: 181–186. Williams SE, Watson JB. (1987). Laryngoscope. 97: 737–739.
2001
118 Health-Related Quality of Life of Living Kidney Donors Ja Hyeon Ku . Hyeon Hoe Kim 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2005
2
Quality of Life Instruments for Living Kidney Transplantation . . . . . . . . . . . . . . . 2006
3
Quality of Life Living Kidney Donors Compared to the General Population . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2007 North and South America . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2007 USA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2007 Canada . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2011 Brazil . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2011 Europe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2011 UK . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2011 Germany . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2011 Italy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2013 Spain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2013 Sweden . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2013 Norway . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2013 Hungary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2014 Turkey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2014 Asia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2014 Japan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2014 Taiwan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2014 Iran . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2015 Australia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2015
3.1 3.1.1 3.1.2 3.1.3 3.2 3.2.1 3.2.2 3.2.3 3.2.4 3.2.5 3.2.6 3.2.7 3.2.8 3.3 3.3.1 3.3.2 3.3.3 3.3.4 4 4.1 4.2 5 5.1 5.2
#
Quality of Life in Living Kidney Donors to According to the Surgical Technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2016 Mini-Incision Donor Nephrectomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2016 Laparosocpic Donor Nephrectomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2016 Other Factors Associated with Quality of Life in Living Kidney Donors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2019 Donor’s Gender . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2019 Donor’s Age . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2019
Springer Science+Business Media LLC 2010 (USA)
2004
118
Health-Related Quality of Life of Living Kidney Donors
5.3 5.4 5.5 5.6
Time Since Donation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2020 Relationship with the Recipient . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2021 Perioperative Complications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2022 Outcomes for the Recipient . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2022
6
Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2022 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2023
Health-Related Quality of Life of Living Kidney Donors
118
Abstract: Kidney transplantation provides improvement in both the length and quality of life (QOL) in patients receiving dialysis. The kidneys used for transplantation are currently obtained from either cadavers or living donors. However, the demand for cadaver organs far exceeds the supply. The shortage of organs and the steadily growing waiting time for a cadaver kidney transplantation have forced the medical community to look for alternatives, such as living kidney donation. Because of the better outcome with the living grafts for the recipients compared to the cadaver transplants, living donor programs have been expanding. The benefits of living donor transplantation have been well documented. In addition, patients have an extensive medical work-up that not infrequently identifies occult medical problems, a boost in self-esteem and an increased sense of well-being. Furthermore, living kidney donation has not been shown to have a detrimental effect on the physical or psychological well-being of donors. However, available data examining the health-related QOL issues of living donors are currently limited. Although there is no consensus on the “gold standard” instrument to measure QOL in this population, to date, the SF-36 has been widely used to measure the QOL of living kidney donors. Most studies have suggested that living kidney donors have similar or higher scores on the QOL questionnaires compared with the healthy general population. The evolution of surgical techniques used for transplantation including the mini-incision and laparoscopy has made living kidney donation a more attractive option for patients and their potential donors. In addition, several variables have been proposed as an influencing the QOL of living kidney donors but these factors are not conclusive. Careful donor selection, with appropriate pre-transplantation psychiatric consulting as well as regular post-donation psychosocial screening and the provision of specific interventions to those in need are recommended for a successful living kidney donor program. List of Abbreviations: > MCS, mental component summary; summary; QOL, quality of life; SF-36, short-form 36
1
> PCS,
physical component
Introduction
Organ replacement from either a living or deceased donor is preferable to dialysis therapy because transplantation provides a better quality of life (QOL) and improved survival. Successful kidney transplantation is the optimal treatment for most patients with end-stage renal disease who require dialysis. The incidence of chronic renal failure has increased considerably over the past decade, and the number of organs from cadavers is not sufficient to meet the current demand for transplantation. The shortage of cadaver kidneys for transplantation means that many individuals must wait for long periods to receive the benefits of transplantation. Living donors provide one way to bridge the gap between the supply and demand. The first successful living donor kidney transplant was performed 50 years ago. Since then, in a relatively brief period of medical history, living kidney transplantation has become the preferred treatment for patients with end stage renal disease. In fact, living donation is an attractive alternative to cadaver transplantation because living donor transplants have superior outcomes: they are less likely to suffer from delayed graft function or acute tubular necrosis, and have a longer half-life than cadaver transplants. Living kidney donation offers other potential benefits, including greater convenience for the recipient, better access to transplantation for the recipient and reduced financial burden to society.
2005
2006
118
Health-Related Quality of Life of Living Kidney Donors
The clinical benefits for the donor are less clear. In the context of living donation, the safety of donors is extremely important. Living donor kidney transplantation is considered an extremely safe procedure, with a worldwide overall mortality rate of 0.03% (Bruzzone et al., 2005). In addition, some benefits for the living kidney donor have been documented, such as a boost in self-esteem and an increased sense of well-being (Toronyi et al., 1998; Westlie et al., 1993). However, the donor risks include the short-term surgical risk as well as the long-term risks of impaired renal function, hypertension and psychological problems (Najarian, 2005). The concerns regarding the psychological impact on donors have become more prominent, with some reports of depression after donation and even of suicide after a recipient’s death (Weizer et al., 1989). Recently, there is growing interest in adding to the complex notion of recovery outcomes to include the patient’s perspective. Thus, measures of physical activity and capacity, as well as physical, emotional, and social well-being, collectively referred to as health-related QOL, have become important outcome factors (Carli and Mayo, 2001). Therefore, recovery following donation in terms of complete restoration of the donor’s working capability as well as QOL following donation is of particular interest. To date, available data examining the QOL issues of living donors are limited. Furthermore, some limitations exist concerning: (1) sample size, (2) the response rate, and (3) validity of the applied psychological instruments-the applied psychological instruments, in the past, often included only a few dimensions and a small number of items. In this chapter, the data regarding health-related QOL of living kidney donors is summarized.
2
Quality of Life Instruments for Living Kidney Transplantation
Standardized instruments should help donors to focus on the issues, help ensure a comprehensive assessment, provide a basis for serial monitoring, and permit auditing that can be reported in a universally interpretable way. There are differences in the levels of literacy, taboo subjects and social desirability among cultures. To define and assess health-related QOL across ethnic groups requires the development of outcome measures in the relevant languages that are culturally appropriate for cross-ethnic studies. Generic QOL tools are used in general populations to assess a wide range of domains applicable to a variety of states of health, conditions and diseases. The main differences among these instruments are the degree to which they emphasize the objective compared to the subjective dimensions, the extent to which various domains are covered (i.e. the number of items and categories) and the format of the questions. The most commonly used instruments are the Nottingham Health Profile, the Sickness Impact Profile, the Short-Form 36 (SF-36), and the LEIPAD Quality of Life. Disease-specific instruments focus on the domains most relevant to the disease or condition under study and on the characteristics of patients in whom the condition is most prevalent. However, despite the high number of both generic and disease-specific instruments, up to now there is no consensus on the “gold standard” instrument to measure the QOL, and this may limit the comparability of results across studies. Experience to date suggests that the SF-36 can be adapted for use in different countries, with relatively minor changes to the content in the form. This feature of the SF-36 provides support for its use in many different translations across multinational clinical trials (Bullinger et al., 1998). Therefore, the SF-36 has been widely used in this field.
Health-Related Quality of Life of Living Kidney Donors
3
118
Quality of Life Living Kidney Donors Compared to the General Population
It remains unknown whether there are any differences between countries, races or social groups, with regard to health-related QOL in living kidney donors. However, most studies have suggested that donating a kidney may be associated with psychological benefits for the donors and have shown generally excellent QOL in kidney donors. This fact has been linked to specific personality traits such as a high self-esteem and positive reinforcement of this specific characteristic by donation. In the US, several studies have shown that living kidney donors have similar or higher scores in QOL domains compared to the general healthy US population (Buell et al., 2005; de Graaf Olson and Bogetti-Dumlao, 2001; Jacobs et al., 1998; Johnson et al., 1999; Perry et al., 2003; Schover et al., 1997). In a study from two UK transplantation centers, living kidney donation had no detrimental effect on the physical or psychological well being of donors 1 year after donation (Lumsdaine et al., 2005). In other countries, similar results were observed. > Tables 118-1–118-3 summarize the results of the health-related QOL of living kidney donors using the SF-36.
3.1
North and South America
3.1.1
USA
Schover et al. (1997) used the Medical Outcomes Study Short-Form Health Survey as a QOL tool and found that donors rated their overall health highly, and had excellent role and social functioning. In this study, 90% would make the same choice again and 83% would strongly encourage others to donate. Jacobs et al. (1998) studied 524 donors. Donors had a higher QOL than the general population, confirming that they have an increased sense of self-worth and positive self-esteem. An overwhelming 96% would donate again. Johnson et al. (1999) described the QOL of living kidney donors using a standardized and validated health survey QOL assessment tool, the SF-36. They sent a questionnaire to 979 American donors, and 60% responded. The donors scored better than the general US population in seven of eight categories and the donors on average scored much better than those with the two disease states tested (congestive heart failure and depression). However, 12% recalled the experience as being stressful or extremely stressful, and 4% regretted the donation. Overall, the vast majority of donors had a positive experience; they would readily donate again if it were possible. de Graaf Olson and Bogetti-Dumlao (2001) mailed the SF-36 questionnaire to all kidney donors. Most donors who had follow-up after donation felt that their experience was good to excellent; yet 50% stated they had no healthcare follow-up after donation. Donors perceived their quality of health after donation as better than the general US population. Perry et al. (2003) evaluated the health-related QOL of patients who underwent laparoscopic and mini-incision open donor nephrectomy in a retrospective fashion. The overall QOL for both the open mini-incision and laparoscopic donor nephrectomy donors was comparable with or higher than the age-matched general US population. Buell et al. (2005) also examined the QOL of laparoscopic and open living donor nephrectomy donors and found that the overall QOL for both open and laparoscopic donor nephrectomy donors was comparable with the general US population.
2007
92
90
75
82
67
92
90
80
Physical function
Rolephysical
Bodily pain
General health
Vitality
Social function
Roleemotional
Mental health
Johnson et al. (1999)
USA Buell et al. (2005)
83.5
97.9
93.6
80.1
89.7
95.4
94.7
97.2
81.8
84.4
86.6
72.5
86.0
81.2
86.5
92.7
81.1
89.4
87.2
63.5
81.5
83.7
86.7
93.8
84.8
85.7
90.5
69.5
83.1
81.6
88.1
94.0
MiniLaparoscopy incision Laparoscopy Open
Perry et al. (2003)
88
100
87.5
60
92
74
50
85
Bergman et al. (2005)
Canada
77
76
83
75
82
71
81
84
Lima et al. (2006)
Brazil
1 month: 81.5 3 months: 91.0
1 month: 86.8 3 months: 87.5
1 month: 75.4 3 months: 82.5
1 month: 77.8 3 months: 85.8
1 month: 88.1 3 months: 89.2
1 month: 63.8 3 months: 79.3
1 month: 78.3 3 months: 88.3
1 month: 53.3 3 months: 70.0
Lombotomy
1 months: 79.1 3 months: 86.4
1 months: 90.0 3 months: 85.7
1 month: 81.7 3 months: 84.8
1 month: 75.0 3 months: 85.3
1 month: 89.8 3 months: 88.3
1 month: 67.9 3 months: 80.6
1 month: 79.3 3 months: 82.9
1 month: 45.8 3 months: 72.3
Subcostal
Aguiar et al. (2007)
118
. Table 118-1 Health-related quality of life in North and South American living kidney donors
2008 Health-Related Quality of Life of Living Kidney Donors
Better scores Equal scores for both than the gen- groups with the eral US popu- general US population lation in seven domains
Equal scores for both open and laparoscopic nephrectomy donors with the general US population Decreased PCS but constant MCS
Better Significant reductions in the SF-36 scores at 1 month and scores a relevant recovery at 3 months (Vitality and General Health) than the control group
Each of the dimension scores was expressed as a value between 0 and 100, with higher scores representing better health. The eight domains are collapsed to create two global components, a physical component score and a mental component score. Scores on the Physical Component Summary are associated with high scores on the Physical Function, Role-Physical, Bodily Pain and General Health scales and low scores on the Role-Emotional and Mental Health scales. For the Mental Component Summary, positive weights are placed on the Mental Health, Role-Emotional, Social Function and Vitality scales, whereas substantial negative ones are placed on the Physical Function and Role-Physical scales. PCS Physical Component Summary; MCS Mental Component Summary; SF-36 short-form 36
Remark
Health-Related Quality of Life of Living Kidney Donors
118 2009
2010
118
Health-Related Quality of Life of Living Kidney Donors
. Table 118-2 Health-related quality of life in European living kidney donors Germany Jackobs et al. (2005) MiniFlank incision incision Physical function
92
Rolephysical
91
Bodily pain
90
General health
80
Vitality
90
Italy
Turkey
Virzi et al. (2007)
Tanriverdi et al. (2004)
78.7
75.3
77.9
68.1
27.7
75.7
56.8
73.7
55.7
73.9
57.7
80.6
–
68.5
66.6
73.8
Schnitzbauer et al. (2007) Mini-incision 1 week: 33.7
Flank incision 1 week: 31.0
3 months: 47.8 3 months: 48.3 86
1 year: 54.8
1 year: 53.6
1 week: 33.7
1 week: 31.1
3 months: 44.7 3 months: 47.0 84
1 year: 53.5
1 year: 52.6
1 week: 42.1
1 week: 37.9
3 months: 51.6 3 months: 52.6 78
1 year: 58.0
1 year: 59.4
1 week: 49.5
1 week: 52.1
3 months: 53.3 3 months: 56.4 70
66
1 year: 56.0
1 year: 57.0
1 week: 49.1
1 week: 48.4
3 months: 55.8 3 months: 58.3 Social function
Roleemotional
Mental health
92
87
1 year: 59.7
1 year: 60.0
1 week: 44.1
1 week: 47.5
3 months: 53.0 3 months: 53.7 90
86
1 year: 55.8
1 year: 54.4
1 week: 37.4
1 week: 39.7
3 months: 46.1 3 months: 49.5 80
73
1 year: 52.2
1 year: 52.9
1 week: 51.0
1 week: 50.6
3 months: 53.7 3 months: 55.2 1 year: 56.3
Remark
Comparable or higher scores in all the SF-36 categories of both groups in comparison to the general US population
1 year: 54.8
Lower QOL scores after 1 week and higher scores after 1 years compared with norm-based US population
Significant No negative worsening in consequences physical func- with respect to tion and pain personal subscales health, family relationships, or energy level
Each of the dimension scores was expressed as a value between 0 and 100, with higher scores representing better health. SF-36 short-form 36; QOL quality of life
Health-Related Quality of Life of Living Kidney Donors
3.1.2
118
Canada
Vlaovic et al. (1999) reported on 104 Canadian donors. Less than 5% of donors complained of the renal donation severely affecting any aspect of their life. Most donors (84%) were able to perform their normal daily activities within 12 weeks of the nephrectomy. Almost one third of the donors lost wages because of their donation, and half incurred significant transportation costs. Almost 90% of donors felt that donating a kidney had positively influenced their relationship with the recipient, and donors felt that their relationships with the recipient were significantly more positive at follow-up. However, Bergman et al. (2005) found that 4 weeks following laparoscopic live donor nephrectomy, a majority of patients seem to have returned to baseline exercise capacity and general > mental health, but have not returned to baseline general physical health.
3.1.3
Brazil
Lima et al. (2006) assessed 100 donors (34 men and 66 women) during a post-operative period lasting longer than 2 years and observed that the QOL of kidney donors was not different than it was for the healthy individuals of the community. The kidney donors had a QOL index similar to the control group in five of the eight parameters evaluated according to the SF-36 survey. Donors had better indexes than the control group for > Vitality and > General Health. Aguiar et al. (2007) reported that donors had significant reductions in the SF-36 scores when the baseline scores were compared to scores 1 month after surgery. However, 3 months after the surgery, there was a significant recovery and the SF-36 scores were similar to those observed before the surgery.
3.2
Europe
3.2.1
UK
In a prospective, longitudinal cohort study, the QOL of 40 donors using the World Health Organisation QOL questionnaire were evaluated before, 6 weeks and 1 year after surgery (Lumsdaine et al., 2005). Before the donation, the donor QOL in the physical domain was significantly higher than the UK values for a healthy person in the general population. They found that 6 weeks after surgery, the score was lower and more compatible with the UK normative levels but improved again at 1 year.
3.2.2
Germany
Giessing et al. (2004) evaluated the impact of kidney donation on German donor’s QOL. They sent questionnaires to donors who could be contacted and analyzed the answers of 106 donors (response rate: 90%). Most donors had an equal or better QOL than the general healthy population (subjects aged more than 14 years and living in the general East and West German population). For three items (> Physical Function, > Role-Physical and General Health), kidney donors had a significantly better score than the matched controls in the general population. For another four
2011
2012
118
Health-Related Quality of Life of Living Kidney Donors
. Table 118-3 Health-related quality of life in Asian living kidney donors Japan
Taiwan
Iran
Australia
Isotani et al. (2002)
Chen et al. (2004)
Zargooshi et al. (2001)
Smith et al. (2004)
Physical function
90.9
84.4
35.2
92.9
Rolephysical
88.9
84.0
62.5
90.6
Bodily pain
85.9
78.4
48.0
80.6
General health
75.2
81.5
30.1
79.3
Vitality
72.2
83.2
34.8
68.9
Social function
92.1
83.9
64.9
89.9
Roleemotional
88.4
79.9
37.9
93.1
Mental health
83.1
78.6
43.4
78.4
Significant lower scores in vendors on all SF-scales compared to controls
Preoperative higher PCS and MCS than community norms and significant decreased of postoperative PCS and MCS
Remark
No significant dif- Worst scores ference from the of “Bodily general US populaPain” and tion and US donors “Mental Health”
Each of the dimension scores was expressed as a value between 0 and 100, with higher scores representing better health. The eight domains are collapsed to create two global components, a physical component score and a mental component score. Scores on the Physical Component Summary are associated with high scores on the Physical Function, Role-Physical, Bodily Pain and General Health scales and low scores on the Role-Emotional and Mental Health scales. For the Mental Component Summary, positive weights are placed on the Mental Health, Role-Emotional, Social Function and Vitality scales, whereas substantial negative ones are placed on the Physical Function and Role-Physical scales. SF-36 short-form 36; PCS physical component summary; MCS Mental Component Summary
items (> Bodily Pain, Vitality, > Social Function, and Mental Health), the donors scored better than the general population, but the differences were not significant. The score for > RoleEmotional was lower for the study population than for the general population, but the difference was not significant. The donors were not concerned about living with one kidney, and most of them would donate again, if this were an option. Jordan et al. (2004) carried out a semi-structured interview and four psychological questionnaires. Donors scored better on a wide range of the psychological scales such as psychological symptoms, health behavior and health consciousness that was expected in comparison with the data from the general German population samples. Nearly all donors (97%) would choose to donate again, and 91% remain entirely satisfied with their decision. After donation, 3% stated that they had psychopathological illness related to the donation and 2% reported distressing events related to the donation. Schnitzbauer et al. (2007)
Health-Related Quality of Life of Living Kidney Donors
118
investigated the donor’s QOL in comparison with the norm-based scores. Preoperatively, the donors had significantly higher scores in the > Physical Component Summary (PCS) and similar scores in the > Mental Component Summary (MCS) compared with a norm-based US population. Thereafter the PCS dropped to lower values after 1 week, increased to similar values after 3 months and to better values 1 year after the procedure. The MSC was similar 1 week after surgery, but 3 months and 1 year after surgery, the scores were significantly better than the norm-based scores. However, Reimer et al. (2006) found that the global QOL was statistically significantly impaired in the kidney donors in the areas “Physical Function” and “Role-Emotional,” whereas the donors scored statistically significantly higher in the areas of “General Health,” “Vitality” and “Mental Health” compared to the general population. The sum scores (PCS and MCS) did not differ from the population norm.
3.2.3
Italy
Recently, 48 Italian donors were evaluated before and 4 months after transplantation, using a clinical interview and psychodiagnostic tests (mini-mental state, Hamilton Rating Scale for Depression, Hamilton Anxiety Scale, Self-Rating Anxiety Scale, and SF-36 Questionnaire) (Virzi et al., 2007). This study demonstrated that living donor kidney transplantation affected the donor QOL, which appeared decreased in physical aspects, especially due to pain symptoms and probably to the concern about living with a solitary kidney. However, there was a general improvement in psychological status and a boost in self-esteem.
3.2.4
Spain
In Spain, 22 donors of living kidneys were evaluated 6 months after donation (Cabrer et al., 2003). All donors stated after donation that they would again favor it and all but one felt completely recovered with the same QOL after donation.
3.2.5
Sweden
Fehrman-Ekholm et al. (2000) assessed the subjective health state of Swedish living donors. Each living donor was mailed and 370 (response rate: 92%) answered the questionnaire. According to the SF-36, the overall subjective health scores of the donors were satisfactory. Donors on all eight health scales scored higher than the age- and gender-adjusted general Swedish population. Less than 1% of the donors regretted the donation.
3.2.6
Norway
Westlie et al. (1993) examined 494 Norwegian donors using a standardized questionnaire containing 19 items related to the QOL. When the donors were compared to the general adult population in a county in mid-Norway, the donors scored significantly better in 13 items out of the 19 QOL items.
2013
2014
118 3.2.7
Health-Related Quality of Life of Living Kidney Donors
Hungary
Toronyi et al. (1998) contacted 78 living kidney donors, and 30 were interviewed and completed a questionnaire. All agreed that the donation did not change their general health. Regarding general attitudes towards living related organ transplantation, all were in favor of blood relative donor transplantation and husband/wife transplantation although opinions regarding non-related transplantation were more mixed.
3.2.8
Turkey
Health-related QOL and mood in Turkish renal transplant donors and the controls were investigated (Ozcurumez et al., 2004; Tanriverdi et al., 2004). The majority of living kidney donors did not experience negative consequences with respect to personal health, family relationships, or energy level and were comfortable with their choice to donate. Donor subjects had lower depression scores on the Beck Depression Inventory than the controls. This might be explained by a highly positive experience with kidney donation and enhanced self-esteem and self-regard related to this act. However, most donors experienced anxiety (based on the Beck Anxiety Inventory) after the transplantation procedure. This is understandable, possibly being associated with worry about the survival of the transplanted kidney, the outcome of the operation in terms of ability to recover and/or reenter the work force and the risks of living with a single kidney. Almost all donors (94.4%) said that they would make the same decision again and would strongly encourage others to donate. Issues related to donor dissatisfaction involved beliefs that preoperative information was inadequate and the perceived negative effects of the transplantation on personal health at the time the survey was conducted.
3.3
Asia
3.3.1
Japan
Isotani et al. (2002) conducted a psychosocial follow-up of Japanese living kidney donors. The mean SF-36 scores of 69 donors (response rate: 66%) were not significantly different from those of the general US population and US donors. In response to the question, “If possible, would you make the same choice again?” 97% of donors said they would agree to donate and 3% believed that donating had had a negative impact on their health. Most (84%) thought that the donation involved only a minor financial burden. In Japan, the cost of the transplantation of a living kidney donation is not borne by the donor, but is included in the cost to the recipient. It is unclear why the authors compared the SF-36 scores to those of the general US population and not to the general Japanese population although normative data for the Japanese population have been published.
3.3.2
Taiwan
Chen et al. (2004) reported the QOL in Taiwanese living kidney donors. In the SF-36 questionnaire, the scores of “Bodily Pain” and “Mental Health” were the worst, possibly
Health-Related Quality of Life of Living Kidney Donors
118
from the long operative wound from the open nephrectomy. Most donors were concerned about the cosmetic problems and pain-related to scar formation. Two of their young female donors underwent wound revision for cosmetic reasons 1 year after the transplantation. One donor was depressed due to graft loss by her son. These findings explain why the “Mental Health” score was low in their series. However, the QOL changes after donation were low and the SF-36 scores were comparable with those of the general population. Because the public health insurance system in Taiwan provides donors with sufficient support to deal with any future medical problems, this may play an important role in the high scores of the other items in the SF-36.
3.3.3
Iran
Paid, living unrelated renal vendors constitute greater than 90% of kidney donors in Iran. Zargooshi et al. (2001) reported on the QOL of Iranian vendors. In this study, 307 vendors (response rate: 97.7%) completed a questionnaire. Iranian kidney vendors had significantly lower scores on all SF-scales compared to controls. They responded that if they had another chance 85% would definitely not provide a kidney again, and 76% strongly discouraged potential vendors from repeating their error. In addition, high rates of self-reported de novo depression and anxiety after vending, and generally negative effects of vending on health and life existed. This study is the first study that provided information regarding vendor QOL.
3.3.4
Australia
There are few prospective psychosocial outcome studies on living kidney donors. Smith et al. (2003, 2004) conducted a psychosocial assessment and monitoring of living kidney donors prospectively. The psychological assessment of the living kidney donors was performed preoperatively and at 4 and 12 months postoperatively. Preoperatively, both physical function (SF-36 PCS score) and psychosocial function (SF-36 MCS score) were significantly higher than the community (state of Victoria) norms. The MCS scores decreased in comparisons between the preoperative period and the 4-month postoperative period, and remained significantly lower 12 months postoperatively. At 4 and 12 months postoperatively, the MCS was no longer significantly higher than the community norms. The PCS scores showed no significant decrease over time although there were significant decreases (between preoperatively and 12 months postoperatively) for the scales of “Bodily Pain,” “General Health” and “Vitality.” The PCS remained significantly higher than the community norms at 4 and 12 months postoperatively. Interestingly, the MCS of donors who developed adjustment and anxiety disorders were significantly lower than were those without a psychiatric disorder. These findings justify the recommendation that donors need to be educated about the extent of the psychosocial impairment that might occur during the postoperative period. The findings of depression, in this group, were in accordance with the results of previous studies (Schover et al., 1997; Westlie et al., 1993). According to previous studies, donors mostly reported a better QOL than their reference groups, independent of gender, kinship and age. This may in part be attributable to the fact that before kidney donation, donors have a significantly higher QOL compared with their reference group, thus possibly a “reserve” occurs for a time after the living donor kidney
2015
2016
118
Health-Related Quality of Life of Living Kidney Donors
transplantation, as Smith et al. (2003) found in a longitudinal study. Donors’ QOL in their study decreased significantly after the operation, but was still better than the reference groups’ QOL because the donors started out with a higher score. However, many studies have several limitations. The QOL of renal donors was compared without matching the reference group for age or gender, without standardized and validated questionnaires, the findings were compared with a reference group that was validated by another cultural background, and the studies involved too few participants or had low response rates in the range of 38–67%. Furthermore, recently, concern has been raised about the safety of donors (Vastag, 2003). Donors were found to be at a higher risk for developing a psychiatric disorder. Between 1 and 5% of donors would not donate again, up to 5% suffer from long-term mental problems, up to 15% of donors believe that donating impacted negatively their health, and between 16 and 23% reported negative financial consequences (Isotani et al., 2002; Smith et al., 2003, 2004). Furthermore, some studies described impaired QOL in donors (Smith et al., 2003, 2004). Therefore, some authors suggested that donors should be monitored for post donation psychosocial problems and counseling offered when needed (Giessing et al., 2004; Smith et al., 2003).
4
Quality of Life in Living Kidney Donors to According to the Surgical Technique
Some information regarding the QOL in living kidney donors according to the surgical technique is currently available but this issue remains controversial. > Table 118-4 summarizes the results of the QOL in living kidney donors according to the surgical technique.
4.1
Mini-Incision Donor Nephrectomy
In one study, the greatest concern among living kidney donors were cosmetic and pain-related scar formation (Chen et al., 2004). Thus, the mini-incision was introduced for living donor nephrectomy. Neipp et al. (2004) demonstrated that the anterior vertical mini-incision for living kidney donation had not only resulted in comparable or even improved QOL, but also reduced morbidity compared to the conventional flank incision. Following the mini-incision donor nephrectomy, the QOL tended to be superior compared to that of the lumbar nephrectomy, although statistical significance was reached only in one of the SF-36 eight categories (Jackobs et al., 2005). Aguiar et al. (2007) reported that the position of the mini-incision (lombotomy or subcostal) has no significant impact on surgical outcomes, pain perception and the QOL of living kidney donors. By contrast, Schnitzbauer et al. (2007) found that the QOL was similar at all time-points (prior to surgery, 1 week, 3 months and 1 year) in the miniincision and conventional incision groups and concluded that the QOL after living donor nephrectomy is not influenced by the surgical technique.
4.2
Laparosocpic Donor Nephrectomy
The evolution of surgical techniques used for transplantation has made kidney living donation a more attractive option for patients and their potential donors. Laparoscopic live donor
Health-Related Quality of Life of Living Kidney Donors
118
nephrectomy allows for kidney retrieval form the donor by multiple but smaller incisions. The advantages of laparoscopic donor nephrectomy include decreased hospital stay, decreased convalescence, less pain, a quick return to normal daily activities, and greater patient acceptance (Kuo and Johnson, 2000; Schweitzer et al., 2000). In addition, the availability of the laparoscopic technique has increased the number of people willing to donate, thus, increasing the pool of potential donors (Kuo and Johnson, 2000; Schweitzer et al., 2000). Perry et al. (2003) reported the first study that compared health-related QOL between laparoscopic and open living donor nephrectomy donors by using a standardized and validated questionnaire. Health-related QOL was significantly higher in the laparoscopy group than in the mini-incision group in three domains that measured “Bodily Pain,” “Physical Function” and “Role-Emotional.” The scores in the other five categories generally favored the laparoscopy group but did not achieve statistical significance. Pace et al. (2003) demonstrated that patients who underwent laparoscopic nephrectomy had higher QOL scores and returned to their preoperative QOL faster compared to those who underwent open nephrectomy. Kok and colleagues (2006) reported on a randomized controlled trial that compared the effects of laparoscopic and mini incision open donor nephrectomy on the QOL in living donors after renal transplant surgery. The authors found that the patients who had the laparoscopy had better scores for physical fatigue (multidimensional fatigue inventory 20) and physical function (SF-36) at 1 year than those that had the mini incision technique. They concluded that in centers with a highly experienced laparoscopic surgeon the laparoscopic technique is superior. Simforrosh et al. (2005) reported that the cosmetic results of laparoscopic donor nephrectomy seem to be much better than the open donor nephrectomy and this may be an important factor for donor satisfaction. However, other investigators have suggested that the influence of the laparoscopic technique on long-term donor satisfaction seems less important than anticipated (Buell et al., 2005; Dahm et al., 2006; Giessing et al., 2005; Rodrigue et al., 2006; Troppmann et al., 2006). Giessing et al. (2004) showed that the applied surgical technique (laparoscopic vs. open) did not differ between the donors scoring better or worse than the general German population. In a subsequent study, Giessing et al. (2005) found that health-related QOL showed significantly different scores only for the item “Mental Health,” with higher scores for the laparoscopy group; the willingness to donate again was not affected by the surgical method. However, most donors choose laparoscopic kidney retrieval, when asked. Buell et al. (2005) did not find that there were significant differences between laparoscopic donor nephrectomy and open donor nephrectomy groups with respect to the SF-36 health survey. In a study by Rodrigue et al. (2006), laparoscopic and open nephrectomy donors did not differ significantly with regard to pain perception, number of surgical/medical complications, physical health problems, financial impact, health-related QOL, or overall satisfaction; although the laparoscopic nephrectomy donors had significantly fewer hospital days and a faster return to work than did the open nephrectomy donors. Dahm et al. (2006) observed that the type of surgical procedure was not a very important factor in the donor’s decision-making process, although the laparoscopic approach offers the short-term benefits of less pain, shorter hospitalization and quicker return to routine activities. After the immediate post-operative period, the advantages of laparoscopy faded, and all live kidney donors experienced a high degree of satisfaction. In a study by Troppmann et al. (2006), which was the first study that compared postoperative psychosocial outcomes in those donating a kidney to a child by open and laparoscopic nephrectomy, there were no significant differences in the SF-36 responses between laparoscopic donors, open donors and a normative sample of the general US population. These findings demonstrate that donors are highly motivated and committed to donation,
2017
2018
118
Health-Related Quality of Life of Living Kidney Donors
. Table 118-4 Health-related quality of life in living kidney donors according to surgical technique Surgical technique
Remark
Mini-incision versus Superior QOL in the mini-incision group and Flank incision significant difference only in “Mental Health” score No significant difference of QOL
References Jackobs et al. (2005)
Schnitzbauer et al. (2007)
Mini-incision versus No significant difference of QOL according to the Aguiar et al. (2007) Mini-incision position (lombotomy or subcostal) Laparoscopy versus Higher QOL scores and faster return in the Flank incision laparoscopy group compared to the open nephrectomy group
Pace et al. (2003)
Different scores only for the item “Mental Health,” with higher scores for the laparoscopy group
Giessing et al. (2005)
No significant difference of QOL
Buell et al. (2005) Rodrigue et al. (2006) Dahm et al. (2006) Troppmann et al. (2006)
Laparoscopy versus Superior QOL in the laparoscopy group and Mini-incision significant difference in “Physical Function,” “Bodily Pain,” “Role-Emotional” score
Perry et al. (2003)
Better scores for physical function at 1 year in the Kok et al. (2006) laparoscopy group Laparoscopy versus No significant difference of QOL HALS
Bargman et al. (2006)
Both lombotomy and subcostal incisions are mini-incision access techniques via a posterior vertical incision and an incision below the ribs for donor nephrectomy, respectively. Laparoscopy is a surgical procedure that uses a thin, lighted tube called a laparoscope inserted through an incision in the abdominal wall. QOL quality of life; HALS hand-assisted laparoscopic surgery
suggesting that an increase in the rate of live kidney donation would be expected by the positive effects of the laparoscopic procedure on the recipients. Several recent comparison studies have reported that hand-assisted donor nephrectomy offers several advantages over standard laparoscopy, including reduced total operative time, time to kidney extraction and the warm-ischemia times (Ruiz-Deya et al., 2001); a lower intraoperative complication rate and cost (Lindstrom et al., 2002), and facilitated teaching with an additional margin of safety for the donor (Velidedeoglu et al., 2002). Therefore, several centers have started using hand assistance for performing donor nephrectomy since standard laparoscopic donor nephrectomy is a challenging advanced laparoscopic procedure requiring experience with vascular dissection and donor nephrectomy is unique in that the life and health of not one, but two, patients are at stake, and the margin for error is small. Recently, Bargman et al. (2006) compared the early results of standard laparoscopic and hand-assisted laparoscopic donor nephrectomy in a randomized study. Patients completed the SF-36 questionnaire
Health-Related Quality of Life of Living Kidney Donors
118
preoperatively and at 1 month and 3 months of follow-up. The mean preoperative, 1-month and 3-month postoperative QOL scores did not differ significantly between the groups.
5
Other Factors Associated with Quality of Life in Living Kidney Donors
Differences in educational, cultural and socioeconomic backgrounds may influence the quality of life of living kidney donors. Some other factors were observed to influence the QOL in living kidney donors (> Table 118-5).
5.1
Donor’s Gender
Johnson et al. (1999) performed logistic regression analysis to determine the risk factors for those who would not donate again (if it were possible) and for those who found the overall experience very stressful. Of several variables including age, gender, highest level of education, relationship to the recipient, perioperative complications, and recipient survival, female donors (odds ratio, 1.8) were more likely to find the overall experience very stressful although these findings were not statistically significant. Jacobs et al. (1998) also reported that donation was self-reported as more stressful when the donors were female. However, Giessing et al. (2004) and Schover et al. (1997) suggested that donors mostly reported a better QOL than the general population, independent of gender. When analyzed according to gender, male donors had better scores for six of eight items although they scored significantly higher only for the item “General Health”. Female donors scored better in seven of eight items, with significant differences for “Physical function,” “Role-Physical,” “Bodily Pain,” and “General Health.” In addition, in Japanese donors, the donor SF-36 scores did not show any significant difference between men and women (Isotani et al., 2002). By contrast, Lima et al. (2006) reported that donors presented better indexes than the control group when “Vitality” and “General Health” were evaluated. This difference was due to the higher score of the female donors.
5.2
Donor’s Age
Giessing et al. (2004) suggested that life-long psychologic counseling should be offered to help cope with the impact of organ donation on donors’ QOL for younger donors since kidney donation had an overall negative impact on QOL for donors aged 31–40 years at the time of the study. Their scores were worse for all eight items compared with the general German population, but significant differences were observed only for the items “Bodily Pain” and “Vitality.” Jacobs et al. (1998) also reported a decline in the QOL of younger donors. They concluded that young donors probably do not “have time to be in the hospital” because of family and career obligations. Fehrmann-Ekholm et al. (2000) also reported a decline in the QOL of younger Swedish donors. Twenty-three percent thought that the nephrectomy had been troublesome. A higher percentage of young donors had felt that the postoperative period was difficult. By contrast, Isotani et al. (2002) showed that donors aged less than 50 years scored better than older donors. Therefore, conclusions must be made cautiously and can only be tentative. A longitudinal study would be necessary to support this speculation.
2019
2020
118 5.3
Health-Related Quality of Life of Living Kidney Donors
Time Since Donation
Johnson et al. (1999) demonstrated that the QOL in US donors was independent of time since the donation; they found no difference in the mean SF-36 scores between those who had
. Table 118-5 Other factors associated with health-related quality of life in living kidney donors Factors Donor’s gender
Remark More stressful experience in female donors
References Jacobs et al. (1998) Johnson et al. (1999)
Better QOL of donors than the general population independent of gender
Schover et al. (1997) Isotani et al. (2002) Giessing et al. (2004)
Better index of “Vitality” and “General Health” than the Lima et al. (2006) control group due to the due to the higher score of female donors Donor’s age
Declined QOL in younger donors
Jacobs et al. (1998) Fehrmann-Ekholm et al. (2000) Giessing et al. (2004)
Time since donation
Higher QOL scores in donors aged less than 50 years than in older donors
Isotani et al. (2002)
No significant difference according to time since donation
Schover et al. (1997) Johnson et al. (1999) Isotani et al. (2002) Giessing et al. (2004)
Significant decrease of MCS (4 and 12 months) and Smith et al. (2004) three domains (Bodily Pain, General Health and Vitality: 12 months), but not PCS (4 and 12 months) Relationship with the recipient
Best scores in parents who donated to offspring and worst scores in donors unrelated to the recipient
Johnson et al. (1999)
Unfavorable outcome in spouse donors
Lima et al. (2006)
Lower scores in donors who were distant relatives
de Graaf Olson W et al. (2001)
No significant difference according to relationship with Schover et al. (1997) recipient Fehrman-Ekholm et al. (2000) Isotani et al. (2002) Smith et al. (2004) Perioperative complications
Lower QOL scores in donors with postoperative complications – lower QOL scores
Jacobs et al. (1998) Johnson et al. (1999) Giessing et al. (2004)
Health-Related Quality of Life of Living Kidney Donors
. Table 118-5 (continued) Factors
Remark
Outcome for the Worse QOL in cases in which the recipient recipient demonstrated graft loss and died
118 References Johnson et al. (1999) Taghavi et al. (2001) Giessing et al. (2004)
No association of graft function and donors’ QOL
Schover et al. (1997) Isotani et al. (2002) Smith et al. (2003, 2004)
The eight domains are collapsed to create two global components, a physical component score and a mental component score. Scores on the Physical Component Summary are associated with high scores on the Physical Function, Role-Physical, Bodily Pain and General Health scales and low scores on the Role-Emotional and Mental Health scales. For the Mental Component Summary, positive weights are placed on the Mental Health, RoleEmotional, Social Function and Vitality scales, whereas substantial negative ones are placed on the Physical Function and Role-Physical scales. QOL quality of life; MCS Mental Component Summary; PCS Physical Component Summary
donated 5 years before responding after the surgery. Schover et al. (1997) also reported that the MOS-20 scores did not differ significantly according to time since donation. Giessing et al. (2004) also found that the time of follow-up (time passed since donation) did not differ between the donors scoring better or worse than the general German population. The donor SF-36 scores did not show a significant difference based on the time since the donation in Japanese donors (Isotani et al., 2002). However, in prospective studies conducted in Australia, the MCS of the SF-36 in living kidney donors fell significantly postoperatively (Smith et al., 2003, 2004). These findings suggest that donors should be alerted to possible psychosocial impairment and monitored postoperatively.
5.4
Relationship with the Recipient
Johnson et al. (1999) reported that when analyzed by donor-recipient relationship, parents who donated to offspring had the best scores and donors unrelated to the recipient, the worst; however, all scores are still the same or better than for the US general population. In the logistic regression analysis, relatives other than first degree were more likely to say they would not donate again, if it were possible, or to find the overall experience very stressful. Lima et al. (2006) also reported that there was a discrepancy in several components of the SF-36 scale among donors according to their relationship with the recipient. They found an unfavorable outcome of spouse donors. de Graaf Olson and Bogetti-Dumlao A (2001) reported that when comparing scores using the relationship of the donor to the recipient, the mean scores of the donors who were “distant relatives” were found to be lower in seven of the eight categories. In a study by Jacobs et al. (1998), relatives other than immediate family members (extended relatives) were more likely to be among the 4% who said they would not donate again. However, for both German (Fehrman-Ekholm et al., 2000) and Japanese (Isotani et al., 2002) donors, the donor SF-36 scores did not show any significant difference according to the relationship with the recipient. The QOL scores also did not differ significantly according to the kinship of the donor and the recipient in a study by Schover et al. (1997). Furthermore, a prospective study also revealed no influence of the kinship with the recipient on donor scores (Smith et al., 2004).
2021
2022
118 5.5
Health-Related Quality of Life of Living Kidney Donors
Perioperative Complications
Donor and recipient complications had a significant impact on German donors’ QOL (Giessing et al., 2004). Donors with postoperative complications had worse scores on all SF-36 items. The significance level was reached for “Physical Function,” “Social Function,” “Role-Emotional” and “Mental Health.” In addition, donors whose recipient had faced a complication scored worse for all items except for “Physical Function,” with significantly worse QOL for the items “Vitality,” “Social Function” and “Bodily Pain.” However, the donors’ willingness to donate again (93.4%) or recommend living donor kidney transplanation (92.4%) was high, irrespective of complications. Jacobs et al. (1998) and Johnson et al. (1999) also suggested that donors who experienced perioperative complications were more likely to find the overall experience very stressful.
5.6
Outcomes for the Recipient
Giessing et al. (2004) reported that German donors’ QOL strongly depended on the QOL of the recipients after kidney transplantation, which is reflected by the close association of donor’s QOL and recipient’s outcome. In their study, the risk of negative effects on the donor was up to ten times higher in cases in which the recipient demonstrated graft loss and died. Johnson et al. (1999) also found that in the logistic regression analysis, the US donors whose recipient died within 1 year of transplantation were more likely to say they would not donate again, if it were possible, or to find the overall experience very stressful. In addition, Westlie et al. (1993) found that donors whose grafts were unsuccessfully transplanted had a worse QOL than “successful” donors. Other studies reported that in a situation with graft failure or recipient death, more than 11% of donors developed suicidal thoughts and 15% of the donors were found to develop depression (Taghavi et al., 2001). Negative attitudes were more frequent among unsuccessful than successful donors, even if most of the donors were happy after their donation. In this regard, Hirvas et al. (1980) pointed out that an unsuccessful operation is always followed by psychological complications. However, others found no association of graft function and the donors’ QOL (Isotani et al., 2002; Schover et al., 1997; Smith et al., 2003, 2004), and thus, the results of the different studies remain controversial. In a prospective study, there was no significant difference in the postoperative MCS of donors grouped according to whether or not their recipient’s transplant had failed (Smith et al., 2003). In a subsequent study, Smith et al. (2004) suggested that it was the emotional state of the recipient rather than the physical state (as measured by length of stay, creatinine level, graft failure, and PCS) that was associated with the donor psychosocial outcome. In Japanese donors, the outcomes for the recipients after donation did not correlate with the donor SF-36 scores (Isotani et al., 2002).
6
Conclusions
The results of methodologically appropriate studies have provided evidence that the donor QOL is at least comparable to that of the general population. Some studies have even indicated a higher QOL or well-being for the kidney donors compared to the general population. However,
Health-Related Quality of Life of Living Kidney Donors
118
increased rates of donors with mental distress and intra-familial conflicts emphasize the need for a careful selection process. Careful donor selection, with appropriate pretransplantation psychiatric consulting, allows those with a normal QOL to donate without significant consequences on their physical and psychological status. Furthermore, regular postdonation psychosocial screening and provision of specific interventions to those in need are recommended. In addition, most studies evaluating the impact of kidney donors’ QOL have limitations such as a small cohort size, a retrospective study design, unmatched reference populations, or low response rates. The retrospective nature of the majority of living donor studies is a significant limitation. The retrospective, cross-sectional study design required patients to recall specific information during their recuperation period following their surgery date. Therefore, the study results may bear some degree of recall bias. Furthermore, retrospective studies have shown that although the majority of donors report that the experience of donation was positive, some were troubled and some even regretted having donated. Nonetheless, these studies suggest the hypotheses that require further evaluation in welldesigned prospective studies. Differences in educational, cultural and socioeconomic backgrounds may influence the QOL of living kidney donors. Therefore, in future studies, consideration of additional important factors such as religion, culture, customs, environment, and other factors influencing the QOL in different countries should be included in the study design. These future prospective studies will contribute to our knowledge of factors that influence the health-related QOL of living kidney donors. In addition, studies on a larger cohort will facilitate identification of the risk factors for dysfunction, which in turn may lead to the establishment of valid screening procedures.
Summary Points Although living kidney donation offers many potential benefits for the recipient, the clinical benefits for the donor are less clear.
To date, the SF-36 has been widely used to measure the QOL of living kidney donors, but
there is no consensus on the “gold standard” instrument to measure the QOL of this population. Most studies have suggested that living kidney donors have similar or higher scores on the QOL questionnaires compared with a healthy general population. Evolution of surgical techniques in transplantation including the mini-incision and laparoscopy has made kidney living donation a more attractive option for patients and their potential donors. The influence of surgical technique on the QOL of living kidney donors remains controversial since donors are highly motivated and committed to donation. Several variables have been proposed as influencing factors on the QOL in living kidney donors but these are not conclusive. Careful donor selection, with appropriate pre-transplantation psychiatric counseling as well as regular post-donation psychosocial screening and the provision of specific interventions to those in need are recommended.
2023
2024
118
Health-Related Quality of Life of Living Kidney Donors
References Aguiar WF, Passerotti CC, Claro JF, Almeida CJ, Gattas N, Cedenho AP, Pestana JO, Ortiz V. (2007). Clin Transplant. 21: 269–276. Bargman V, Sundaram CP, Bernie J, Goggins W. (2006). J Endourol. 20: 717–722. Bergman S, Feldman LS, Mayo NE, Carli F, Anidjar M, Klassen DR, Andrew CG, Vassiliou MC, Stanbridge DD, Fried GM. (2005). Am J Transplant. 5: 2489–2495. Bruzzone P, Pretagostini R, Poli L, Rossi M, Berloco PB. (2005). Transplant Proc. 37: 2436–2438. Buell JF, Lee L, Martin JE, Dake NA, Cavanaugh TM, Hanaway MJ, Weiskittel P, Munda R, Alexander JW, Cardi M, Peddi VR, Zavala EY, Berilla E, Clippard M, First MR, Woodle ES. (2005). Clin Transplant. 19: 102–109. Bullinger M, Alonso J, Apolone G, Leplege A, Sullivan M, Wood-Dauphinee S, Gandek B, Wagner A, Aaronson N, Bech P, Fukuhara S, Kaasa S, Ware JE. Jr. (1998). J Clin Epidemiol. 51: 913–923. Cabrer C, Oppenhaimer F, Manyalich M, Paredes D, Navarro A, Trias E, Lacy A, Rodriguez-Villar C, Vilarrodona A, Ruiz A, Gutierrez R. (2003). Transplant Proc. 35: 1631–1632. Carli F, Mayo N. (2001). Br J Anaest. 87: 531–533. Chen CH, Chen Y, Chiang YJ, Wu CT, Chen HW, Chu SH. (2004). Transplant Proc. 36: 1920–1921. Dahm F, Weber M, Muller B, Pradel FG, Laube GF, Neuhaus TJ, Cao C, Wuthrich RP, Thiel GT, Clavien PA. (2006). Nephrol Dial Transplant. 21: 2563–2568. de Graaf Olson W, Bogetti-Dumlao A. (2001). Prog Transplant. 11: 108–115. Fehrman-Ekholm I, Brink B, Ericsson C, Elinder CG, Duner F, Lundgren G. (2000). Transplantation. 69: 2067–2071. Giessing M, Reuter S, Deger S, Tullmann M, Hirte I, Budde K, Fritsche L, Slowinski T, Dragun D, Neumayer HH, Loening SA, Schonberger B. (2005). Transplant Proc. 37: 2011–2015. Giessing M, Reuter S, Schonberger B, Deger S, Tuerk I, Hirte I, Budde K, Fritsche L, Morgera S, Neumayer HH, Loening SA. (2004). Transplantation. 78: 864–872. Hirvas J, Enckell M, Kuhlback B. (1980). Acta Med Scand. 208: 285–287. Isotani S, Fujisawa M, Ichikawa Y, Ishimura T, Matsumoto O, Hamami G, Arakawa S, Iijima K, Yoshikawa N, Nagano S, Kamidono S. (2002). Urology. 60: 588–592. Jacobs C, Johnson E, Anderson K, Gillingham K, Matas A. (1998). Adv Ren Replace Ther. 5: 89–97.
Jackobs S, Becker T, Luck R, Jager MD, Nasham B, Gwinner W, Schwarz A, Klempnauer J, Neipp M. (2005). World J Urol. 23: 343–348. Johnson EM, Anderson JK, Jacobs C, Suh G, Humar A, Suhr BD, Kerr SR, Matas AJ. (1999). Tranaplantation. 67: 717–721. Jordan J, Sann U, Janton A, Gossmann J, Kramer W, Kachel HG, Wilhelm A, Scheuermann E. (2004). J Nephrol. 17: 728–735. Kok NF, Lind MY, Hansson BM, Pilzecker D, Mertens zur Borg IR, Knipscheer BC, Hazebroek EJ, Dooper IM, Weimar W, Hop WC, Adang EM, van der Wilt GJ, Bonjer HJ, van der Vilet JA, IJzermans JN. (2006). Br Med J. 333: 221–224. Kuo PC, Johnson LB. (2000). Transplantation. 69: 2211–2213. Lima DX, Petroianu A, Hauter HL. (2006). Quality of life and surgical complications of kidney donors in the late post-operative period in Brazil. Nephrol Dial Transplant. 21: 3238–3242. Lindstrom P, Haggman M, Wadstrom J. (2002). Handassisted laparoscopic surgery (HALS) for live donor nephrectomy is more timeand cost-effective than standard laparoscopic nephrectomy. Surg Endosc. 16: 422–425. Lumsdaine JA, Wray A, Power MJ, Jamieson NV, Akyol M, Andrew Bradley J, Forsythe JL, Wigmore SJ. (2005). Transpl Int. 18: 975–980. Najarian JS. (2005). Transplant Proc. 37: 3592–3594. Neipp M, Jackobs S, Becker T, zu Vilsendorf AM, Winny M, Lueck R, Klempnauer J, Nashan B. (2004). Transplantation 78: 1356–1361. Ozcurumez G, Tanriverdi N, Colak T, Emiroglu R, Zileli L, Haberal M. (2004). Transplant. Proc. 36: 114–116. Pace KT, Dyer SJ, Stewart RJ, Honey RJ, Poulin EC, Schlachta CM, Mamazza J. (2003). Surg Endosc. 17: 143–152. Perry KT, Freedland SJ, Hu JC, Phelan MW, Kristo B, Gritsch AH, Rajfer J, Schulam PG. (2003). J Urol. 169: 2018–2021. Reimer J, Rensing A, Haasen C, Philipp T, Pietruck F, Franke GH. (2006). Transplantation 81: 1268–1273. Rodrigue JR, Cross NJ, Newman RC, Widows MR, Guenther RT, Kaplan B, Morgan MA, Howard RJ. (2006). Prog Transplant. 16: 162–169. Ruiz-Deya G, Cheng S, Palmer E, Thomas R, Slakey D. (2001). J Urol. 166: 1270–1273. Schnitzbauer AA, Hornung M, Seidel U, Kruger B, Kramer BK, Schlitt HJ, Obed A. (2007). Clin Transplant. 21: 235–240.
Health-Related Quality of Life of Living Kidney Donors Schover LR, Streem SB, Boparai N, Duriak K, Novick AC. (1997). J Urol. 157: 1596–1601. Schweitzer EJ, Wilson J, Jacobs S, Machan CH, Philosophe B, Farney A, Colonna J, Jarrell BE, Barlett ST. (2000). Ann Surg. 232: 392–400. Simforoosh N, Basiri A, Tabibi A, Shakhssalim N, Hosseini Moghaddam SM. (2005). BJU Int. 95: 851–855. Smith GC, Trauer T, Kerr PG, Chadban SJ. (2003). Transplantation. 76: 807–809. Smith GC, Trauer T, Kerr PG, Chadban SJ. (2004). Transplantation. 78: 1384–1389. Taghavi R, Mahdavi R, Toufani H. (2001). Transplant Proc. 33: 2636–2637. Tanriverdi N, Ozcurumez G, Colak T, Duru C, Emiroglu R, Zileli L, Haberal M. (2004). Transplant Proc. 36: 117–119. Toronyi E, Alfoldy F, Jaray J, Remport A, Mathe Z, Szabo J, Gati Z, Perner F. (1998). Transplant Int. 11 (Suppl 1): S481–S483.
118
Troppmann C, Johnston WK. 3rd, Pierce JL, McVicar JP, Perez RV. (2006). Pediatr Nephrol. 21: 1052–1054. Vastag B. (2003). J Am Med Assoc. 290: 181–182. Velidedeoglu E, Williams N, Brayman KL, Desai NM, Campos L, Palanjian M, Wockik M, Bloom R, Grossman RA, Mange K, Barker CF, Naji A, Markmann JF. (2002). Transplantation. 74: 169–172. Virzi A, Signorelli MS, Veroux M, Giammarresi G, Maugeri S, Nicoletti A, Veroux P. (2007). Transplant Proc. 39: 1791–1793. Vlaovic PD, Devins GM, Abbey S, Wright E, Robinette MA. (1999). Can J Urol. 6: 859–864. Weizer N, Weizman A, Shapira Z, Yussim A, Munitz H. (1989). Psychother Psychosomat. 51: 216–219. Westlie L, Fauchald P, Talseth T, Jakobsen A, Flatmark A. (1993). Nephrol Dial Transplant. 8: 1146–1150. Zargooshi J. (2001). J Urol. 166: 1790–1799.
2025
119 Quality of Life and Tryptophan Degradation D. Fuchs . K. Schroecksnadel . G. Neurauter . R. Bellmann-Weiler . M. Ledochowski . G. Weiss 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2028
2 Tryptophan Metabolism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2029 2.1 Tryptophan Degradation: TDO, IDO, and INDOL 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2030 3
Tryptophan Metabolism in “Healthy Individuals” and in Chronic Diseases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2031
4
Immune-Mediated Tryptophan Degradation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2032
5
Immune Modulation by IFN-g and IDO: Immunodeficiency in Chronic Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2034
6 Tryptophan Degradation and Quality of Life . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2034 6.1 Possibilities to Improve Patients’ Quality of Life . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2035 7 Weight Loss and Tryptophan Degradation in Chronic Disease . . . . . . . . . . . . . . . . . . 2036 7.1 General Mechanisms of Tumor Cachexia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2036 7.2 Tryptophan and Weight Loss . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2037 8
Fatigue and Anemia of Chronic Disease and Enhanced Tryptophan Catabolism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2037 8.1 Fatigue and Tryptophan Metabolism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2038 8.2 Anemia of Chronic Disease and Cellular Immune Activation . . . . . . . . . . . . . . . . . . . . . 2039 8.3 Anemia of Chronic Disease and Tryptophan Degradation . . . . . . . . . . . . . . . . . . . . . . . . . 2040 9 “Neuropsychiatric” Side Effects of Enhanced Tryptophan Catabolism . . . . . . . . . . 2040 9.1 Depression and Tryptophan Metabolism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2041 9.2 Cognitive Impairment and Tryptophan Degradation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2042 10
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2043 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2043 Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2044
#
Springer Science+Business Media LLC 2010 (USA)
2028
119
Quality of Life and Tryptophan Degradation
Abstract: Patients suffering from autoimmune disease, cancer or chronic inflammatory diseases mostly suffer from severe “complications” of their underlying illness, like weight loss, physical inefficiency, immunodeficiency, anemia, chronic fatigue, mood disorders or even depression. All these symptoms can severely impair the quality of life of patients. The percentage of patients with chronic diseases complaining about poor quality of life is high, irrespective of their underlying disease. Interestingly, most patients suffering from chronic inflammatory disease have a strongly disturbed > tryptophan metabolism. Within cellular immune response the essential amino acid tryptophan is degraded by the enzyme > indoleamine-2,3-dioxygenase (IDO), leading to lowered tryptophan serum/plasma concentrations and increased levels of tryptophan catabolites like kynurenine. Tryptophan is not only necessary for the growth and proliferation of various cells as well as pathogens, it is also the precursor of the important neurotransmitters serotonin (5-hydroxytryptamin) and nicotinamide adenine dinucleotide (NAD). Increased > tryptophan degradation within chronic inflammatory cascades thus leads to a diminished tryptophan availability, which might not only decrease the immunoresponsiveness of patients, but may also influence their mood, physical strength and haematopoiesis. In fact, immune-mediated tryptophan degradation is supposed to contribute importantly to the development of fatigue, weight loss, and neuropsychiatric disorders. This review provides an overview, how enhanced IDO-activation might contribute to the development of various symptoms impairing the quality of life of patients with chronic disease. List of Abbreviations: ACMSA, aminocarboxymuconic semialdehyde; ATP, adenosine triphosphate; GCH, guanine triphosphate cylohydrolase I; HIV, human immunodeficiency virus infection; IDO, indoleamine 2,3-dioxygenase; IFN-g, > interferon-g; INDOL 1, indoleaminepyrrole 2,3-dioxygenase-like 1; Kyn/trp, kynurenine to tryptophan ratio; LPS, lipopolysaccharide; NAD, nicotinamide adenine dinucleotide; TDO, tryptophan pyrrolase; > Th1 type immune response, T-helper cell type 1 immune response; TNF-a, tumor necrosis factor-a; 5-HT, 5-hydroxytryptamin (= serotonin); 1-MT, 1-methyl-tryptophan
1
Introduction
Tryptophan metabolism is disturbed strongly in patients suffering from chronic inflammatory disease. The catabolism of the essential amino acid tryptophan is enhanced, lowered tryptophan serum/plasma concentrations and increased levels of the tryptophan catabolites like kynurenine are encountered in patients. Tryptophan degradation takes place as a consequence of on-going immune activation, within cellular immune response the enzyme indoleamine2,3-dioxygenase (IDO) is induced to convert tryptophan to N-formyl-kynurenine (which is then further catabolized to kynurenine). Patients suffering from autoimmune disease, cancer or chronic inflammatory diseases mostly suffer from severe “complications” of their underlying illness, like weight loss, immunodeficiency, anemia, chronic fatigue, mood disorders or even depression. All these symptoms can severely impair the quality of life of patients. Nowadays it becomes more and more important not only to treat the underlying disease, but also to improve the patient’s quality of life. The percentage of patients with chronic diseases complaining about poor quality of life is high, irrespective of their underlying disease. Interestingly, patients suffering from different kinds of chronic disease present with similar symptoms, and thus a key role of inflammatory processes and immune activation cascades has been proposed for several years.
Quality of Life and Tryptophan Degradation
119
Independently of specific pathways important in the progression of tumors or autoimmune disease, inflammatory cascades, which are activated in all these diseases, might be involved in the development of neuropsychiatric as well as physical “side effects” of chronic diseases. Among other important biochemical pathways, which are disturbed in chronic disease, immunemediated tryptophan degradation is supposed to contribute importantly to the development of the above described symptoms. Tryptophan is not only an essential amino acid, which is necessary for the growth and proliferation of various cells as well as pathogens, it is also the precursor of the important neurotransmitters serotonin (5-hydroxytryptamin) and NAD. This article provides an overview, how enhanced tryptophan degradation by IDO-activation might contribute to the development of various symptoms of patients with chronic disease, thus impairing the quality of life of patients with chronic disease. First of all, tryptophan metabolism will be adumbrated shortly, afterwards the role of IDO-activation in chronic disease is described and data regarding the role of tryptophan degradation in the development of several symptoms are presented.
2
Tryptophan Metabolism
Tryptophan is an essential amino acid required for the biosynthesis of proteins. Furthermore tryptophan serves as precursor for several biologically important compounds like serotonin or nicotinamide adenine dinucleotide (NAD). The least abundant of all amino acids is characterized by an indole ring, which cannot be synthesized by animals and humans. Tryptophan therefore is either obtained from protein degradation, or must be supplied by diet, the required daily dose lies between 175 and 250 mg for adults (Peters, 1991). Ingested proteins are hydrolyzed into the constituent amino acids in the digestive system, and dietary tryptophan is then delivered to the hepatic portal system (Moffett and Namboodiri, 2003). There it can either be distributed through the blood stream or can be directly used for protein synthesis. Inadequate tryptophan ingestion leads to disturbances of protein metabolism and consecutively, to the loss of muscles and weight (Brown et al., 1996). Inadequate tryptophan supply by diet confers a major problem in the third world, whereas an average diet in the Western world contains about 3 to 4 times higher tryptophan doses than needed (Brown et al., 1991), e.g. more than 200 mg tryptophan is supplied by a single serving of meat, fish or cheese. Therefore, in the Western world, decreased tryptophan plasma/serum concentrations are mostly not due to dietary restriction, but are rather a consequence of enhanced tryptophan degradation. Tryptophan is usually bound to human serum albumin, however, if this binding is saturated it can lead to elevated levels of free tryptophan. Free tryptophan, which constitutes only about 5% of the body tryptophan, can cross the blood brain barrier via an amino acid transport system, which transports the neutral amino acids tryptophan, tyrosine, phenylalanine, valine and isoleucine. In the brain (10–20%) as well as in entero-chromaffine cells of the gut (80–90%) tryptophan is hydroxylated to 5-hydroxytryptophan by tryptophan 5-hydroxylase (EC 1.14.16.4), which is then further decarboxylated to the neurotransmitter 5-hydroxytryptamin (5-HT = serotonin). Several studies indicate that the concentration of brain serotonin is dependent on the availability of its precursor tryptophan (Review by Neumeister, 2003), even if only 1% of the available tryptophan is really catabolized by this pathway.
2029
2030
119
Quality of Life and Tryptophan Degradation
Very small amounts of tryptophan are also catabolized in the large intestine by bacteria (to indole derivatives), resulting in the formation of indican, which is then excreted in the urine. However, the vast majority of tryptophan is catabolized via the so called kynurenine pathway leading to the formation of nicotinic acid, niacin and nicotinamide adenine dinucleotides as end products (see > Figure 119-1.) . Figure 119-1 Tryptophan metabolism
2.1
Tryptophan Degradation: TDO, IDO, and INDOL 1
Three enzymes catalyze the oxidation of tryptophan to N-formylkynurenine – a process, which is irreversible, as cleavage of the five-membered indole ring occurs (see also > Figure 119-1). Tryptophan 2,3-dioxygenase (TDO, tryptophan pyrrolase, EC 1.13.1.2) and indoleamine-2,3dioxygenase (IDO; EC 1.13.11.42) are the two established enzymes, whereas the function of the recently identified enzyme indoleamine-pyrrole 2,3-dioxygenase-like 1 (INDOL 1; EC 1.13.11) is not fully clarified. INDOL 1 is expressed primarily in the placenta in human beings (mouse: kidney) and shares 43% amino acid identities with IDO, genes of both enzymes are located on the short arm of human chromosome 8 separated only by a short intergenic region (see also Murray, 2007). TDO, on the other hand, is located in the liver and oxidizes excess dietary tryptophan to generate energy, aminocarboxymuconic semialdehyde (ACMSA) is the precursor of adenosine
Quality of Life and Tryptophan Degradation
119
triphosphate (ATP). TDO expression is induced by corticosteroids, while IDO expression is mainly stimulated by proinflammatory cytokines, IFN-g is the most potent trigger for IDO activation. In contrast to TDO, which can only use tryptophan as substrate, IDO can utilize also other indoleamine derivatives including L-, and D-tryptophan, tryptamine, 5-hydroxytryptophan and serotonin. IDO can be induced in many different cells (Werner et al., 1989), among them antigen-presenting cells such as monocyte-derived macrophages, dendritic cells and fibroblasts. Under inflammatory conditions, IFN-g strongly induces IDO-mediated tryptophan degradation, resulting in increased kynurenine and decreased tryptophan serum/plasma concentrations. In vivo, enhanced cytokine-induced degradation of tryptophan is observed, whenever the cellular immune system is activated, and a decrease of serum tryptophan and a parallel increase of kynurenine and other catabolites of tryptophan can be detected (see also Wirleitner et al., 2003). Additional determination of immune activation markers like IFN-g, cytokines or neopterin enables to distinguish between TDO- and IDO-mediatated tryptophan oxidation (Moroni, 1999; Widner et al., 1997). If kyn/trp correlates with the extent of immune activation, immune-mediated IDO-activation can be assumed. Tryptophan degradation by the above mentioned enzymes leads to the formation of N-formylkynurenine, which can be further metabolized to kynurenine and consecutively, to a series of other compounds, depending on the enzyme repertoire of cells (see also > Figure 119-1). Some of these catabolites, e.g. quinolinic acid are proposed to be neurotoxic (Moroni, 1999), probably causing convulsion and excitoxicity. In the liver, quinolinic acid can be converted to nicotinic acid which enters the NAD metabolism and provides metabolic energy.
3
Tryptophan Metabolism in “Healthy Individuals” and in Chronic Diseases
In healthy individuals, the “tryptophan status,” as characterized by protein and serotonin synthesis and protein degradation, is balanced. Excess dietary tryptophan is converted to generate the main energy donors of cells: ATP + CO2 and water. Concentration of free serum tryptophan ranges from 40 to 100 mmol/L in healthy individuals (mean SD: 73 15 mmol/L) and is dependent to some extent on dietary intake (Peters, 1991; Widner et al., 1997). Women usually present with 20% lower serum/plasma tryptophan concentrations than men. Tryptophan catabolism to kynurenine is constitutively active in the liver, therefore an average of 1.92 0.58 mmol/L serum kynurenine is found in healthy individuals (Widner et al., 1997). Calculation of the ratio of the enzyme substrate divided by the resulting product, namely, calculating the kynurenine (kyn) to tryptophan (trp) ratio (kyn/trp) allows to assess the degree of tryptophan degradation. Kyn/trp is suited better to compare tryptophan metabolism in different individuals than absolute tryptophan or kynurenine concentrations. Decreasing plasma tryptophan concentrations in parallel with increasing plasma levels of proinflammatory cytokines, kynurenine or other tryptophan degradation products are indicative of enhanced IDO activation in patients, while reduced dietary intake of tryptophan leads to lowered endogenous tryptophan and low kynurenine concentrations, respectively. However, in inflammatory conditions, which are mostly observed in patients with chronic disease, the role of TDO-mediated oxidation is negligible, as tryptophan catabolism by IDO activation occurs at a high rate. In accordance with this finding, patients with, e.g., autoimmune disease, cancer, neurodegenerative disease and coronary artery disease present with decreased serum/plasma
2031
2032
119
Quality of Life and Tryptophan Degradation
tryptophan, increased kynurenine and inflammation/immune activation marker concentrations indicating immune-mediated IDO-activation (Schrocksnadel et al., 2006). Various chronic diseases, in which enhanced IDO-activation has been demonstrated are listed in > Table 119-1.
. Table 119-1 Chronic diseases associated with enhanced IDO-mediated tryptophan degradation Malignant diseases
Hematological disorders Lung cancer Gynecological tumors Gastrointestinal tumors Malignant melanoma
Autoimmune diseases
Systemic lupus erythematosus Rheumatoid arthritis Systemic sclerosis
Viral diseases
HIV-infection
Bacterial diseases
Lyme neuroborreliosis
Cardiovascular disease
Coronary artery disease
Neurodegenerative disorders
Alzheimer’s disease
Cardiac disorders Huntington’s disease Parkinsons’s disease
4
Immune-Mediated Tryptophan Degradation
Different roles have been ascribed to the enzyme IDO, which was first discovered in 1967 by Prof. Hayaishi and co-workers. Since the late 1970s IDO has been regarded as key enzyme of the host defense armament against a variety of infectious pathogens (Taylor and Feng,). Intracellular tryptophan depletion by IDO was observed to exert anti-parasitic, anti-viral and anti-chlamydial effects (see also MacKenzie et al., 2007). In most cases, IDO induction is observed in infected tissues only. By withdrawal of the essential amino acid from the micro-environment, protein biosynthesis and thus proliferation of pathogens like influenza or Chlamydia (see Review MacKenzie et al., 2007), is arrested. However, also systemic IDO induction can occur, in mice endotoxin shock induced by intraperitoneal interleukin-12 injection, was demonstrated to induce tryptophan degradation by IDO strongly (Takikawa et al., 1986). Enhanced IDO-mediated tryptophan degradation thus was identified early as very potent mechanism of the host’s immune system to inhibit pathogen proliferation. However, interestingly, since the 1990s, also other, rather unexpected “target cells” for IDO have attracted interest: Host T-cell proliferation is suppressed very efficiently by IDO-activation (Munn et al., see also > Figure 119-2). In 1996 we and others have described an increasing activity of IDO throughout human pregnancy (Schro¨cksnadel et al., 1996), which might relate to the memory loss, which seems to develop frequently in pregnant women (Sharp et al., 1993).
Quality of Life and Tryptophan Degradation
119
. Figure 119-2 Interferon-g-mediated pathways and feedback regulation by indoleamine-2,3-dioxygenase (IDO). Within cellular immune response T-lymphocytes release the proinflammatory cytokine interferon-g (IFN- g), which induces several important biochemical pathways in human monocytes: formation of the pteridine neopterin, release of other proinflammatory cytokines like tumor necrosis factor-a (TNF-a) and induction of the enzyme indoleamine-2,3-dioxygenase (IDO). IDO degrades the essential amino acid tryptophan to kynurenine, thus depleting tryptophan from the local environment. Tryptophan deprivation represents a very important mechanism to inhibit the proliferation of activated T-lymphocytes, leading to a down-regulation of Th1-type immune response
A few years later, in 1998 and 1999 Munn and co-workers (Munn et al., 1998, Munn et al., 1999) could show that IDO activation during pregnancy is required to prevent rejection of the allogenic fetus by the maternal immune system. These experiments led to extensive studies investigating the function of IDO in various immuno-competent cells, because tryptophan degradation by IDO was recognized as biochemical pathway of high impact and with a great potential for the understanding, treatment and prevention of human disease. Consecutively, a much deeper understanding of the role of IDO-activation in the interaction of different immunocompetent cells has been established (see also Mellor and Munn, 2004). Also the relationship between cancer and IDO-activation was reevaluated, as IDO can be expressed by both, tumor and host immune cells (Uyttenhove et al., 2003). IDO-expression by malignant tumor cells was suggested to enable cancer progression by immune escape, as local tryptophan depletion inhibits T-cell response against the tumor. However, although increased systemic and local IDO activation as reflected by enhanced kyn/trp or IDO expression has been observed in many different tumors (see also > Table 119-1), it is not always clear, whether enhanced kynurenine formation is really due to spontaneous IDO activity in tumor cells or due to immune-mediated IDO activation. Correlations between Th1-type immune response marker neopterin and kyn/trp in patients with various kinds of cancer rather indicate that immune-mediated tryptophan catabolism takes place, resulting in down-regulation of immune cascades or consecutively, tumor tolerance.
2033
2034
119 5
Quality of Life and Tryptophan Degradation
Immune Modulation by IFN-g and IDO: Immunodeficiency in Chronic Disease
Pro-inflammatory cytokines like IFN-g and tumor necrosis factor alpha (TNF-a) orchestrate immune activation cascades by induction of various biochemical pathways in immunocompetent cells. IDO induction during acute infection down-regulates these cascades efficiently, thus providing feedback regulation (> Figure 119-2). IFN-g induces IDO-activation strongest, but also other cytokines, mostly interferons, and lipopolysaccharide induce or superinduce kynurenine formation (Taylor and Feng, 1991; Werner et al., 1987). Local and also systemic > tryptophan deprivation represents the most important immunomodulatory effect of IDO. However, also toxic down-stream metabolites of kynurenine have been proposed to have immunosuppressive properties (see also below in further detail). When growth of viruses and chlamydia was suppressed by IDO expressing cells in vitro, tryptophan supplementation enabled pathogens to proliferate again. Similarly, addition of tryptophan to T-cells reversed IDO mediated inhibition of lymphocyte proliferation (Mellor and Munn, 2004; Sakurai et al., 2002; von Bubnoff et al., 2002). Tryptophan withdrawal does not only induce cell-cycle arrest, e.g. of in vitro-activated human and mouse T cells, but may also render arrested cells more sensitive to apoptosis (Mellor and Munn, 2004). In fact, stimulated cells lacking an essential compound like tryptophan for their proliferation, will preferentially undergo apoptosis. Furthermore, also tryptophan metabolites such as quinolinic acid, picolinic acid and 3-hydroxyanthranilic acid may exert antiproliferative and cytotoxic effects: human T cells were found to be sensitive to higher doses of exogenously added kynurenine metabolites (Frumento et al., 2002; Terness et al., 2002). Although doses of metabolites used in these experiments were much higher than concentrations detected in vivo in the serum/plasma of patients, high local concentrations of metabolites might influence surrounding cells to some extent. Additionally, tryptophan catabolites are able to modulate toxic radical formation by either scavenging or stabilizing radicals (Weiss et al., 2002), more detailed information about immunoregulatory activities of kynurenine metabolites is given elsewhere (Grohmann et al., 2003; Moffett and Namboodiri, 2003).
6
Tryptophan Degradation and Quality of Life
Apart from its immunoregulatory properties, IDO appears to represent a key enzyme influencing the quality of life of patients suffering from chronic disease. Patients with chronic diseases mostly complain about a strongly deteriorated quality of life. Irrespective of their underlying disease, patients present with similar symptoms, namely weight loss, immunodeficiency, anemia, chronic fatigue, mood disorders or even depression. All these symptoms can severely impair the quality of life of patients. Interestingly, the development of most symptoms has been related with immune activation and enhanced IDO-activation. Thus, the following chapters will present data of studies investigating the relationship between enhanced tryptophan degradation and the above mentioned symptoms, namely weight loss, fatigue, anemia and mood disorders (see also > Figure 119-3). However, before really going into detail, also data about the association
Quality of Life and Tryptophan Degradation
119
. Figure 119-3 Relationship between immune activation, tryptophan metabolism and symptoms impairing patients’ quality of life in chronic disease
between increased tryptophan catabolism and the quality of life of different patient populations shall be presented: In patients suffering from colorectal cancer, immune-mediated tryptophan degradation was found to be significantly correlated with impaired quality of life, serum tryptophan concentrations were predictive for impaired quality of life scores determined by two different questionnaires (the “Rotterdam Symptom Checklist physical symptom” and the “Sickness Impact Profile”). Also in patients suffering from all kinds of malignant tumors, tryptophan metabolism was recently found to be associated with patients’ assessment of their quality of life. Patients’ quality of life was furthermore dependent on their physical performance status determined by Karnofsky index and ECOG and of course, on the stage of tumor progression, which was also reflected by tryptophan concentrations. Similarly in patients with HIV-infection, IDO-activation in patients increases with progressive CDC-stage, and quality of life as well as depression scores of patients were significantly associated with the degree of immune activation. In patients without antidepressant treatment, enhanced tryptophan degradation coincided with moderate to severe depression scores, however, quality of life scores were not related with tryptophan metabolism in this relatively heterogenous population. Th1 type immune response marker neopterin on the other hand was predictive for both, an impaired quality of life and the presence of depression (Schroecksnadel et al., 2008).
6.1
Possibilities to Improve Patients’ Quality of Life
As depression/psychic distress often also enforces a further impairment of patients’ quality of life, diagnosis of mood disorders and adequate treatment is important. Nowadays it is not only essential to treat the underlying disease, i.e. cancer or HIV, but also to improve the patient’s quality of life. It is well established that a multi-disciplinary e.g. oncologic therapy concept is able to improve the patient’s attitude towards his disease and thus, also improve his
2035
2036
119
Quality of Life and Tryptophan Degradation
compliance during therapy. Symptomatic treatment of certain symptoms is often performed, if causal treatment is either only partially or not effective and therefore also cannot avert “complications.” With regard to the fact that tryptophan degradation has been related with nearly all symptoms impairing the quality of life (as described in the following chapters), it would of course be tempting to “correct” tryptophan metabolism, e.g. by supplementation of tryptophan to depleted patients. However, tryptophan supplementation is like a two-edged sword, it could in fact also worsen the situation of patients. Tryptophan is depleted in chronic disease slowing down the proliferation of pathogens or tumor cells and thus disease progression by nutrient starvation. By supplying tryptophan, the immunoresponsiveness of T-cells might improve, but also the underlying disease could be fed, and potentially neurotoxic tryptophan catabolites may further accumulate. Another approach targeting IDO directly is inhibition of IDO by competitive inhibitors. This approach has been used successfully in animal models, e.g. in mice. Treatment of mice with 1-methyl tryptophan was used first in experiments investigating maternal tolerance of the allogeneic fetus (Munn et al., 1999). Based on these fundamental experiments IDO inhibition was also performed in transplanted mice (Brandacher et al., 2007). Furthermore, treatment of mice with the competitive IDO inhibitor 1-methyl-tryptophan (1-MT) is able to prevent IDO-mediated depression like behavior (Dantzer et al., 2008), suggesting that IDO-inhibition might possibly also have beneficial effects on other symptoms associated with enhanced tryptophan degradation, that might affect the quality of life of patients. Still, IDO-blockade by 1-MT or other known IDO inhibitors like tryptophan-beta-(3-benzofuranyl)-DL-alanine and beta-(3-benzo(b)thienyl)-DLalanine is supposed to be too toxic for human beings and data of mice are also so far limited.
7
Weight Loss and Tryptophan Degradation in Chronic Disease
Accelerated tryptophan degradation in patients with chronic diseases might in fact be involved in the pathogenesis of several severe “side effects” of chronic disease like the development of cachexia (Schrocksnadel et al., 2006). More than 50% of all advanced-stage cancer patients experience a wasting syndrome called cachexia as a consequence of metabolic changes leading to loss of adipose tissue, skeletal muscle mass and anemia (Bruera, 1997).
7.1
General Mechanisms of Tumor Cachexia
Cancer-associated malnutrition can be due to local effects of the tumor, host response to the tumor or anti-cancer therapies. Reduced food intake, alterations in nutrient metabolism and resting energy expenditure may also favor the development of cachexia. Several agents produced by the tumor directly, or systemically in response to the tumor, such as pro-inflammatory cytokines and hormones, have been implicated in the pathogenesis of malnutrition and cachexia. These systemic effects are supposed to be a type of paraneoplastic syndrome, which is mediated by circulatory tumor-produced catabolic factors acting either alone or in concert with certain cytokines such as IFN-g (see Review by Brandacher et al., 1997). However, also other immune-mediated pathways like increased production of the proinflammatory cytokine TNF-a, which was first named “cachectin” before its molecular
Quality of Life and Tryptophan Degradation
119
identification, are certainly involved in the pathogenesis of cachexia and muscle loss in chronic disease. By influencing the brain’s serotonergic system or by mimicking leptin, a key target for neuropeptidergic effector molecules regulating food intake and energy expenditure via the sympathetic nervous system, proinflammatory cytokines might regulate the homeostatic loop of body weight regulation in patients.
7.2
Tryptophan and Weight Loss
As tryptophan is essential for many cellular functions, including protein biosynthesis and cell proliferation, an intracellular tryptophan deficiency may alter these cellular functions substantially. In patients with hematologic neoplasias, lower tryptophan and higher neopterin concentrations were associated with low serum albumin concentrations and weight loss (Denz et al., 1993; Iwagaki et al., 1997). Lower albumin concentrations were observed in patients with weight loss, indicating enhanced protein-turnover or decreased protein synthesis. In colorectal carcinoma patients substantial serum tryptophan reduction was observed in patients with liver metastasis and a history of more than 1 kg weight loss over the previous month, however in the other patients no relationship between weight loss and tryptophan degradation was seen (Huang et al., 2002). Also studies performed in HIV-infected patients point to a role of IFN-g and IFNmediated pathways in the pathogenesis of weight loss: A recent study performed in 152 HIV-infected patients showed a significant correlation between weight and plasma tryptophan concentrations (see > Figure 119-4). Also earlier data confirm that immune activation might be responsible for the development of cachexia: Neopterin levels were not only associated with weight loss (Zangerle et al., 1993), they were also shown to be predictive for the development of cachexia. Although unfortunately clear data regarding the role of IDO-mediated tryptophan degradation in the development of cachexia are missing so far, several other studies propose that a disturbed tryptophan metabolism in HIV-infected patients might contribute importantly to weight loss and disturbed energy and protein balance: Lots of energy generated by tryptophan oxidation are wasted in HIV-infected patients, and even during the asymptomatic phase patients show an increased basal metabolic rate (see Review by Murray, 2007). Enhanced IDO-activity and consecutively, enforced protein degradation, reflecting the combat between virus and host immune response were supposed very early to play a key role (Brown et al., 1991). Also the finding that children infected with HIV presenting with lower tryptophan concentrations failed to thrive, points towards the same direction (Brown et al., 1996). Immune-mediated tryptophan depletion probably slows down protein biosynthesis, while in parallel enforcing accelerated breakdown of muscle proteins.
8
Fatigue and Anemia of Chronic Disease and Enhanced Tryptophan Catabolism
Tryptophan degradation may also be involved in the development of fatigue and anemia of chronic disease. Weight loss and anemia have earlier been regarded as main causes for cancerrelated fatigue. However, both are only one among other factor contributing to the complex fatigue syndrome encountered in cancer patients (Holzner et al., 2002). Recent studies have shown that the etiology of cancer-related fatigue is multifactorial and fatigue is strongly related with physical performance and psychological distress.
2037
2038
119
Quality of Life and Tryptophan Degradation
. Figure 119-4 Relationship between tryptophan and weight and anemia, respectively. Correlations between plasma tryptophan concentrations on the one hand and weight (a) as well as hemoglobin concentrations (b) of HIV-infected patients on the other hand. Dark grey triangles represent HIV-infected patients under antiretroviral treatment (ART), while light grey squares represent patients without treatment
Fatigue can be defined as the subjective sensation of having reduced energy, loss of strength or becoming easily tired. The vast majority of tumor patients complains about fatigue and decreased physical efficiency, depending on the tumor type and progression. Special questionnaires like the FACT-F and the FACIT-AN have been developed to estimate the extent of fatigue (see also other chapters of this book).
8.1
Fatigue and Tryptophan Metabolism
Recent data of our group show that fatigue is significantly associated with disturbed tryptophan metabolism in patients with bronchus carcinoma (Schroecksnadel K, Pircher M, Fiegl M, Weiss G, Denz H, Fuchs D, Pteridines 2008, > Figure 119-5), furthermore also inflammation/immune activation markers C-reactive protein (CRP) and neopterin are associated with fatigue. Both the subjective assessment of patients regarding their extent of fatigue (on a scale from 1 to 5) and also the FACT-fatigue, and FACT-anemia-scores, respectively, were correlated with increased tryptophan degradation, patients with higher tryptophan concentrations experienced less fatigue. Similarly, also another study performed in patients with different kinds of cancer could demonstrate that the subjective fatigue assessment of patients as well as their subjective quality of life assessment was related with tryptophan degradation and tumor stage (Schroecksnadel et al., 2007). Patients with progressive tumor disease had the worst fatigue feeling, patients who were still alive after 3 months experienced less fatigue than patients who died within that time (Schroecksnadel et al., 2007). In the same study also decreased hemoglobin concentrations were associated with fatigue, but not with patients’ quality of life. Still, anemia has been established to be a significant
Quality of Life and Tryptophan Degradation
119
. Figure 119-5 Tryptophan degradation and quality of life. Decreasing plasma tryptophan concentrations coincide with decreasing quality of life scores (a,c; self-assessment score of patients scoring their quality of life on a scale from 1 to 5: 1 = very good and 5 = very bad quality of life) and increasing fatigue scores (b,d; score 1–5: 1 = no fatigue at all, 5 = very severe fatigue) in a heterogenous population of tumor patients (a,b) as well as in patients with bronchus carcinoma (c, d)
predictor of patients’ quality of life, especially in cancer patients, but also in patients suffering from chronic infection like HIV-infection.
8.2
Anemia of Chronic Disease and Cellular Immune Activation
On-going inflammatory and immune activation cascades are well established to trigger the development of anemia of chronic disease (Weiss and Goodnough, 2005). In accordance with
2039
2040
119
Quality of Life and Tryptophan Degradation
this finding, hemoglobin concentrations were demonstrated to be correlated inversely with IFN-g and neopterin concentrations in patients with HIV-infection and hematologic disease, respectively (Denz et al., 1993; Fuchs et al., 1991) Also in vitro studies indicate that pro-inflammatory cytokines IFN-g and TNF-a are very potent inhibitors of erythropoiesis, both cytokines suppressed the growth and differentiation of erythroid progenitor cells strongly (Fuchs et al., 1991). Furthermore pro-inflammatory cytokines also greatly disturb iron metabolism by shifting iron from the circulation to the reticuloendothelial system, i.e. in monocytes and macrophages. Therefore iron is less available for erythropoiesis, leading to the development of anemia.
8.3
Anemia of Chronic Disease and Tryptophan Degradation
However, not only withdrawal of iron might play a role, also tryptophan depletion by enhanced tryptophan catabolism might be involved: Recently an association between decreasing tryptophan concentrations and dropping hemoglobin levels was observed in patients with anemia of inflammation (Weiss et al., 2004). And also recent data by our group (Schroecksnadel et al., 2008) show a relationship between hemoglobin concentrations and tryptophan metabolism in HIV-infected patients and in cancer patients: Higher tryptophan plasma concentrations coincide with higher hemoglobin levels (> Figure 119-4), while higher IDOactivation (as reflected by increased kyn/trp) is observed in patients with anemia. Therefore, deprivation of the essential amino acid tryptophan might represent a limiting factor for erythropoiesis as well, cytokine-mediated inhibition of erythroid progenitor cells might in fact be mediated by tryptophan starvation of progenitor cells, which are arrested in their cell cycle and cannot proliferate any more. Recently IDO-activation in bone marrow stromal cells was shown to be able to inhibit T-cell proliferation (Meisel et al., 2004) suggesting that also the proliferation of hematopoietic progenitor cells might be suppressed by stromal cells. Also leukopoiesis might be affected, in HIV-infected patients significant inverse associations were demonstrated between urine and serum neopterin concentrations on the one hand and between CD4+ cell counts on the other hand (Fuchs et al., 1991). Urinary neopterin and serum soluble 75 kD TNF-receptor were furthermore shown to be jointly predictive for the percentage of CD4+ T-cell count loss. Conclusively, decreased bone marrow function certainly represents a major problem in patients with tumor disease as well as autoimmune disease, because it also impairs their quality of life significantly. Significant associations between hemoglobin concentrations and patients’ quality of life earlier resulted in the suggestion to improve patients’ quality of life by “correcting” low hemoglobin concentrations by blood transfusions or erythropoietin supplementation. Like tryptophan deprivation, also iron withdrawal is an important aspect among the antitumoral strategies of the immune system. Thus, iron overload of patients by erythrocyte transfusions may support tumor growth and could result in further immuno-suppression. Also tumor-promoting effects of erythropoietin have been suspected. Thus, “correction of anemia” should be performed not only with regard to the patient’s acute situation, but thinking also about possible deteriorating long-term effects.
9
“Neuropsychiatric” Side Effects of Enhanced Tryptophan Catabolism
Profound fatigue might coincide and also be interrelated with other distressing symptoms affecting patients with malignant disease, autoimmune disease or HIV-infection, namely
Quality of Life and Tryptophan Degradation
119
severe mood changes, subtle cognitive changes and depression. When an accelerated tryptophan catabolism was recognized in patients with HIV infection, its potential relevance for the precipitation of neuropsychiatric symptoms in the patients became immediately obvious (Fuchs et al., 1988), as a role of immune activation cascades in the pathogenesis of mood disturbances and depression had been suggested already earlier. A recent review by Dantzer and coworkers gives a very good and comprehensive overview of mechanisms that might be responsible for inflammation-mediated sickness and depression (Dantzer et al., 2008). Treatment of patients suffering from hepatitis C or chemotherapy-resistant tumors with interferon was shown to lead to major neuropsychiatric complications, implicating that interferons are central mediators of major depressive disorders. In line with that hypothesis, mice infected with adenovirus carrying the cDNA for murine IFN-g show impaired ingestive and motivated behavior, and animal models studying depression-like behavior induced by pro-inflammatory cytokines have been established. Interestingly, treatment of mice with the IDO-inhibitor 1-MT is able to prevent lipopolysaccharide (LPS)-induced depression-like behavior, but not sickness behavior (Dantzer et al., 2008). However, despite a dramatic decrease of tryptophan concentrations during LPS-treatment, serotonin turn-over was not significantly affected, probably indicating that also mechanisms other than decreased serotonin synthesis might play a role (Dantzer et al., 2008), e.g. alteration of glutamergic neurotransmission by kynurenine derivatives. In fact, depression and depression like behavior are probably triggered also by many other factors, like the personality of patients, their social environment and also the acute situation.
9.1
Depression and Tryptophan Metabolism
In fact, decreased tryptophan availability might not lead to depressive symptoms immediately, rather tryptophan depletion might render patients more susceptible to depression. In line with this thesis, acute tryptophan depletion was demonstrated to decrease mood in vulnerable patients who have a family history of depression (Ruhe et al., 2007). Further studies in animals and also human beings may provide promising new data how tryptophan depletion/kynurenine metabolites are involved in the development of depression. Anyway, depression is a common complication in patients with chronic diseases, like cancer patients, patients suffering from autoimmune disease or chronic infection. A recent study performed in HIV-infected patients showed that more than 50% of patients suffered from depression, however, only a small percentage of them were also taking antidepressant medication (Schroecksnadel et al., 2008). Significant associations were observed between depression scores on the one hand, and neopterin concentrations and the kynurenine/tryptophan ratio only in patients without antidepressant medication on the other (see also > Figure 119-6). Decreased tryptophan and serotonin concentrations found in the blood of HIV-infected patients also suppose that IDO-activation might enforce the development of depression (Launay et al., 1988). Also in other diseases a relationship between major depression and disturbed tryptophan metabolism has been described (Kohl and Sperner-Unterweger, 2007; Maes et al., 1994). Lowered serum tryptophan levels were observed in patients with major depression in parallel with elevated IFN-g and neopterin concentrations (Maes et al., 1994). Similarly, a relationship between lower tryptophan concentrations and an increased susceptibility to depression was reported in malignant melanoma patients during treatment with interferon-a (Capuron et al., 2003).
2041
2042
119
Quality of Life and Tryptophan Degradation
. Figure 119-6 Immune activation and depression. Depressive HIV-infected patients without antidepressant treatment present with enhanced tryptophan degradation (as reflected by increased Kyn/trp-p < 0.05, a) and increased neopterin concentrations (p < 0.01, b) compared to patients without depression. Results of 152 patients are depicted as boxplots: medians (black line), 25th and 75th percentile (box) and range (whiskers) of patients with (dark grey boxplots) and without antidepressant medication (light grey boxplots)
9.2
Cognitive Impairment and Tryptophan Degradation
Also cognitive function might be impaired by decreased tryptophan availability. Immunemediated tryptophan degradation by IDO may thus initiate the development of several neuropsychiatric symptoms (Wirleitner et al., 2003). In HIV-infected patients decreased tryptophan concentrations were related with progressive cognitive inability (Fuchs et al., 1990), antiretroviral treatment was demonstrated to improve cognitive impairment (Suarez et al., 2001) and also reduce IDO-activation and liquor quinolinic acid formation (Look et al., 2000). Also in patients with late phase Alzheimer’s and Huntington’s disease an association between tryptophan degradation and cognitive ability was observed (Kohl and SpernerUnterweger, 2007; Wirleitner et al., 2003). Several studies in HIV infected patients have suggested that enhanced tryptophan catabolism may be involved in the development of AIDS-related dementia and peripheral neuropathy (Fuchs et al., 1990; Wirleitner et al., 2003). Not only tryptophan deprivation, but also the accumulation of neurotoxic kynurenine metabolites might contribute to neuropsychiatric changes: Concentrations of kynurenine and the potent neurotoxin quinolinic acid raise in the cerebrospinal fluid of patients with HIV infection and SLE (further literature – see Wirleitner et al., 2003). Similarly, elevated quinolinic acid concentrations were shown in models of inflammatory neurological disorders such as experimental allergic encephalitis, bacterial and viral infections (see also Review by Wirleitner et al., 2003). Quinolinic acid may not only influence the neuroendocrine system by interference with the NMDA receptor, also NMDA-mediated induction of apoptosis may represent a crucial mechanism.
Quality of Life and Tryptophan Degradation
10
119
Conclusion
Data of several studies propose a role of enhanced tryptophan degradation in the development of neuropsychiatric symptoms, weight loss and anemia of chronic disease. Still, it has to be outlined that a causal role of IDO activation and/or kynurenine metabolites in the development of above-mentioned symptoms in human beings has not been proven so far, even if clinical data obtained so far fit very well with our hypothesis. The coincidence of enhanced tryptophan catabolism with above described symptoms in fact strongly indicates an important role of IDO activation and tryptophan depletion, but it has to be kept in mind that most studies cited here are mainly based on serum/plasma measurements and data of tryptophan/ serotonin metabolism in the brain are scarce. Thus, further investigations involving the brain and cerebrospinal fluid, respectively, might enable a deeper insight into the relationship between tryptophan metabolism and neuropsychiatric symptoms. Similarly, also further studies are needed to clarify whether decreased tryptophan availability really affects haematopoiesis, and how IDO might modulate muscle and skeletal metabolism (and thus lead to cachexia).
Summary Points Patients suffering from chronic diseases like cancer, autoimmune disease or chronic infection often experience a deteriorated quality of life.
Independently of their underlying disease, patients suffering from chronic disease present
with similar symptoms, namely increased fatigue feeling, decreased physical efficiency, weight loss, anemia, mood disorders or even depression. Within Th1-type immune response the proinflammatory cytokine interferon-g induces several biochemical pathways, among them activation of the enzyme indoleamine-2,3dioxygenase (IDO). IDO degrades the essential amino acid tryptophan to kynurenine leading to decreased tryptophan availability. In patients suffering from chronic disease IDO-mediated tryptophan degradation is strongly enhanced, patients present with lowered tryptophan and increased kynurenine concentrations in parallel with increased inflammation/immune activation marker concentrations like neopterin. Tryptophan deprivation represents a very important mechanism to inhibit/slow down the proliferation of pathogens as well as T-lymphocytes. On-going IDO activation within chronic diseases appears to contribute importantly to the development of immunodeficiency in patients. Immune-mediated IDO-activation might play an important role in the development of several symptoms known to deteriorate patients’ quality of life: neuropsychiatric symptoms, anemia of chronic disease and weight loss are known to coincide with enhanced IDO-activation. Trytophan is the precursor of the important neurotransmitters serotonin (5-hydroxytryptamin) and NAD. Enhanced tryptophan degradation might trigger the development of neuropsychiatric symptoms by influencing neurotransmission.
2043
2044
119
Quality of Life and Tryptophan Degradation
Increased tryptophan degradation has been associated with fatigue, sleeping disorders, cognitive impairment/dementia as well as depressive disorders.
Cytokine cascades within chronic disease impair haematopoiesis resulting in anemia of chronic disease. Increased IDO-activation was demonstrated in patients with anemia of chronic disease suggesting that additionally to restricted iron supply also limited tryptophan availability might contribute to decreasing bone marrow function.
Appendix Key features of tryptophan:
Tryptophan is an essential amino acid. Tryptophan is degraded by three enzymes: tryptophan dioxygenase (TDO), indoleamine2,3-dioxygenase (IDO) and indoleamine-pyrrole 2,3-dioxygenase-like 1 (INDOL 1).
Tryptophan is the precursor of serotonin and nicotinamide dinucleotide. Tryptophan deprivation represents an important mechanism to inhibit the growth of various pathogens and cells.
Enhanced tryptophan degradation takes place during cellular immune response. Immune-mediated tryptophan degradation might play an important role in the development of several symptoms frequently encountered in patients with chronic disease: fatigue, weight loss, anemia, neuropsychiatric complications.
References Brandacher G, Margreiter R, Fuchs D. (2007). Curr Drug Metab. 8: 273–282. Brandacher G, Amberger A, Schroecksnadel K, Margreiter R, Fuchs D. (2007). Predictive value of IFN-g induced Indoleamine 2,3-dioxygenase (IDO) expression in cancer patients. Cancer Management in Man. Brown BR, Baranowski MD, Kulig JW, Stephenson JN, Perry B. (1996). Adolescence. 31: 253–264. Brown RR, Ozaki Y, Datta SP, Borden EC, Sondel PM, Malone DG. (1991). Adv Exp Med Biol. 294: 425–435. Bruera E. (1997). BMJ. 315: 1219–1222. Capuron L, Neurauter G, Musselman DL, Lawson DH, Nemeroff CB, Fuchs D, Miller AH. (2003). Biol Psychiatry. 54: 906–914. Dantzer R, O’Connor JC, Freund GG, Johnson RW, Kelley KW. (2008). Nat Rev Neurosci. 9: 46–56. Denz H, Orth B, Weiss G, Herrmann R, Huber P, Wachter H, Fuchs D. (1993). Clin Investig. 71: 37–41.
Frumento G, Rotondo R, Tonetti M, Damonte G, Benatti U, Ferrara GB. (2002). J Exp Med. 196: 459–468. Fuchs D, Moller AA, Reibnegger G, Stockle E, Werner ER, Wachter H. (1990). J Acquir Immune Defic Syndr. 3: 873–876. Fuchs D, Shearer GM, Boswell RN, Lucey DR, Clerici M, Reibnegger G, Werner ER, Zajac RA, Wachter H. (1991). AIDS. 5: 209–212. Fuchs D, Werner ER, Dierich MP, Wachter H. (1988). Ann Neurol. 24: 289. Grohmann U, Fallarino F, Bianchi R, Vacca C, Orabona C, Belladonna ML, Fioretti MC, Puccetti P. (2003). Adv Exp Med Biol. 527: 47–54. Holzner B, Kemmler G, Greil R, Kopp M, Zeimet A, Raderer M, Hejna M, Zochbauer S, Krajnik G, Huber H, Fleischhacker WW, Sperner-Unterweger B. (2002). Ann Oncol. 13: 965–973. Huang A, Fuchs D, Widner B, Glover C, Henderson DC, len-Mersh TG. (2002). Br J Cancer. 86: 1691–1696.
Quality of Life and Tryptophan Degradation Iwagaki H, Hizuta A, Uomoto M, Takeuchi Y, Saito S, Tanaka N. (1997). Acta Med Okayama. 51: 233–236. Kohl C, Sperner-Unterweger B. (2007). Curr Drug Metab. 8: 283–287. Launay JM, Copel L, Callebert J, Corvaia N, Lepage E, Bricaire F, Saal F, Peries J. (1988). J Acquir Immune Defic Syndr. 1: 324–325. Look MP, Altfeld M, Kreuzer KA, Riezler R, Stabler SP, Allen RH, Sauerbruch T, Rockstroh JK. (2000). AIDS Res Hum Retroviruses. 16: 1215–1221. MacKenzie CR, Heseler K, Muller A, Daubener W. (2007). Curr Drug Metab. 8: 237–244. Maes M, Scharpe S, Meltzer HY, Okayli G, Bosmans E, D’Hondt P, Vanden Bossche BV, Cosyns P. (1994). Psychiatry Res. 54: 143–160. Meisel R, Zibert A, Laryea M, Gobel U, Daubener W, Dilloo D. (2004). Blood. 103: 4619–4621. Mellor AL, Munn DH. (2004). Nat Rev Immunol. 4: 762–774. Moffett JR, Namboodiri MA. (2003). Immunol Cell Biol 81: 247–265. Moroni F. (1999). Eur J Pharmacol. 375: 87–100. Munn DH, Zhou M, Attwood JT, Bondarev I, Conway SJ, Marshall B, Brown C, Mellor AL. (1998). Science. 281: 1191–1193. Munn DH, Shafizadeh E, Attwood JT, Bondarev I, Pashine A, Mellor AL. (1999). J Exp Med. 189: 1363–1372. Murray MF. (2007). Curr Drug Metab. 8: 197–200. Neumeister A. (2003). Psychopharmacol Bull. 37: 99–115. Peters JC. (1991). Adv Exp Med Biol. 294: 345–358. Ruhe HG, Mason NS, Schene AH. (2007). Mol Psychiatry. 12: 331–359. Sakurai K, Zou JP, Tschetter JR, Ward JM, Shearer GM. (2002). J Neuroimmunol. 129: 186–196. Schro¨cksnadel H, Baier-Bitterlich G, Dapunt O, Wachter H, Fuchs D. (1996). Obstet Gynecol. 88: 47–50. Schroecksnadel K, Fiegl M, Prassl K, Winkler C, Denz HA, Fuchs D. (2007). J Cancer Res Clin Oncol. 133: 477–485. Kurz-Schroecksnadel K, Pircher M, Fiegl M, Weiss G, Denz H, Fuchs D. (2008). Pteridines. 19: 57–58.
119
Schroecksnadel K, Sarcletti M, Winkler C, Mumelter B, Weiss G, Fuchs D, Kemmler G, Zangerle R. (2008). Brain Behav Immun Epub ahead of print. Schrocksnadel K, Wirleitner B, Winkler C, Fuchs D. (2006). Clin Chim Acta. 364: 82–90. Sharp K, Brindle PM, Brown MW, Turner GM. (1993). Br J Obstet Gynaecol. 100: 209–215. Suarez S, Baril L, Stankoff B, Khellaf M, Dubois B, Lubetzki C, Bricaire F, Hauw JJ. (2001). AIDS. 15: 195–200. Takikawa O, Yoshida R, Kido R, Hayaishi O. (1986). J Biol Chem. 261: 3648–3653. Taylor MW, Feng GS. (1991). FASEB J. 5: 2516–2522. Terness P, Bauer TM, Rose L, Dufter C, Watzlik A, Simon H, Opelz G. (2002). J Exp Med. 196: 447–457. Uyttenhove C, Pilotte L, Theate I, Stroobant V, Colau D, Parmentier N, Boon T, Van den Eynde BJ. (2003). Nat Med. 9: 1269–1274. von Bubnoff BD, Matz H, Frahnert C, Rao ML, Hanau D, de la SH, Bieber T. (2002). J Immunol. 169: 1810–1816. Weiss G, Diez-Ruiz A, Murr C, Theurl I, Fuchs D. (2002). Pteridines. 13: 139–143. Weiss G, Goodnough LT. (2005). N Engl J Med. 352: 1011–1023. Weiss G, Schroecksnadel K, Mattle V, Winkler C, Konwalinka G, Fuchs D. (2004). Eur. J. Haematol. 72: 130–134. Werner ER, Hirsch-Kauffmann M, Fuchs D, Hausen A, Reibnegger G, Schweiger M, Wachter H. (1987). Biol Chem. 368: 1407–1412. Werner ER, Werner-Felmayer G, Fuchs D, Hausen A, Reibnegger G, Wachter H. (1989). Biochem J. 262: 861–866. Widner B, Werner ER, Schennach H, Wachter H, Fuchs D. (1997). Clin Chem. 43: 2424–2426. Wirleitner B, Neurauter G, Schrocksnadel K, Frick B, Fuchs D. (2003). Curr Med Chem. 10: 1581–1591. Zangerle R, Reibnegger G, Wachter H, Fuchs D. (1993). AIDS. 7: 175–181.
2045
120 L-Carnitine Supplementation on Quality of Life and Other Health Measures G. Mantovani . A. Maccio` . C. Madeddu . G. Gramignano 1 1.1 1.2 1.3 1.4 1.4.1 1.4.2 1.5
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2048 Significance of Quality of Life as a Main End-Point in Cancer . . . . . . . . . . . . . . . . . . 2048 L-Carnitine Physiological Role in Health and Diseases . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2049 L-Carnitine Supplementation in Several Clinical Settings . . . . . . . . . . . . . . . . . . . . . . . . . 2051 L-Carnitine Supplementation in Cancer Patients: Effects on QL . . . . . . . . . . . . . . . . . . 2052 Cancer-Related Anorexia/Cachexia Syndrome (CACS) . . . . . . . . . . . . . . . . . . . . . . . . . . . 2052 Cancer-Related Fatigue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2055 L-Carnitine Supplementation on QL in Other Clinical Settings . . . . . . . . . . . . . . . . . . . 2057
2
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2058 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2059 Appendix 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2060 Most commonly used QL questionnaires and related scoring system . . . . . . . . . . . 2060 Multidimensional Fatigue Symptom Inventory-Short Form (MFSI-SF) . . . . . . . . . 2063
#
Springer Science+Business Media LLC 2010 (USA)
2048
120
L-Carnitine
Supplementation on Quality of Life and Other Health Measures
Abstract: Chronic diseases often have a relapsing and remitting course with substantial impact on function and > quality of life (QL). For chronic illnesses where there is no cure, it is important to recognize that therapy really makes people feel better. Thus, survival per se is no longer perceived to be the only end point; the goal is to improve, restore, or preserve QL. Quality of life and > supportive care are complementary concepts in the care of cancer patients. Nutritional support is a critical component of supportive care. However, both palliative and curative treatment of cancer should be accompanied by specific and patienttailored nutritional intervention aiming at improving performance status and QL. Among the oral > nutritional supplements, L-carnitine with its nutritional, antioxidant, antimyopathic effects is one of the most promising. Several studies, carried out in cancer patients as well as in other chronic illnesses, demonstrated the ability of L-carnitine supplementation to in ameliorating QL, particularly > fatigue, and other symptoms, such as weight loss, pain, neuropathy and chemotherapy side effects, which in turn influence patient’s QL. In detail a recently published study of our group demonstrated the ability of L-carnitine oral supplementation (6g/day for 4 weeks) to induce increase of LBM, amelioration of fatigue and appetite. Several authors, in accordance to our results, showed that L-carnitine supplemented to cancer patients with fatigue and carnitine deficiency was able to ameliorate fatigue, sleep disorders and mood. Beside cancer patients, L-carnitine supplementation demonstrated to be able to improve quality of life in hemodialized patients and to ameliorate fatigue in celiac disease and multiple sclerosis patients as well as in elderly subjects. More clinical trials, however, are needed to identify those patients and diseases that may derive the greatest benefit from the supplementation with this agent. Indeed, while several studies have examined the role of L-carnitine supplementation in different comorbid conditions, very few have assessed its effect on QL. List of Abbreviations: AIDS, acquired immune deficiency syndrome; BDI, beck depression inventory; CACS, > cancer-related anorexia/cachexia syndrome; CoA, coenzyme A; COPD, chronic obstructive pulmonary disease; CPT, carnitine palmitoyltransferase; CRF, chronic renal failure; CRP, C-reactive protein; EORTC-QLQ-C30, European Organization for Research and Treatment of Cancer, Quality of Life Questionnaire; ESRD, end-stage renal disease; IL, interleukin; LBM, lean body mass; MFSI-SF, multidimensional fatigue symptom inventoryshort form; MMSE, mini-mental state examination; MPA, medroxyprogesterone acetate; OS, > oxidative stress; QL, quality of life; ROS, reactive oxygen species; SF-36, medical outcome study 36-item short form; TNF, tumor necrosis factor; VAS, visual analogue scales; WHO, World Health Organization
1
Introduction
1.1
Significance of Quality of Life as a Main End-Point in Cancer
Chronic diseases often have a relapsing and remitting course with substantial impact on function and quality of life (QL). For chronic illnesses where there is no cure, it is important to recognize that therapy really makes people feel better. Thus, survival per se is no longer perceived to be the only end point; the goal is to improve, restore, or preserve QL. The World Health Organization (WHO) defines QL as ‘‘an individual perception of their position in life in the context of the culture and value systems in which they live and in relation
L-Carnitine
Supplementation on Quality of Life and Other Health Measures
120
to their goals, expectations, standards and concerns’’ (The World Health Organization, 1995). The concept of QL is generally described as multidimensional comprising individuals’ perceived physical, psychosocial and emotional functioning. QL encompasses the concept of health-related quality of life and other domains such as environment, family and work. Health related quality of life is the extent to which one’s usual or expected physical, emotional, and social well-being are affected by medical conditions or their treatment. In particular, when considering the concept of QL in cancer patients different interrelated components have been defined (Caplan, 1987). The first is the physical aspect which includes physical symptoms, response to treatment and treatment toxicities, body image and mobility. The second includes psychological and social functioning, interpersonal relationships, happiness, spirituality and financial issues. The third aspect refers to the individual perception of QL and it is influenced by a person’s culture, philosophy and politic context. Despite the difficulties of actually quantifying patient perceptions of QL, the number of instruments available to measure QL psychometrically has rapidly increased in the last few years. Assessments can now be made in a variety of distinctive ways using both specific and generic measures. Generic scales include items that cover all major aspects of a person’s health and are applicable whatever the person’s condition. Specific scales include only those items likely to be affected by the disease concerned or its treatment, and have been developed for particular disease categories, principally cancer (ranging from cancer in general to specific treatments/phases). Most QL instruments have been designed for self-administration and are relatively short. There are no gold-standard questionnaires, and the choice is based on psychometric properties, research objectives and study design (Boini et al., 2004; Carr et al., 1996; Fitzpatrick et al., 1992; Fletcher et al., 1992; Ware et al., 1981). Among the generic instruments the most commonly used are the WHO Quality of life instrument (WHOQOL100) (The World Health Organization, 1998) and the Medical Outcome Study 36-Item Short Form (SF-36) (Ware and Sherbourne, 1992). To evaluate QL in disease, i.e., cancer, several questionnaires and visual analogue scales (VASs) are commonly used: some addressing specific symptoms, such as the Beck Depression Inventory (BDI) (Beck et al., 1996; Katz et al., 2004) for depression or the Multidimensional Fatigue Symptom Inventory-Short Form (MFSI-SF) for fatigue, some specifically developed for the disease, such as the European Organization for Research and Treatment of Cancer, Quality of Life Questionnaire (EORTC-QLQ-C30) (See > Appendix 1).
1.2
L-Carnitine
L-carnitine
Physiological Role in Health and Diseases
is a natural nutrient and essential for the > beta-oxidation of fatty acids in mitochondria to generate ATP. L-Carnitine is a low-molecular weight, trimethylamine compound with vitamin-like properties, which is synthesized in the liver, kidney, and brain via the conversion of two essential amino acids, lysine and methionine (Calabrese et al., 2006). The endogenous synthesis of L-carnitine is catalyzed by the concerted action of five different enzymes. This process requires the two essential amino acids lysine and methionine, iron (Fe2+), vitamin C, vitamin B6 and niacin in the form of nicotinamide adenine dinucleotide. Human requirements for carnitine are usually met with a combination of diet and endogenous biosynthesis. The primary dietary sources of carnitine are red meat, poultry,
2049
2050
120
L-Carnitine
Supplementation on Quality of Life and Other Health Measures
fish and dairy products. In comparison, vegetable products provide fairly small amounts. Dietary carnitine appears to be rapidly absorbed from the intestinal lumen across the mucosal membrane by both passive and active transport mechanisms. Carnitine is then taken up from the portal circulation by the liver and subsequently released into the systemic circulation (Hoang et al., 2007). The carnitine system, which includes carnitine, carnitine esters (acetylcarnitine, propionylcarnitine), several specific intracellular enzymes and membrane transporters, plays a critical role in cellular homeostasis (Arrigoni-Martelli and Caso, 2001; Rebouche, 1999). Indeed, b-oxidation, the major process by which long-chain fatty acids are oxidized in mitochondria, is ubiquitously dependent on this system (> Table 120-1). Carnitine is a cofactor required for transformation of the free long-chain fatty acids into acyl-carnitines and for their subsequent transportation into the mitochondrial matrix, where they undergo b-oxidation for cell energy production (> Figure 120-1). Fatty acids in cytoplasm are transformed to long-chain acyl-coenzyme A (CoA) and transferred into the mitochondrial matrix by the action of three carnitine-dependent enzymes to produce acetylCoA through the b-oxidation pathway. The relation between CoA and carnitine is pivotal for energy metabolism. CoA is required for beta-oxidation, for the metabolism of several amino acids, for pyruvate dehydrogenase, and thus for tricarboxylic acid cycle. The pyruvate dehydrogenase complex, the key irreversible rate-limiting step in carbohydrate oxidation, is modulated by the intramitochondrial ratio of acetyl-CoA to CoA. An increased ratio results in inhibition of pyruvate dehydrogenase activity: carnitine, by converting acetyl-CoA into acetylcarnitine CoA, thus removes a powerful inhibitor of pyruvate dehydrogenase. Thus, the activity of L-carnitine in the modulation of the intramitochondrial ratio of acetyl CoA to CoA affects glucose oxidation. The resulting increased access to the tricarboxylic acid cycle leads to an increased availability of ATP and NADH/FADH2 production (Gramignano et al., 2006).
. Table 120-1 Key facts of carnitine L-carnitine
is a trimethylated amino acid, roughly similar in structure to choline
Carnitine’s primary mechanism of action is apparently attributable to its role as a cofactor in the transformation of free long-chain fatty acids into acylcarnitines and for their subsequent transport into the mitochondrial matrix, where they undergo b-oxidation for cellular energy production Carnitine is involved in the metabolism of ketones for energy and the conversion of branched-chain amino acids – valine, leucine, and isoleucine – into energy L-carnitine
is supplied exogenously as a component of the diet and can also be synthesized endogenously Both primary and secondary carnitine deficiencies do occur Conditions that appear to benefit from exogenous supplementation of L-carnitine include anorexia, chronic fatigue, cardiovascular disease, diphtheria, hypoglycemia, male infertility, muscular myopathies, and Rett syndrome Preterm infants, dialysis patients, cancer patients and HIV-positive individuals seem to be prone to a deficiency of L-carnitine and benefit from supplementation
L-Carnitine
Supplementation on Quality of Life and Other Health Measures
120
. Figure 120-1 A schematic diagram of the metabolic pathways of carnitine. Carnitine’s role in long-chain fatty acid (acyl group) translocation into the mitochondrial matrix, for subsequent oxidation is highlighted in red, whereas the role of carnitine as a buffer of excess acetyl-CoA production is highlighted in blue. PDC, pyruvate dehydrogenase complex; TCA, tricarboxylic acid cycle; CAT, carnitine acetyltransferase; CACT, carnitine acylcarnitine translocase; CPT, carnitine palmitoyltransferase
1.3
L-Carnitine
Supplementation in Several Clinical Settings
Carnitine may be essential for several groups of people including both children and adults suffering from a variety of genetic, infectious and injury-related illnesses. Some childhood cardiomiopathies are due to carnitine metabolic errors or deficiencies. There is data that support the treatment of some myocardial dysfunctions with L-carnitine supplementation (Winter et al., 1995). Treatment of diseases, such as immunodeficiency virus type 1 infection/ acquired immune deficiency syndrome (AIDS), may elicit or cause carnitine deficiency problems (Mintz, 1995). Other possible clinical applications of L-carnitine are: cardiovascular diseases; Alzheimer’s disease and other neurodegenerative disorders; epilepsy; depression; physical and mental fatigue; pain and neuropathies; maculodegeneration; diabetes and diabetic neuropathy; chemotherapy-induced neuropathy; sexual dysfunction; emphysema and chronic obstructive pulmonary disease (COPD); kidney chronic failure (Hoang et al., 2007). While several studies have examined the role of L-carnitine supplementation in different comorbid conditions, very few have assessed its effect on QL. Complaints like weakness, poor exercise tolerance and fatigue, affect QL and L-carnitine may be beneficial in relieving these symptoms.
2051
2052
120 1.4
L-Carnitine
L-Carnitine
Supplementation on Quality of Life and Other Health Measures
Supplementation in Cancer Patients: Effects on QL
L-carnitine
has been shown by various studies to benefit cancer patients with cachexia, fatigue, pain and neuropathy and chemotherapy side effects. Some studies carried out on advanced cancer patients undergoing anticancer therapy have shown the efficacy of L-carnitine administration in increasing lean body mass, and improving fatigue and quality of life (Hoang et al., 2007).
1.4.1
Cancer-Related Anorexia/Cachexia Syndrome (CACS)
Cancer-related anorexia/cachexia syndrome (CACS) and oxidative stress (OS) play a key role in the progression and outcome of neoplastic disease. CACS is present in 80% of terminally ill cancer patients, in about 50% of patients who undergo antineoplastic therapies and in 2–3% of patients who receive adjuvant chemotherapy (Gramignano et al., 2006). CACS is a multifactorial syndrome characterized by tissue wasting, loss of body weight, particularly of lean body (muscle) mass and to a lesser extent adipose tissue, metabolic alterations, fatigue, reduced performance status, very often accompanied by anorexia leading to a reduced food intake (Brennan, 1977; Bruera, 1992; Heber et al., 1986; Nelson and Walsh, 1991). In addition to decreased food intake, important abnormalities in carbohydrate, protein and lipid as well as energy metabolism have been observed in patients with cancer which may account for CACS. Abnormalities in protein metabolism include: increased protein turnover, loss of skeletal muscle, and increased gluconeogenesis from amino acids and fatty acids. Loss of skeletal muscle proteins occurs through increased rate of skeletal muscle protein breakdown and decrease in the rate of skeletal muscle protein synthesis (Gramignano et al., 2006). In particular, glucose intake is severely compromised by the presence of symptoms such as nausea, vomiting and anorexia. The reduced glucose intake induces the activation of gluconeogenesis from lactate, muscle amino acids and free fatty acids, finally leading to depletion of fat and protein stores. The cycle converting lactate to pyruvate and glucose is named the Cory cycle. The Cory cycle activity is increased from 20% (value observed in healthy subjects) to 50% in cancer patients with CACS. The utilization of lactate and glycogenetic amino acids for the synthesis of glucose in the liver is a process associated with high energy consumption. Increased gluconeogenesis has been proposed as the main cause of increased energy expenditure of cancer patients. The increase of glucose turnover is strictly related to histotype, stage of disease and grade of cachexia. Several studies have analyzed the relationship between glucose metabolism and changes of body weight. Patients without weight loss have a normal Cory cycle activity, whilst those with progressive weight loss have an increased Cory cycle activity associated with an increased lactate production. However, the compensatory increased gluconeogenesis is associated with reduced synthesis of insulin and insulin resistance. In fact, the most important carbohydrate abnormalities observed in cachectic cancer patients are increased glucose synthesis, gluconeogenesis and Cory cycle activity, insulin resistance, and decreased glucose tolerance. This adaptation redirects glucose to the liver and other viscera and away from skeletal muscle because hepatic glucokinase is not affected by insulin, unlike hexokinase in myocytes and elsewhere. The energy needs of muscle are met by oxidation of non-essential amino acids, which contributes to negative nitrogen balance. In addition to having increased glucose production and glucose intolerance, cancer patients show a clear insulin resistance that involves adipose tissue, skeletal muscle and liver. The increased hepatic glucose production is partially the result of a lack of inhibition of gluconeogenesis by insulin
L-Carnitine
Supplementation on Quality of Life and Other Health Measures
120
due to a certain degree of liver insulin resistance. Similarly, glucose utilization by skeletal muscle is reduced both in experimental animals and cancer patients, this being the result of clear insulin resistance (Maccio` et al., 2006). CACS may results from circulating factors produced by the tumor or by the host immune system in response to the tumor, such as cytokines released by lymphocytes and/or monocytes or macrophages (Mantovani et al., 2004). Several proinflammatory cytokines, including Interleukin (IL)-6, IL-1 and Tumor Necrosis Factor (TNF)-a, are involved in the pathogenesis of cachexia associated with human cancer (Tisdale, 2005). An imbalance between oxidants and antioxidants in favor of the oxidants, can potentially lead to cell damage and is termed ‘‘oxidative stress.’’ These oxidants, also termed reactive oxygen species (ROS), are present as a normal product of aerobic metabolism but can be produced at elevated rates under pathophysiological conditions. They are superoxide radicals, hydroxyl radicals, hypochlorous acid, peroxyl radicals and singlet oxygen, and can cause cell damage by oxidizing nucleic acid, proteins and membrane lipids (Lundholm et al., 1994). In advanced cancer patients the high levels of ROS are caused by: (1) An increased production due to hypermetabolism, increased activation of the immune system and chronic inflammation with the associated release of proinflammatory cytokines, C-reactive protein (CRP) and fibrinogen; (2) An inadequate detoxification due to altered glucose metabolism in addition to symptoms such as anorexia/cachexia, nausea, and vomiting, that prevent a normal nutrition and thereby a normal supply of nutrients such as glucose, proteins and vitamins, leading to accumulation of ROS (Maccio` et al., 2006). ROS may act at different stages in the establishment and progression of cancer. It has recently been well documented that ROS are also involved in tissue wasting and CACS. Moreover, some cytokines involved in CACS, especially TNFa, are inducers of ROS, which seem to be part of the final common pathway through which cell damage takes place. Thus, OS, which is one of the causes of and in turn may be worsened by CACS, is promoted by an excess of ROS and some proinflammatory cytokines, such as IL-1, IL-6 and TNFa (Beck et al., 1990; Mantovani et al., 2001). Mitochondrial b-oxidation is a major source of reactive oxygen species, and it has been suggested that OS, due to reduced efficiency of cellular bioenergetic processes, contributes to the development of metabolic abnormalities (Petersen et al., 2003). The use of carnitine to optimize mitochondrial function, therefore, may result in reduced oxidative stress leading to relevant clinical outcomes, such as increased appetite, preserved lean body mass, and reduced morbidity. The rationale for the use of carnitine in wasting diseases is further supported by the fact that that skeletal muscle is the organ with the highest concentration of carnitine. Carnitine deficiency induces a decrease in muscle strength, hypotonia, lipid storage myopathy (predominantly in type I fiber), and type II fiber atrophy, including myolysis (Reda et al., 2003). The carnitine system is also important for the control and regulation of fuel partitioning within the skeletal muscle, which may contribute to muscle structure and function. Therefore, low serum levels of carnitine in advanced cancer patients, which are due to a decreased dietary intake as well as to an impaired endogenous synthesis, could be an important factor in the development of CACS (Gramignano et al., 2006). In accordance to this rationale, a study by Gramignano et al. (2006) recently demonstrated that advanced cancer patients supplemented with 6 g daily of oral L-carnitine for 4 weeks reported a significant increase of lean body mass (LBM) following carnitine supplementation (> Figure 120-2). LBM may be considered the most important nutritional/functional symptom of CACS. In CACS decreased LBM may contribute to impair the performance status, the physical and social functioning, the patient perceived self-image, and, therefore, QL.
2053
2054
120
L-Carnitine
Supplementation on Quality of Life and Other Health Measures
. Figure 120-2 Changes of lean body mass (LBM) during L-carnitine supplementation in cancer patients. Histograms represent the mean LBM values (kg). The difference between LBM at baseline (t0), after 2 (t1) and 4 (t2) weeks of treatment was statistically significant for both t1 and t2 versus t0 (p = 0.001 and p < 0.05, respectively) (Gramignano G et al., 2006)
Moreover, in the above cited study, L-carnitine supplementation was also found to improve appetite (> Figure 120-3) (Gramignano et al., 2006). The orexigenic effects of carnitine are of particular interest, as they may help us to better understand the pathogenesis of cancerassociated anorexia, whose etiology is complex and involves many different factors, including lipid metabolism. Experimental animal evidence (He et al., 2006) suggests that under physiological conditions in the hypothalamic area which regulates food intake, increased malonyl-CoA concentrations inhibit carnitine palmitoyltransferase (CPT)-dependent fatty acid oxidation and reduce food intake. In human cancer it may be hypothesized that CPT activity is depressed and this may lead to increased malonyl-CoA concentrations further inhibiting b-oxidation and food intake. Restoration or enhancement via carnitine supplementation of fatty acid oxidation acting on the hypothalamic sensing areas, may thus result in improved appetite (Laviano et al., 2006). On the basis of these promising results, we have recently included L-carnitine supplementation in an ongoing randomized phase III study to assess the effectiveness and safety of an innovative treatment against CACS/OS. The study plans to enroll more than 300 cancer patients randomized to five different arms of treatment. Eligibility criteria are: hystologically confirmed tumors of any site; loss of body weight 5% in the last 3 months (clinical CACS) and/or with abnormal values of proinflammatory cytokines, ROS and antioxidant enzymes predictive of the onset of CACS; life expectancy >4 months. Patients may be treated with either antineoplastic therapy or supportive care. The study aims to test which is the most effective and safest treatment of cancer-related anorexia/cachexia syndrome and oxidative stress in improving identified primary endpoints: increase of lean body mass, decrease of resting energy expenditure, increase of total daily physical activity, decrease of interleukin-6 and tumor necrosis factor-alpha, and improvement of fatigue assessed by the
L-Carnitine
Supplementation on Quality of Life and Other Health Measures
120
. Figure 120-3 Changes of appetite during L-carnitine supplementation in cancer patients. Histograms represent the mean appetite VAS score. The appetite increased significantly after 2 weeks (t1) (6.16 + 2.4, p < 0.05) and 4 weeks (t2) (6.83 + 1.9, p = 0.001) of treatment with LC supplementation in comparison to baseline (t0) (Gramignano G et al., 2006)
MFSI-SF questionnaire. All patients receive as basic treatment polyphenols plus antioxidant agents alpha-lipoic acid, carbocysteine, and vitamins A, C, and E, all orally. Patients are then randomized to one of the following five arms: (1) medroxyprogesterone acetate/ megestrol acetate; (2) pharmacologic nutritional support containing eicosapentaenoic acid; (3) L-carnitine; (4) thalidomide; or (5) medroxyprogesterone acetate (MPA)/megestrol acetate plus pharmacologic nutritional support plus L-carnitine plus thalidomide. Treatment duration is 4 months. The preliminary results (Mantovani et al., 2008) on 125 patients showed that L-carnitine supplementation (4 g/day for 4 months) was able to induce a significant improvement of fatigue and performance status. A second interim analysis (submitted to AACR Annual Meeting, 2008) was carried out in October 2007 after the enrolment of 204 patients. It showed a significant improvement of fatigue in arms 3 (L-carnitine) and 5, a significant decrease of IL-6 in arm 3 (L-carnitine) and a significant decrease of TNF-a in arms 3 (L-carnitine) and 5. The interim analysis showed that arm was clearly significantly less effective than the others for primary efficacy endpoints: therefore it will be withdrawn from the study. As for toxicity, no toxicity of any grade was observed. Only one patient, randomized in arm 1, discontinued MPA for a deep vein thrombosis event which occurred during treatment.
1.4.2
Cancer-Related Fatigue
Fatigue is a multidimensional symptom defined by the National Comprehensive Cancer Network as ‘‘an unusual, persistent, subjective sense of tiredness related to cancer or cancer treatment that interferes with usual functioning’’ (Mock et al., 2000). Fatigue can be described
2055
2056
120
L-Carnitine
Supplementation on Quality of Life and Other Health Measures
in terms of perceived energy, mental capacity, and psychological status and can impair daily functioning, self-care capabilities and desire to continue treatment (Winningham et al., 1994). Fatigue may be caused by the disease, the antineoplastic therapies and/or a broad range of physical and psychological comorbidities. Suggested pathogenetic mechanisms include: an imbalance in energy metabolism due to increased energy requirement and decreased availability of metabolic substrates, an abnormal production of substances that impair metabolic homeostasis or normal muscle functioning (e.g., cytokines) and anemia. Other suggested mechanisms refer fatigue to the pathophysiology of sleep disorders and major depression (Gramignano et al., 2006). Some studies demonstrated the ability of L-carnitine supplementation to ameliorate fatigue in cancer patients (> Table 120-2). In detail, Graziano et al. (2002) treated 50 non anemic cancer patients undergoing chemotherapy with 4 mg/day of L-carnitine: after 1 week fatigue (assessed by the Functional assessment of Cancer Therapy-Fatigue questionnaire) significantly improved in almost all patients. In a subsequent study, Cruciani et al. (2004) showed that 1 week of carnitine supplementation (from 250 mg to 1,750 mg/day) in cancer patients with fatigue and carnitine deficiency resulted in improved symptoms of fatigue, reduced depression and sleep disruption, but no change in performance status. More recently, two studies have been published. The first by Gramignano et al. (2006) supplemented cancer patients undergoing chemotherapy with 6 g daily of oral L-carnitine for 4 weeks and reported a
. Table 120-2 Summary of clinical studies on L-carnitine supplementation and QL outcomes in cancer patients Study design
Study population
L-carnitine
treatment
Outcomes
Findings
References
Prospective open non randomized phase II study
12 advanced cancer patients with fatigue; tumor at different sites
6 g/day for 4 weeks
Fatigue by MFSI-SF; QL by QoL-OS and EQ-5D vas; LBM; grip strength, appetite, ROS, proinflammatory cytokines
Amelioration of fatigue; Increase of LBM and appetite
Gramignano et al. (2006)
Phase I/II study
21 advanced cancer patients with carnitine deficiency and moderate to severe fatigue
Escalating doses (250, 750, 1,250, 1,750, 2,250,2,750, 3,000 mg/day) for 7 days
Fatigue by BFI; depressed mood; quality of sleep; Performance Status
Decrease of fatigue and depressed mood, improvement of sleep quality
Cruciani et al. (2006)
Prospective non randomized phase II study
50 non anemic cancer patients undergoing chemotherapy
4 g/day for 7 days
Fatigue by FACT-F
Amelioration Graziano of fatigue in et al. (2002) 45/50 patients
Abbreviations: MFSI-SF multidimensional fatigue symptom inventory-short form; QoL-OS QoL focused on symptoms of oxidative stress; LBM lean body mass; ROS reactive oxygen species; BFI brief fatigue inventory; FACT-F functional assessment of cancer therapy-fatigue
L-Carnitine
Supplementation on Quality of Life and Other Health Measures
120
. Figure 120-4 Changes of fatigue assessed by the Multidimensional Fatigue Symptom Inventory-Short Form (MFSI-SF) questionnaire during L-carnitine supplementation in cancer patients. Histograms represent the mean score of Multidimensional Fatigue Symptom Inventory-Short Form (MSFI-SF) questionnaire. The difference between mean values of MFSI-SF score at baseline (t0) versus 2 (t1) and 4 weeks (t2) of treatment with LC supplementation was statistically significant (p < 0.05 and p < 0.001, respectively) (Gramignano G et al., 2006)
significant reduction of fatigue as measured by the MFSI-SF (> Figure 120-4) (Stein et al., 2004). Moreover, they reported that lean body mass and appetite increased significantly while ROS decrease not significantly following carnitine supplementation. The second by Cruciani et al. (2006) was a phase I/II clinical trial in 38 carnitine-deficient patients with advanced cancer showed that L-carnitine supplementation (250, 750, 1,250, 1,750, 2,250, 2,750, 3,000 mg/day administered in two daily doses for 7 days) was very well tolerated up to doses of 3 g/day, increased plasma carnitine and improved fatigue in a dosedependent manner as well as mood and sleep quality in most patients. This study suggests that the prevalence of carnitine deficiency in cancer patients is higher than reported.
1.5
L-Carnitine
Supplementation on QL in Other Clinical Settings
It has been widely established that patients with end-stage renal disease (ESRD) undergoing chronic hemodialysis treatment have reduced plasma and tissue L-carnitine levels. Disturbances in carnitine homeostasis during long-term hemodialysis have been implicated as a contributing factor for certain clinical problems in ESRD patients, including skeletal myopathies, poor exercise performance, cramps, weakness and fatigue. Therefore, a large number of studies have been conducted over the past 20 years to assess the efficacy of L-carnitine supplementation in treating certain dialysis-related clinical disorders (Reuter et al., 2008).
2057
2058
120
L-Carnitine
Supplementation on Quality of Life and Other Health Measures
A single blind, randomized, placebo-controlled clinical trial on 20 patients on maintenance hemodialysis showed that L-carnitine supplementation (20 mg/kg for 8 weeks) resulted in a significantly greater increase in the total SF-36 score, the overall physical health score, overall mental health score and the scores of the physical functioning, general health, vitality, social functioning and mental health, in comparison to the placebo group (Rathod et al., 2006).The results of this study are substantiated by those of another recent doubleblind, randomized, controlled study in 50 hemodialyzed patients, which showed that the role-physical and SF-36 physical component summary scores improved significantly from baseline in the L-carnitine (2 g i.v.) group compared to the control group (Steiber et al., 2006). Moreover, carnitine supplementation in patients on hemodialysis may reduce insulin resistance, enhance the response to erythropoietin resulting in improved anemia, inflammatory and antioxidant status, protein balance, lipid profile, and cardiac function (Guarnieri et al., 2007). Although the administration of L-carnitine in patients with chronic renal failure may improve subjective symptoms such as malaise, muscle weakness, intradialytic cramps, hypotension, and quality of life in selected maintenance dialysis patients, the full evidence is insufficient to recommend its routine administration in chronic renal failure (CRF) patients (NKF-K/DOQI, 2000). In clinical settings other than CRF such as celiac disease patients, mutliple sclerosisis patients and elderly patients, L-carnitine deficiency has been assessed and its role in reducing energy production and thus inducing fatigue has been hypothesized. Oral L-carnitine (2–6 g daily) was effective in ameliorating fatigue in celiac disease patients (Ciacci et al., 2007) and 2 g daily in patients with multiple sclerosis treated with interferon b (Lebrun et al., 2006). A recent randomized, placebo controlled, double blind study evaluated the efficacy of L-carnitine (2 g daily) on physical and mental fatigue and cognitive functions in a population of 66 centenarians. At the end of the study period, the L-carnitine-treated patients, compared with the placebo group, showed significant improvements in total muscle mass, physical and mental fatigue, fatigue severity and cognitive function (Mini-Mental State Examination (MMSE)) (Malaguarnera et al., 2007).
2
Conclusion
If health-related quality of life is the key goal for health promotion, then it is captured only partly by the existing mortality and morbidity indexes. Researchers now urge government agencies and health care providers to begin collecting quality-of-life data on the populations they serve. In the Healthy People 2000 report, the chief goal of health promotion was to increase the span of healthy life. The focus was on mortality and morbidity data and symptom checklists as the principal measures of ill health. In contrast, the new emphasis in the Healthy People 2010 report is on quality of life and overall well-being. Helping people to increase life expectancy and improve their quality of life is the primary goal of the Healthy People 2010 report (Drewnowski and Evans, 2001). Quality of life and supportive care are complementary concepts in the care of cancer patients. Neither is easy to define. Both have received increasing attention in the medical literature of recent years. From the clinical perspective, supportive care is one means toward the end of improving patients’ quality of life. In order to evaluate our degree of success in this endeavor, we must agree on operational definitions of those aspects of care and the outcomes
L-Carnitine
Supplementation on Quality of Life and Other Health Measures
120
we wish to study, then devise, validate and apply appropriate measures. Supportive care covers a variety of topics including symptom control, anti-infective measures, nutritional supplements and psychosocial support. The aspects of quality of life studied include physical, emotional, psychological and (less commonly) spiritual wellbeing. Symptoms influenced by the disease or its treatment are often included in the assessment and QL scales have been used as outcome measures in comparing treatments (Coates, 1997). Nutritional support is a critical component of supportive care. However, both palliative and curative treatment of cancer should be accompanied by specific and patient-tailored nutritional intervention aiming at improving performance status and QL (Capra et al., 2001). During curative treatment, nutritional support has an additional and specific role, i.e., to increase the response to treatment, decrease the rate of complications and reduce morbidity. In palliative care, nutritional support aims at improving patient’s QL by improving clinical symptom management (nausea, vomiting, pain, fatigue, etc.). It should be emphasized that, although nutritional intervention is not a primary part of specific cancer treatment, it is necessary in all stages of the disease and of the therapeutic strategy. It helps to control cancerrelated symptoms, reduce postoperative complications and infection rate, shortens length of hospital stay, improves tolerance to treatment and enhance immunometabolic host response. Moreover, timely nutritional support is associated with improved QL (Peltz, 2002). The relationship between > nutritional status and QL is becoming a critical issue in oncology particularly if we consider that the significant advances in cancer therapy led to dramatically prolonged survival time or many patients, who should be able to fully live. Cancer patients reported a significant influence of nutritional status on QL parameters like physical function, psychological state and social well-being. Weight loss and other nutritional symptoms are associated with poor QL. Also a clear correlation between reduced food intake and QL exists. A low QL in turn is associated to reduced response to antineoplastic treatments (Marin Caro et al., 2007). Among the oral nutritional supplements, L-carnitine demonstrated in several studies, in cancer patients as well as in other chronic illnesses, to ameliorate QL, particularly fatigue (> Table 120-2). Moreover, several studies showed the effectiveness of L-carnitine supplementation in improving muscle strength, body weight, LBM, pain, peripheral neurotoxicity that, in turn, influence patient’s QL. More clinical trials, however, are needed to identify those patients and diseases that may derive the greatest benefit from the supplementation with this agent. Indeed, while several studies have examined the role of L-carnitine supplementation in different comorbid conditions, very few have assessed its effect on QL.
Summary Points
L-carnitine
has several antioxidant and energetic properties which make it essential in several disease. Carnitine is essential for b-oxidation of long-chain fatty acids in the mitochondria. The impact of carnitine on quality of life has a strong rationale and this has led to several researches centered especially on fatigue in different disease settings. L-carnitine supplementation has been shown to improve fatigue symptoms in cancer patients. In the present chapter we have especially examined the studies on L-carnitine effect on QL outcomes.
2059
2060
120
L-Carnitine
Supplementation on Quality of Life and Other Health Measures
Appendix 1 Most commonly used QL questionnaires and related scoring system
L-Carnitine
Supplementation on Quality of Life and Other Health Measures
120
2061
2062
120
L-Carnitine
Supplementation on Quality of Life and Other Health Measures
L-Carnitine
Supplementation on Quality of Life and Other Health Measures
120
Multidimensional Fatigue Symptom Inventory-Short Form (MFSI-SF)
2063
2064
120
L-Carnitine
Supplementation on Quality of Life and Other Health Measures
L-Carnitine
Supplementation on Quality of Life and Other Health Measures
120
2065
2066
120
L-Carnitine
Supplementation on Quality of Life and Other Health Measures
L-Carnitine
Supplementation on Quality of Life and Other Health Measures
120
2067
2068
120
L-Carnitine
Supplementation on Quality of Life and Other Health Measures
L-Carnitine
Supplementation on Quality of Life and Other Health Measures
120
References Arrigoni-Martelli E, Caso V. (2001). Drugs Exp Clin Res. 27: 27–49. Beck A, Streer R, Brown G. (1996). Manual for the Beck Depression Inventory-II. Psychological Corporation, San Antonio, TX, USA. Beck SA, Groundwater P, Barton C, Tisdale MJ. (1990). Br J Cancer. 62: 822–825. Boini S, Briancon S, Guillemin F, Galan P, Hercberg S. (2004). Health Qual Life Outcomes. 2: 4. Brennan MF. (1977). Cancer Res. 37: 2359–2364. Bruera E. (1992). Oncology. 49(Suppl 2): 35–42. Calabrese V, Giuffrida Stella AM, Calvani M, Butterfield DA. (2006). J Nutr Biochem. 17: 73–88. Caplan K. (1987). In: Aaronson NK, Beckmann J (eds.) The Quality of Life of Cancer Patients. Raven Press, New York, pp 1–9. Capra S, Ferguson M, Ried K. (2001). Nutrition. 17: 769–772. Carr AJ, Thompson PW, Kirwan JR. (1996). Br J Rheumatol. 35: 275–281. Ciacci C, Peluso G, Iannoni E, Siniscalchi M, Iovino P, Rispo A. et al. (2007). Dig Liver Dis. 39: 922–928. Coates A. (1997). Support Care Cancer. 5: 435–438. Cruciani RA, Dvorkin E, Homel P, Culliney B, Malamud S, Shaiova L. et al. (2004). Ann N Y Acad Sci. 1033: 168–176. Cruciani RA, Dvorkin E, Homel P, Malamud S, Culliney B, Lapin J, et al. (2006). J Pain Symptom Manage. 32: 551–559. Drewnowski A, Evans WJ. (2001). J Gerontol A Biol Sci Med Sci. 56(Spec No. 2): 89–94. Fitzpatrick R, Fletcher A, Gore S, Jones D, Spiegelhalter D, Cox D. (1992). BMJ. 305: 1074–1077. Fletcher A, Gore S, Jones D, Fitzpatrick R, Spiegelhalter D, Cox D. (1992). BMJ. 305: 1145–1148. Gramignano G, Lusso MR, Madeddu C, Massa E, Serpe R, Deiana L. et al. (2006). Nutrition. 22: 136–145. Graziano F, Bisonni R, Catalano V, Silva R, Rovidati S, Mencarini E. et al. (2002). Br J Cancer. 86: 1854–1857. Guarnieri G, Biolo G, Vinci P, Massolino B, Barazzoni R. (2007). J Ren Nutr. 17: 23–29. He W, Lam TK, Obici S, Rossetti L. (2006). Nat Neurosci. 9: 227–233. Heber D, Byerley LO, Chi J, Grosvenor M, Bergman RN, Coleman M. et al. (1986). Cancer. 58: 1867–1873. Hoang BX, Graeme Shaw D, Pham P, Levine S. (2007). Med Hypotheses. 69: 262–272. Katz MR, Kopek N, Waldron J, Devins GM, Tomlinson G. (2004). Psychooncology. 13: 269–280. Laviano A, Meguid MM, Guijarro A, Muscaritoli M, Cascino A, Preziosa I. et al. (2006). Curr Opin Clin Nutr Metab Care. 9: 442–448.
Lebrun C, Alchaar H, Candito M, Bourg V, Chatel M. (2006). Mult Scler. 12: 321–324. Lundholm K, Gelin J, Hyltander A, Lonnroth C, Sandstrom R, Svaninger G. et al. (1994). Cancer Res. 54: 5602–5606. Maccio` A, Madeddu C, Mantovani G. (2006). In: Mantovani G (ed.) Cachexia and Wasting: A Modern Approach. Springer, New York, pp 195–203. Malaguarnera M, Cammalleri L, Gargante MP, Vacante M, Colonna V, Motta M. (2007). Am J Clin Nutr. 86: 1738–1744. Mantovani G, Maccio A, Madeddu C, Gramignano G, Serpe R, Massa E, et al. (2008). Nutrition. 24: 305–313. Mantovani G, Maccio A, Massa E, Madeddu C. (2001). Drugs. 61: 499–514. Mantovani G, Madeddu C, Maccio A, Gramignano G, Lusso MR, Massa E. et al. (2004). Cancer Epidemiol Biomarkers Prev. 13: 1651–1659. Marin Caro MM, Laviano A, Pichard C. (2007). Clin Nutr. 26: 289–301. Mintz M. (1995). J Child Neurol. 10(Suppl 2): S40–S44. Mock V, Atkinson A, Barsevick A, Cella D, Cimprich B, Cleeland C. et al. (2000). Oncology (Williston Park). 14: 151–161. Nelson K, Walsh D. (1991). Cancer Bull. 43: 403–406. NKF-K/DOQI. (2000). Clinical Practice Guidelines for Nutrition in chronic renal failure K/DOQI, National Kidney Foundation. Am J Kidney Dis. 2000; 35(6 Suppl 2): S1–140. Peltz G. (2002). Nutr J. 1: 1. Petersen KF, Befroy D, Dufour S, Dziura J, Ariyan C, Rothman DL. et al. (2003). Science. 300: 1140–1142. Rathod R, Baig MS, Khandelwal PN, Kulkarni SG, Gade PR, Siddiqui S. (2006). Indian J Med Sci. 60: 143–153. Rebouche C. (1999). In: Shike M, Ross A (eds.) Nutrition in Health and Disease, 9th ed. Williams & Wilkins, Baltimore, pp. 505–512. Reda E, D’Iddio S, Nicolai R, Benatti P, Calvani M. (2003). Acta Diabetol. 40(Suppl 1): S106–S113. Reuter SE, Faull RJ, Evans AM. (2008). Nephrology (Carlton). 13: 3–16. Stein KD, Jacobsen PB, Blanchard CM, Thors C. (2004). J Pain Symptom Manage. 27: 14–23. Steiber AL, Davis AT, Spry L, Strong J, Buss ML, Ratkiewicz MM, et al. (2006). JPEN J Parenter Enteral Nutr. 30: 10–15. The World Health Organization. (1995). Soc Sci Med. 41: 1403–1409. The World Health Organization. (1998). Quality of Life Assessment (WHOQOL): development and
2069
2070
120
L-Carnitine
Supplementation on Quality of Life and Other Health Measures
general psychometric properties. Soc Sci Med. 46: 1569–1585. Tisdale MJ. (2005). Physiology (Bethesda). 20: 340–348. Ware JE Jr., Brook RH, Davies AR, Lohr KN. (1981). Am J Public Health. 71: 620–625. Ware JE Jr., Sherbourne CD. (1992). Med Care. 30: 473–483.
Winningham ML, Nail LM, Burke MB, Brophy L, Cimprich B, Jones LS. et al. (1994). Oncol Nurs Forum. 21: 23–36. Winter S, Jue K, Prochazka J, Francis P, Hamilton W, Linn L. et al. (1995). J Child Neurol. 10(Suppl 2): S45–S51.
121 Quality of Life Among Diabetic Subjects: Indian Perspectives K. Vijayakumar . R. T. Varghese 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2072
2
Financial Burden of Diabetes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2075
3
Ethnic Predisposition of Indians to Diabetes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2075
4 4.1 4.2 4.3 4.4 4.5 4.6 4.7
Determinants of QOL among Indian Diabetic Subjects . . . . . . . . . . . . . . . . . . . . . . . . . 2076 Demographic and Socioeconomic Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2076 Laboratory Information and Effect of Patient Education on QOL . . . . . . . . . . . . . . . . 2078 Behavioral Factors and Its Impact on QOL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2079 Psychological Factors and QOL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2080 Comorbidities/Complications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2083 Treatment Regime . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2086 Type 1 Diabetes and GDM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2087 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2088 Appendix-1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2088
#
Springer Science+Business Media LLC 2010 (USA)
2072
121
Quality of Life Among Diabetic Subjects: Indian Perspectives
Abstract: The ultimate aim of any health care giver is to relieve suffering and improve the > quality of life (QOL) of the patients. Diabetes, both due to its chronic nature and multisystemic involvement seriously affect the QOL of the patient and to an extent that of the bystanders too. Developing countries like India are witnessing a rapid surge in the prevalence of lifestyle diseases including diabetes. Given the ethnic predisposition of Indians to develop diabetes and to have poor Health Related Quality of Life (HRQOL), as evidenced by various studies, more efforts must be taken to tackle this problem of diabetes and the burden it places. The development of complications worsens the suffering and the burgeoning cost on disease management further worsens the QOL. Measures to improve QOL in Indian diabetic subjects deserves special mention in the light of the fact that India could well be the diabetes capital of the world unless serious measures are taken to stop the diabetic epidemic. This chapter deals with the quality of life among diabetic subjects in the Indian perspective. We have also dealt with the factors affecting QOL in these subjects. The role of demographic and socioeconomic factors, patient education, behavioral factors, psychological factors, co morbidities, complications and treatment regime on QOL have been dealt in detail. An effort is also made to evaluate QOL in type 1 diabetes and gestational diabetes. List of Abbreviations: DM, > diabetes mellitus; HRQOL, health related quality of life; QOL, quality of life
1
Introduction
India is a country which is undergoing demographic and epidemiological transition. This results in twin burden of communicable and non communicable diseases. This is most marked in states like Kerala that underwent this change early. Migrant workers from Kerala to the Middle East and Europe are sending in foreign currency which translated into huge amounts against a devalued rupee. With the current economic growth, most states in India are catching up with this epidemiological trend. The prevalence of all lifestyle diseases like hypertension, coronary artery disease and type 2 diabetes have been rising in India (Hypertension study group, 2001; Kutty et al., 2000; Raghupathy et al., 2007). Diabetes receives special attention because studies show a predisposition of Indians to develop diabetes (Radha et al., 2003). King et al. predict that world wide prevalence of diabetes in adults will rise from 4.0 (1995) to 5.4% (2025). The number of adults with diabetes in the world is expected to rise to 300 million in the year 2025 and the major part of this numerical increase will occur in developing countries attributing 170%. Thus, by the year 2025, >75% of people with diabetes will reside in developing countries and the countries with the largest number of people with diabetes will be India, China, and USA (King et al., 1998). > Figure 121‐1 (Wild, 2004) gives a comparison of the total number of diabetic subjects in various countries in 2000 and a projection for the same in the year 2030. > Figure 121‐2 (Mohan, 2004) shows the rise in prevalence of diabetes in India in the recent years. Recent surveys (2001) like the National Urban Diabetes Survey (NUDS), carried out in six cities, reported that for urban India the age-standardized prevalence rates is 12% for diabetes and 14% for impaired glucose tolerance (IGT) (Ramachandran et al., 2001). This assumes
Quality of Life Among Diabetic Subjects: Indian Perspectives
121
. Figure 121‐1 The number of diabetic subjects in various countries in 2000 AD and 2030 AD. (NOTE – The ten countries chosen were the countries having maximum prevalence of diabetes in 2000 AD. Data for Russian federation and Italy in 2030 was not projected in the study. Source of data: ‘‘Copyright 2004 American Diabetes Association. From Diabetes Care, Vol. 27: 1047–1053, 2004. Reprinted with permission from The American Diabetes Association’’)
. Figure 121‐2 The rise in the prevalence of diabetes in India. Source: ‘‘Mohan V, Madan Z, Jha R, Deepa R, Pradeepa R (2004). Int J Diab Dev Countries, 24: 29–35.’’ ‘‘Copyright – The International Journal of Diabetes in Developing Countries’’
2073
2074
121
Quality of Life Among Diabetic Subjects: Indian Perspectives
importance in the light of the fact that studies found a much higher rate of progression to diabetes among Indian individuals with IGT than reported in Caucasian populations (Prabhakaran et al., 2007). This points to a rapid increase in prevalence in the near future. The ratio of IGT to frank diabetics is small in states like Kerala compared to the Caucasian population. The variations in this pattern should be explored in the Indian/developing nation context. Two population based studies, the Chennai Urban Population Study (CUPS) and Chennai Urban Rural Epidemiological Study (CURES) showed a marked increase in the prevalence of diabetes within a short span of 5 years. These studies confirm our fear of a diabetic epidemic in our country. Thus India could get the dubious distinction of being the diabetes capital of the world, which would be bearing the brunt of the epidemic nature of diabetes. Worldwide surveillance of diabetes is therefore a necessary first step toward its prevention and control. All lifestyle diseases cause considerable pain, agony and suffering on account of the disease and its complications. This coupled with the financial burden of treatment ultimately results in worsening of QOL. A chronic and multi-systemic disease like diabetes and the associated complications ranging from retinopathy to foot ulcers adversely affect the QOL of the subject in a profound manner. QOL may be thought of as a multidimensional concept incorporating an individuals’ subjective perception of physical, emotional and social wellbeing, including both a cognitive component (satisfaction) and an emotional component (happiness). Kumar (2004) described three major goals of human existence in India including: hedonistic (pleasure oriented), collective (QOL and wellbeing of an individual as inextricably bound with that of others), and transcendental (QOL and wellbeing in spiritual terms) (Kumar, 2004). Almost every person with diabetes is burdened by the manifold demands of the disease and its management. (> Table 121‐1) (Mohan, 2004) The magnitude of impact of diabetes on HRQOL was reported to be equivalent to that of having cardiovascular conditions, cancer and chronic respiratory disease) (Sprangers et al., 2000). Treatments for chronic disorders may damage QOL of patients even if they improve their health. In evaluating outcomes of diabetes care, it is essential to assess the impact of diabetes
. Table 121-1 The burden due to diabetes for an Indian patient Direct costs
Indirect costs
Intangible costs
Medical
Non medical
Consultation
Transportation
Man days lost
Reduced life expectancy
Investigation
Time utilized for care
Low productivity
Pain and discomfort
Medication
Disability payment
Stress
Management
Social security
Anxiety
Hospitalization
Depression
Treating complications
Insecurity Inconvenience
Health programs
Lower quality of life
Source: ‘‘Mohan V, Madan Z, Jha R, Deepa R, Pradeepa R (2004). Int J Diab Dev Countries. 24:29–35’’
Quality of Life Among Diabetic Subjects: Indian Perspectives
121
on QOL. It informs us not only about the patients’ experience of living with the condition, but also shows us ways in which we could improve diabetes care. If QOL is made a target of clinical and research efforts and seen as at least as important as the target of improved health, we are more likely to achieve both. It is thus suggested that targets of diabetes management are more likely to be achieved if the importance of protecting and improving QOL is recognized and monitored alongside biomedical outcomes such as blood glucose levels (Singh and Bradley, 2006). For measuring QOL, ADDQOL (Audit of Diabetes-Dependent Quality-of-Life) questionnaire was found to be culturally appropriate, valid and reliable in a sample of Asian patients with diabetes seen at a tertiary healthcare institution in Singapore. Individualized instruments such as the ADDQOL may potentially be a useful alternative to CAT (Computerized Adaptive Testing) in developing countries (Wee et al., 2006a).
2
Financial Burden of Diabetes
The economic impact of a disease like diabetes is also a cause of worry, for the individual and the country. India accounts for 23.5% of the world’s disability adjusted life years lost due to diabetes (DALYs) (World Bank, 1993). A study conducted in families having type 1 diabetes in Chennai reported that a median figure of Rs. 13,980 (US$310) was spent annually for diabetes by the families and the median percentage of annual income spent on diabetes treatment was 22% for the entire group (Shobhana et al., 2002). The Cost of Diabetes in India (CODI) study revealed that the mean direct annual cost for outpatient care for diabetes individuals was Rs. 4,724 and the total indirect cost for non-earning diabetic patients was estimated to be Rs. 9,748 while for earning members it was Rs. 16,831 (Kapur, 2000). Diabetes has thus become a great economic challenge as it drains between 5 and 25% of the family income of an average Indian (Shobhana et al., 2000). International studies show that the median annual direct medical costs for subjects with diet-controlled type 2 diabetes, BMI 30 kg/m2, and no micro vascular, neuropathic, or cardiovascular complications were $1,700 for white men and $2,100 for white women. A 10 kg/m2 increase in BMI, treatment with oral anti diabetic or antihypertensive agents, diabetic kidney disease, cerebrovascular disease, and peripheral vascular disease were each associated with 10–30% increases in cost. Insulin treatment, angina, and myocardial infarction were each associated with 60–90% increases in cost. Dialysis was associated with an 11-fold increase in cost (Brandle et al., 2003).
3
Ethnic Predisposition of Indians to Diabetes
The role of ethnicity in predisposing Indians to diabetes also needs to be looked into. The prevalence rates of diabetes, premature coronary artery disease [CAD] and cardiovascular disease [CVD] are higher in Asian Indians compared to other ethnic groups (Ma et al., 2003) and there is also strong evidence that Indians have a stronger genetic predisposition to diabetes (Radha et al., 2003). This is thought to be due to typical Asian Indian phenotype with higher percentage of body fat and increased waist to hip ratio for any given body mass index (BMI) which predisposes to diabetes and the metabolic syndrome (Joshi, 2003).
2075
2076
121
Quality of Life Among Diabetic Subjects: Indian Perspectives
Thumboo et al. (2003) report that ethnicity and socioeconomic status are independently associated with clinically important differences in Health Related QOL in a multi-ethnic, urban Asian population. Although Indians reported the highest overall QOL, independent of socio-economic and health status, in non diabetic subjects from a population of Indians, Chinese and Malays in Singapore (Ng et al., 2005), studies in Asian diabetic subjects showed that, in addition to having the highest risk of mortality (Ma et al., 2003), diabetic Indians were also at the highest risk of impaired HRQOL, compared with diabetic patients of other ethnicities (Wee et al., 2006b). Few international studies have shown that compared to host populations, migrant Indians have higher mortality rates in diabetic subjects compared with non-diabetic subjects (Ma et al., 2003). This has to be viewed in the light of higher propensity of Asian Indians to develop diabetes and its complications. This in spite of studies showing that immigrants tend to enjoy better health status than the native-born population, even when those immigrants are lower in socioeconomic status warns about the propensity to develop diabetes in Indian population However, studies state that with increasing length of stay in the United States and adaptation to mainstream behavior, the health status of immigrants deteriorates (Hummer et al., 1999).
4
Determinants of QOL among Indian Diabetic Subjects
4.1
Demographic and Socioeconomic Factors
In the developed countries, the majority of people with diabetes are aged at or above 65 years while in developing countries, the majority of people (King et al., 1998) with diabetes are in the age range of 45–64 year (> Figure 121‐3) (Mohan, 2004). This fact warrants urgent
. Figure 121-3 The age wise distribution of the prevalence of diabetes in India. Source: ‘‘Mohan V, Madan Z, Jha R, Deepa R, Pradeepa R (2004). Int J Diab Dev Ctries. 24:29–35.’’ ‘‘Copyright – The International Journal of Diabetes in Developing Countries’’
Quality of Life Among Diabetic Subjects: Indian Perspectives
121
attention as in countries like India it translates into more morbidity, economic expenditure and more complications for a longer period all independently contributing to a poor QOL. Many overseas studies (Lori, 2003) conducted have concluded that youth with type 1 diabetes enjoyed similar quality of life as to a non-diabetic youth population. Studies from the state of Kerala in India also point to the fact that diabetic subjects in the younger age group, i.e., 10–29 were found to be enjoying twice as better a quality of life compared to other age groups proving that young subjects have a better QOL. The lesser duration of diabetes and lesser tendency to have any complications in this group could be a reason. This age group is more likely to avail better medical facilities and is likely to enjoy more family support (Varghese et al., 2007). Though there are more women than men with diabetes worldwide (King et al., 1998) only 55% of women were found to have a good quality of life while 72.5% men had good QOL, in Kerala (Varghese et al., 2007). This calls for the setting up of more support groups for women with diabetes. The fact that the social support for better QOL focused on the nuclear family (Sridhar et al., 2007) also calls for better support from spouse and children especially for women. Sridhar et al. (2007) report that compared to women, men had better adjustment with disease, coped better, integrated better, had better quality of life and well-being. Better social life and physical activity might contribute to higher satisfaction levels in men. Studies have shown that men were more confident of their ability to control diabetes and reported a higher QOL (Rubin and Payrot, 1999). In the elderly non diabetic population it is seen that women are poorer and generally suffer more morbidity than men in old age, even though their death rates are lower. The better-off among the elderly enjoy a quality of life much superior to their poor brethren (Vijayakumar et al., 1994). Thus, in transitional societies such as Kerala, socioeconomic status and gender play a significant role in determining the quality of life of the elderly, a finding which may have some policy implications with emphasis to be laid on women especially the poor. Studies prove that age- and sex-standardized prevalence of IGT and NIDDM was remarkably similar across ethnic (religious) groups (16.2 and 12.4% in Hindu Indians, 15.3 and 13.3% in Muslim Indians (Dowse et al., 1990). Religious participation may be associated with a number of socioeconomic, lifestyle, ethnic, and geographic factors that may affect health. Spirituality, as measured by The Spiritual Transcendence Scale (STS), was hypothesized to predict more positive health behaviors and less psychological distress (Vohra, 2006) among South Asians. Sharma (2002) suggested that religious rituals and activities can provide a ‘‘high degree of leisure,’’ and are considered to be ‘‘a powerful cultural force that integrates individuals to the larger society’’ and keeps people healthy physically and mentally. Studies prove that socioeconomic status is associated with clinically important differences in health related QOL in a multi-ethnic, urban Asian population (Thumboo et al., 2003). Ramachandran et al. (2002) in a study conducted in southern India observe that diabetic subjects from a lower socioeconomic group have a higher prevalence of cardiac disease, neuropathy and cataract but a lower prevalence of retinopathy compared to those from a higher socioeconomic group and that the risk variables including hyperglycemia, dyslipidemia, hypertension, smoking and alcohol consumption were more in the low socioeconomic group. This study suggests that while the prevalence of diabetes is higher among the urban higher socioeconomic group, higher rates of complications of diabetes are observed in the lower socioeconomic group in the developing countries.
2077
2078
121
Quality of Life Among Diabetic Subjects: Indian Perspectives
Studies in Kerala (Varghese et al., 2007), the state which has the best health indices, show that income and quality of life showed a linear relationship and that those who face economic hardships are reluctant to place their own medical needs above the needs of their own family members. Similar results have been reported from other countries also (AHRQ, 2000; Rubin, 2000). Those who followed a mixed diet (non-vegetarians) were found to have a better quality of life compared to pure vegetarians (odds ratio = 3.28) (Varghese et al., 2007). This maybe due to the fact that as the disease progresses and complications start to occur, the patients are compelled to change their lifestyle. In Kerala only very few people follow a complete vegetarian diet and they are those, who had to change their diet. Such people tend to have lower levels of satisfaction, due to diet and more importantly due to the disease condition. In contrast the non-vegetarians maybe in their initial stages of disease enjoy a better quality of life. It is also noted that people on diet and exercise as mode of treatment are most happy which may be due to the fact that they are in the initial stages of disease with not much of complications. Studies (Sekar et al., 2006) have also shown that there is a therapeutic role of pulse based diets in type 2 diabetes, but caution that it is a matter of conjecture whether the therapeutic effects of modified pulse carbohydrate diet on type 2 diabetes of differing severity will be elicited. Education level and mode of treatment had a linear relationship with quality of life. As the educational level increased the quality of life increased. Those who are better educated will have a better understanding of the disease, its effect on them, and will avail themselves the best treatment they can afford. People who have a higher education generally belong to the upper strata of society and have a better financial status. So they are able to afford better treatment facility (Varghese et al., 2007).
4.2
Laboratory Information and Effect of Patient Education on QOL
In general, Asian Indians have greater levels of total and central fat than Europeans for a similar Body Mass Index (WHO, 2004) thus predisposing them to diabetes. The key, in prevention of type 2 diabetes, and reduction of the risk of cardiovascular disease in people (have higher) lies in reduction of Body Mass Index. In planning national measures for the prevention of type 2 diabetes, both high-risk group and the general population should be targeted simultaneously with lifestyle modification the primary goal through a stepwise approach. In addition, it is important that all activities are tailored to the specific local situation (Alberti et al., 2007). Lau et al. (2004) report that improved diabetic control is associated with improved mental, but not physical QOL over a 1 year period in the community setting. This may reflect both mental empowerments garnered from proactive disease management and the burden of more complex anti-diabetic regimens. A decrease in HbA1c values was associated with a concomitant improvement in mental QOL scores over the study period. However, there is no evidence that a decrease in HbA1c values is associated with concomitant changes in physical QOL scores over 1 year. Behavior and lifestyle changes are the keys to successful self-management of diabetes and this call for stress to be laid on patient education for improving the QOL.
Quality of Life Among Diabetic Subjects: Indian Perspectives
121
Studies have proved that twice as many patients who were aware of diabetes and its consequences were free of complications when compared to those who were not aware and had similar diabetes duration. Subjects with strong belief in self-efficacy and an optimistic outlook had better satisfaction and quality of life (Rose et al., 2002). For improving the QOL patient education must be comprehensive and should include: information about diabetes and its treatment, nutritional management, physical activity, medications, glycemic monitoring, prevention, detection and treatment of complications, and finally psychosocial adjustment. The patient education is best given by a team which should have a collective (Rayappa et al., 1999) combination of expertise in medical treatment, medical nutrition, therapy, teaching skills, and behavioral psychology (Mensing et al., 2002). Studies suggests that, pharmacist facilitated patient counseling has an impact in improving the perception about disease, diet and life style changes and in turn on glycemic control and overall quality of life in diabetic patients. This procedure can be considered as an important element in implementing the disease management program (Adepu et al., 2007). Trials have shown significant benefits to quality of life from training in insulin dose adjustment. This provides dietary freedom without loss of diabetes control (Singh and Bradley, 2006). The importance of questionnaires in local languages have been stressed upon in many studies. A study in U.K reported that 48% of Asian patients were unable to read, while only 26% spoke and 20% read English, emphasizing the need for education in Asian languages using oral and visual teaching methods (Wilson, 1993). In Asian diabetic patients cultural differences were identified such as the use of alternative therapy to supplement treatment (33%) and the large number of vegetarians (61%) which also has to be borne in mind whilst implementing education programs (Wilson, 1993).
4.3
Behavioral Factors and Its Impact on QOL
It was seen that sleep disorders were more common in diabetic subjects (33.7 vs. 8.2% in controls; P < 0.01) and they report higher rates of insomnia and excessive day time sleepiness (Sridhar and Madhy, 1994). Younger patients or those had onset of diabetes at a younger age had higher levels of sleep disturbances. There was a significant association of sleep disturbances with the presence of cough, dyspnea, nocturnal cramps, paresthesia and burning of soles (Sridhar and Madhu, 1994). In type 2 diabetes sleep disturbances may be related to obesity and obesity related sleep disorders like obstructive sleep apnea (Alon et al., 2006), but it was observed that body weight in type 2 diabetes mellitus was not related to the prevalence of sleep disorders in south Indian samples (Sridhar and Madhu, 1994; Sridhar and Venkata, 2006). Sleep disturbances may be due to physical discomfort, psychosocial factors, and fluctuations in metabolic control and perhaps also hypoinsulinemia. Rapid changes in glucose levels during sleep cause awakenings and complaints of insomnia. Quality of life is affected and coping with the disease is made difficult by sleep disorders (Sridhar and Madhu, 1994). Painful Diabetic Peripheral Neuropathy is also associated with considerable sleep impairment. Given the recognized association between sleep impairment, type 2 diabetes and metabolic and affective disturbance, and the known adverse impact of affective disturbance on diabetes self-care, addressing these features- pain, sleep, and affective disturbance-is
2079
2080
121
Quality of Life Among Diabetic Subjects: Indian Perspectives
an important aspect of care for patients with painful Diabetic Peripheral Neuropathy (Zelman et al., 2006). Thus, physicians caring for persons with diabetes must be able to recognize, diagnose and manage sleep disturbances in their patients, when they occur. Bladder complaints occur in diabetes due to increased urine production associated with Diabetes mellitus and also due to Detrusor laxity, which leads to large atonic bladders, which thus cannot generate sufficient intravesicular pressure that is required to initiate urination. The bladder is hyper distended and this causes ‘‘overflow incontinence’’ (Gayathri, 2001). Patients with overactive bladder reported more health problems and had an average of 84% more yearly visits to a physician than those without overactive bladder (Wagner et al., 2002). It was observed that the persons for whom urinary incontinence was not a problem had an odds ratio of 4.93 for good QOL. For those whom frequency of micturition is above normal, the habit will be a nuisance and a source of embarrassment especially in social functions. They will be unable to take long trips. And their sleep is likely to be disturbed. It is clear that this will lead to lesser satisfaction levels and hence a lower quality of life for the patient (Varghese et al., 2007). Many elderly would be thankful for the marked relief of nocturia and the good quality of life that insulin bestows on them. Hence the compliance with insulin regimens is usually good among the elderly (Ganesan et al., 1994). The knowledge about smoking and diabetic complications is poor in India. A survey shows that only 0.01% of respondents know that smoking aggravates diabetes (Mohan et al., 2005). A few works suggest that some diabetic smokers were unable or unwilling to acknowledge their increased risk for complications and were also less motivated to follow the physicians’ advice, which may have contributed to the disparities in the recommended care for diabetes (Hosler et al., 2007). A study in Kerala found that smokers were 1.25 times as happy as non-smokers. This odd finding may be because those who admitted to smoking were still in their early stages of disease and enjoyed a relatively better quality of life free from complications (Varghese et al., 2007). Studies have shown that smoking aggravates the micro-vascular complications of diabetes mellitus like foot ulcers and cessation of smoking helps in its prevention (Singh et al., 2005) and thus improving QOL. Wei et al. observed an elevated risk of developing type 2 diabetes in both nondrinkers and men with high alcohol intakes, when compared with men who reported moderate alcohol intake. Men with a high alcohol intake may be able to reduce their risk of developing type 2 diabetes if they drink less (Wei et al., 2000). Diabetic subjects consuming alcohol were found to have at least 1.67 times a better quality of life than the non-alcoholics, probably as they were still in their early stages of disease and enjoyed a relatively better quality of life free from complications (Varghese et al., 2007).
4.4
Psychological Factors and QOL
In a study conducted in Kerala, India, (Varghese et al., 2007) (> Tables 121‐2 and > 121‐3) multivariate analysis established that satisfaction in social relationships and friendships, time set apart for management of diabetes, feeling of physical illness and worry of diabetic
Quality of Life Among Diabetic Subjects: Indian Perspectives
121
. Table 121-2 Determinants of QOL among diabetic subjects, its OR and 95% CI- A study done in Trivandrum, Kerala, India Variable
Odds ratio
95% CI
Significant variables accounting for maximum variation in QOL Satisfaction in social relationships & friendships
17.84
3.95–22.02
Time set apart for management of diabetes
27.45
11.93–64.53
Feeling of physical illness
3.21
1.32–8.07
Worry of diabetic complications
8.57
2.75–29.72
Satisfaction with treatment
9.23
3.95–22.02
Gender (Men)
2.16
1.13–4.15
Diet (Non Vegetarian)
3.28
1.32–8.29
Physical activity (Present)
8.33
3.55–20.63
Family history (Present)
1.92
1.03–3.57
Variables found to be significant as determinants of QOL
Diabetic friends (Present)
2.62
1.37–4.99
Sex life (Good)
3.25
0.93–12.44
Urinary incontinence (Absent)
4.93
2.49–9.86
Expression of diabetes frequently
1.53
0.82–2.84
Overcrowding
1.93
0.82–4.58
Co morbidities
0.97
0.51–1.87
Diabetic complications
1.49
0.77–2.92
Smoking
1.25
0.45–3.64
Alcohol
1.67
0.62–4.69
Family support
2.41
0.66–9.17
F.B.S
1.10
0.57–2.13
Age
2.50
0.26–59.87
Variables not found to be significant as determinants of QOL
‘‘Reprinted from Varghese RT, Salini R, Pradeep A, Reeshma KK and Vijayakumar K (2007). Diabetes Metab Syndr Clin Res Rev. 1(7): 7. Copyright (September 2007) with permission from Elsevier’’
complications, contribute to major variation in QOL and accounted for 46.5% variation of QOL in men. Whereas in women these factors coupled with treatment satisfaction accounted for 65.4% variation (Varghese et al., 2007). A study in Japan showed that four factors separated out, accounting for 50.3% of the variance in HRQOL in diabetic patients. They are – loneliness factor from restricted social activities (factor 1), getting worse factor (factor 2), bleak future factor from physical and psychological troubles in daily life with diabetes (factor 3), and negative feeling factor from severe diet control (factor 4) (Adachi and Oyamada, 2006). The factors are thus similar in Indian and Japanese patients except for the diet factor in the Japanese. Indians may not feel diet modification worsening their QOL
2081
2082
121
Quality of Life Among Diabetic Subjects: Indian Perspectives
. Table 121-3 Major determinants of QOL among men and women with adjusted OR- A study done in Trivandrum, Kerala, India Adjusted O.R
Determinant of QOL
B
S.E
SIG
Satisfaction with respect to social relationships and friendships (social rel)
M
4.4
1.476 0.549 0.007
F
5.5
1.697 0.582 0.004
Feel physically ill on account of diabetes (ill)
M
4.3
1.456 0.667 0.029
F
2.5
0.91
M
3.3
1.198 0.420 0.004
F
4.8
1.577 0.674 0.019
Satisfaction with respect to time for management of diabetes M (includes time for medical checkups, lab for assessing sugar F levels) (time)
1.9
0.641 0.213 0.003
2.1
0.749 0.195 0.000
Satisfaction with respect to medical treatment (happy rx)
M
-
-
F
2.4
0.884 0.647 0.172
Worry about diabetic complications (complication)
0.578 0.116
-
-
‘‘Reprinted from Varghese RT, Rekha S, Pradeep A, Reeshma KK, Vijayakumar K (2007). Diabetes Metab Syndr Clin Res Rev. 1(7), 7.Copyright (September 2007) with permission from Elsevier’’
due to better family support and dietary adjustment by all members of family or by special efforts to make food palatable. Clinician’s awareness and early intervention for psychosocial problems (for example, inadequate family support) could thus improve diabetes control (Tan et al., 2005). Studies in India also prove that the social support was focused on the nuclear family (spouse principally, and children to a lesser extent) (Sridhar et al., 2007). The role of the family in the management of type-I diabetes is gaining recognition. In countries where the adolescent is dependent on the family for medical needs, the family’s role is all the more important (Sudhir et al., 2003). Inclusion of evidence-based medicine concepts, role of family support, and enhancement in customized content and easier feedback mechanism in the web-sites can be a significant development in the direction of patient-centered diabetes care (Takurdesai et al., 2004). Active family nutritional support, as measured by culturally relevant categories, is significantly associated with control of triglyceride, cholesterol, and HbA1c levels. The findings suggest that the family is a more useful unit of intervention for individuals than for the individual alone when designing diabetes care strategies (Epple et al., 2003). This could perhaps be due to family support helping in diet modification and support for healthy lifestyle changes like exercises. The increased involvement of the pharmacist resulted in better use of their abilities and more time for physicians to spend with patients who had complications. In a study, it was clear that a period of physicians’ time was saved as a result of the pharmacist’s clearly explaining instructions on drug use to the test group of patients (Palaian et al., 2006). It is also possible that some of the observations may suggest cost-effective delivery of health care by primary care physicians rather than sub-specialist endocrinologists or diabetologists: a multi centric study of processes and outcomes among groups of specialist physicians and generalists showed
Quality of Life Among Diabetic Subjects: Indian Perspectives
121
that there was no significant difference in outcomes with specialist physicians, other than better patient satisfaction (Greenfeld et al., 2002).
4.5
Comorbidities/Complications
The population based study done by Dr. Mohan and his group following the cohort from the Chennai Urban Population Study is an important mortality study conducted in diabetic and non-diabetic subjects from India. The overall mortality rate was higher in diabetic, compared to non-diabetic, subjects (18.9 vs. 5.3 per 1,000 person years). Cardiovascular and renal diseases were the commonest causes of death among diabetic subjects, whereas mortality due to gastrointestinal, respiratory, lifestyle-related and unnatural causes were observed only among non diabetic subjects. In urban India, mortality rates are twofold higher in people with diabetes compared to non-diabetic subject (Mohan et al., 2006), portraying the severity of diabetes and its complications. Observations, that waist circumference and triglyceride levels, which are markers of obesity and CAD, are found to be lower among the non-survivors compared to the survivors (Mohan et al., 2006) in south India. This has to be viewed with caution as it may denote the genetic preponderance of these individuals and the risk that Asiatic Indians have for complications of diabetes. Subjects with diabetes and multiple co-existing chronic medical conditions have poorer HRQOL than those without these conditions (deVisser et al., 2002; Lloyd et al., 2001). Studies show that the presence of other chronic medical conditions in subjects with diabetes led to further lowering of HRQOL, the effect of which was generally additive rather than synergistic. The findings further underscore the importance of preventing and treating complications of diabetes to prevent further deterioration in HRQOL among subjects with diabetes, and also highlight the need to identify factors that may be modulated to improve HRQOL in these subjects (Hwee et al., 2005). A study (Ramachandran et al., 2002) conducted in southern India reported, that diabetic subjects from a lower socioeconomic group have a higher prevalence of cardiac disease, neuropathy and cataract but a lower prevalence of retinopathy compared to those from a higher socioeconomic group. Thus while the prevalence of diabetes is higher among the urban higher socioeconomic group, higher rates of complications of diabetes are observed in the lower socioeconomic group in the developing countries. A lower perceived QOL in Asians compared with white Europeans with End Stage Renal Disease (ESRD) has been reported. Analysis of QOL indicates that Asian patients in particular perceive kidney disease as a social burden, even if successfully transplanted (Bakewell et al., 2001). Renal transplant is the preferred mode of treatment, as it is most cost-effective and provides a better quality of life. But due to financial constraints and non-availability of organs, only about 5% of ESRD patients undergo transplant surgery Due to limited access to Renal Replacement Therapy (RRT), the ideal approach in countries like India should be to reduce the incidence of ESRD and attempt preventive measures. Preemptive transplant, reducing the duration of dialysis prior to transplant, use of immunosuppression for only up to 1 year, and availability of more deceased donor organs may be helpful to make RRT options within the reach of the common man (Singh and Bhandari, 2004).
2083
2084
121
Quality of Life Among Diabetic Subjects: Indian Perspectives
Not long ago, it was believed that erectile dysfunction is a natural sequel of ageing and when it occurred in younger men it was often labeled as psychogenic. But the situation has changed significantly and younger men are presenting with erectile dysfunction of organic etiology. Many factors like obesity, smoking, sedentary lifestyle, diabetes, coronary artery disease have been identified as potential risk factors (Kekre, 2006). Diabetes as a disease has independently been found to affect the sex life especially that of males in whom Erectile Dysfunction is more (De Berardis et al., 2002). Studies in India have shown that those happy with their sex life were found to have an odds ratio of 3.25 to have higher satisfaction levels than the unhappy group. Sex in itself is an important contributor towards satisfaction in life. This could be because most people having a good sex life were in the early stages of the disease or of younger age and still enjoying better health and likely to have few complications (Varghese et al., 2007). Studies were conducted in India wherein Sildenafil was administered orally, 1 h before intercourse, preferably on an empty stomach and an improvement in erections was seen in 56% of those with diabetes (Singh et al., 2005). Those with diabetes may require 100 mg. This could thus improve QOL. Diabetic neuropathy is a very painful condition which significantly worsens the QOL of these patients. Measures relieving the pain will thus help improving the QOL. Studies suggest that oxcarbazepine administered as monotherapy is an efficacious and safe option for the symptomatic treatment of pain associated with symmetrical diabetic neuropathy (Beydoun et al., 2004; Erdemoglu and Varlibas, 2006). First-generation antiepileptic drugs have been shown to be effective in neuropathic pain (Aaron, 2004; GutierrezAlvarez et al., 2007). The evidence supporting the use of new antiepileptic drugs in painful diabetic neuropathy like Sodium Valproate has also been shown in studies (Kochar et al., 2004). The role of GTN spray, a well tolerated drug, providing significant improvement in painful diabetic neuropathy has also been reported (Agrawal, 2007). The main conditions affecting the eye in diabetes include retinopathy and cataract. Vision impairment is associated with a significant decrease across all domains of quality of life and visual function in an older population of rural south India. These findings are consistent with reports of populations elsewhere. Participants with cataract reported difficulty across all domains of quality of life and visual function, suggesting that cataract extraction may improve quality of life and visual function in this population (Nirmalan et al., 2005). Decreased QOL was associated with the presence of glaucoma or corneal disease independent of visual acuity and with cataract or retinal disease as a function of visual acuity (Nutheti et al., 2006). Improvement in health-related QOL and visual function occurred within 3 months of cataract extraction. The speed of visual acuity recovery after phacoemulsification matched the improvement across health-related QOL functions, resulting in rapid recovery of the patients’ functional independence and health status (Mamidipudi et al., 2003). Laser photocoagulation and vitrectomy have improved the quality of life for patients with diabetic retinopathy and prevented debilitating visual loss (Sharma et al., 2000, 2001). Among persons diagnosed as having diabetes mellitus, the prevalence of foot ulcers is 4–10%, the annual population-based incidence is 1.0–4.1%, and the lifetime incidence may be as high as 25%. These ulcers frequently become infected, cause great morbidity, engender considerable financial costs, and are the usual first step to lower extremity amputation (Singh et al., 2005) and immediate worsening of QOL. In general, lower-leg amputation is suffered by
Quality of Life Among Diabetic Subjects: Indian Perspectives
121
5–25 people per 100,000; among people with diabetes, the figure is 6–8 for every 1,000. With relatively low investment, government can advance education and prevention that will result in lower rates of amputation than the unacceptable figures we see today (Das and Joshi, 2005). Diabetic foot problems are thus common throughout the world, and result in major economic consequences for the patients, their family and society. Foot ulcers are more likely to be of neuropathic origin or due to vascular problems and delayed wound healing (Das and Joshi, 2005), and therefore eminently preventable in developing countries, which will experience the greatest rise in the prevalence of type 2 diabetes in the next 20 years. People at greatest risk of ulceration can easily be identified by careful clinical examination of the feet: education and frequent follow-up is indicated for these patients. When assessing the economic effects of diabetic foot disease, it is important to remember that rates of recurrence of foot ulcers are very high, being greater than 50% after 3 years. Ulceration of the foot in diabetes is common and disabling and frequently leads to amputation of the leg. Mortality is high and healed ulcers often recur. The pathogenesis of foot ulceration is complex, clinical presentation variable, and management requires early expert assessment. Interventions should be directed at infection, peripheral ischemia, and abnormal pressure loading caused by peripheral neuropathy and limited joint mobility. Despite treatment, ulcers readily become chronic wounds. Diabetic foot ulcers have been neglected in health-care research and planning. Furthermore, the pathological processes are poorly understood and poorly taught and communication between the many specialties involved is disjointed and insensitive to the needs of patients (Jeffcoate and Harding, 2003). Costing should therefore include not only the immediate ulcer episode, but also social services, home care, and subsequent ulcer episodes. A broader view of total resource use should include an estimate of quality of life and the final outcome. An integrated care approach with regular screening and education of patients at risk requires low expenditure and has the potential to reduce the cost of health care (Boulton et al., 2005). Depression is twice as common among diabetes patients as in the general population, with 15–30% of diabetes patients meeting criteria for depression (Piette et al., 2004). Among both type 1 and type 2 diabetes depression is a prevalent and recurrent condition. Depression complicates diabetes by promoting poor glycemic control and increasing risk of other complications. Prompt treatment of depression has significant and favorable effects on mood and quality of life and beneficial effects on glycemic control among the diabetic patients (Chowdhury, 2004) with type 2 diabetes and depressive symptoms exhibited lower adherence with self-care. Studies show that higher depressive-symptom scores were associated with poor self-care behaviors significantly with poor participation in education programs (odds ratio OR = 1.21, 95% confidence interval CI = 1.06–1.38) and poor diet (OR = 1.11, 95% CI = 1.01–1.22), and marginally with poor medication taking (OR = 1.14, 95% CI = 1.00–1.31) (Hye Sook et al., 2004) thus affects diabetes management by directly affecting patients’ healthrelated quality of life, reducing physical activity levels, limiting adherence to self-care regimens, and impairing patients’ ability to communicate effectively with clinicians. Small randomized trials suggest that both antidepressant medication and cognitive behavioral therapies (CBT) or related approaches may improve not only diabetic patients’ depressive symptoms, but their physical health as well. This calls for systematic identification of diabetic patients and quality-of-care reviews, proactive patient monitoring between outpatient encounters, intensive efforts to coordinate treatment providers, increased access to CBT
2085
2086
121
Quality of Life Among Diabetic Subjects: Indian Perspectives
or related therapies addressing patients’ depressive symptoms and diabetes self care, and an emphasis on promoting physical activity to address both depressive symptoms and physiologic dysregulation. The conceptual framework developed for the care of diabetic patients will also offer general insights into the management of patients with multiple chronic medical disorders (Piette et al., 2004).
4.6
Treatment Regime
Moderately intensive aerobic exercise besides helping glycemic control in diabetes has been observed to help in normalizing ALT levels in patients with Non Alcoholic Steatohepatitis. Thus exercise boosts the QOL of all diabetic subjects especially those with cirrhosis. This calls for primary intervention in the form of diet and exercise in diabetic subjects (Baba, 2006). Studies suggest that intensive glucose control in advanced type 2 diabetes mellitus (DM) has no effect on health status over 2 years. The successful lowering of glycemia does not improve health-related quality of life nor do the increased demands of an intensive therapy regimen worsen it (Pitale et al., 2005). The researchers estimated that diet and exercise or metformin would delay diabetes, reduce complications, and improve survival. Progression of IGT to diabetes is high in native Asian Indians. Both lifestyle modification and metformin significantly reduced the incidence of diabetes in Asian Indians with impaired glucose tolerance; there was no added benefit from combining them (Ramachandran et al., 2006). The effect of diet and exercise was more favorable than the effect of metformin. Compared with no prevention, diet and exercise counseling cost about $8,800 and metformin would cost about $29,000 per quality-adjusted life-year saved. While diet and exercise had favorable cost effectiveness at any adult age, metformin was not cost-effective after 65 years of age (Hoerger et al., 2005). Inhaled insulin offers an alternative noninvasive option for premeal insulin administration, with glycemic efficacy slightly less than subcutaneous regular insulin and increased patient acceptability. Until long-term safety data are available, inhaled insulin should be reserved for non-pregnant adults with diabetes who are opposed to injections and who would otherwise delay appropriate and timely therapy with insulin (Ceglia et al., 2006). The key benefit of inhaled insulin appears to be that patient satisfaction and quality of life are significantly improved, presumably due to the reduced number of daily injections required (Mudaliar, 2007). Insulin pen devices are discreet and offer patients convenience and flexibility. These features may give patients the confidence to overcome issues of needle anxiety and the social embarrassment associated with self-injection and, therefore, may lead to improved adherence to recommended insulin dosing schedules and compliance with multiple-injection regimens (Korytkowski et al., 2006). India has a high prevalence of tropical chronic pancreatitis and such patients who undergo the Puestow procedure not only have relief from pain but also improvement of diabetes (Sidhu et al., 2001) (and thus better QOL). Insulin-dependent diabetes mellitus is associated with renal failure, diabetic retinopathy, neuropathy and vasculopathy. Simultaneous pancreas-kidney transplant was successfully
Quality of Life Among Diabetic Subjects: Indian Perspectives
121
done in India in a young diabetic with renal failure. The dual transplant has cured diabetes and renal failure and thus had a beneficial effect on his neuropathy, retinopathy and quality of life (Guleria et al., 2005). Significant relationship between participation in yoga and QOL was found in a study in India (Damodaran et al., 2002). The study found that yoga and life style intervention have several beneficial effects, in the form of improvement in glycemic control, reduction in body weight and favorable alteration in lipid profile parameters. There was also a favorable response on various risk factors like BMI, W/H ratio and quality of life. The study shows that adjunctive use of yoga life style interventions significantly improves glycemic status and various risk factors. There was also a significant reduction in doses of OHA and insulin. Being a traditional and cost effective life style management it can be used as an alternative therapy in the patients of diabetes mellitus (Agarwal et al., 2003). Overall, studies suggest beneficial changes in several risk indices, including glucose tolerance and insulin sensitivity, lipid profiles, anthropometric characteristics, blood pressure, oxidative stress, coagulation profiles, sympathetic activation and pulmonary function, as well as improvement in specific clinical outcomes. Yoga may improve risk profiles in adults with DM 2, and may have promise for the prevention and management of cardiovascular complications in the population (Innes et al., 2005). Additional high quality Randomized Control Trials are needed to confirm and further elucidate the effects of standardized yoga programs on specific indices of CVD risk and related clinical endpoints.
4.7
Type 1 Diabetes and GDM
Studies show that Ethnic Chinese youths showed better diabetic control than Malays and Indians (mean HbA1c 9.1%, 10.3%, and 11.0% respectively). Young people with better diabetic control (HbA1c < 10%) were more likely to have better quality of life (Tan et al., 2005). Behavioral intervention can be included as an effective adjunct to routine medical care in the management of young type I diabetics, especially in the management of compliance and metabolic control, enhancement of knowledge and quality of life. Studies in India also prove a significant hypoglycemic effect of camel milk when given as an adjunctive therapy. The action is presumed to be due to presence of insulin/insulin like protein in it. Its therapeutic efficacy may be due to lack of coagulum formation of camel milk in acidic media. There is no doubt that the discovery and development of oral insulin for therapeutic use is a Himalayan task. It has been observed that oral administration of insulin initiated at clinical onset of type 1 diabetes did not prevent the deterioration of beta cell function. Pozzilli et al. in IMDIAB VII study indicate that addition of 5 mg of oral insulin does not modify the course of the disease in the first year after diagnosis and probably does not statistically affect the humeral immune response against insulin. It is important to note that a certain level of scientific testing on camel milk has been already attempted and documented, particularly, insulin levels in camel milk and this scientific wisdom can be a remarkable achievement for diabetic patients. Gestational diabetes is a strong predictor of high birth weight in the newborn, which predisposes to future diabetes.
2087
2088
121
Quality of Life Among Diabetic Subjects: Indian Perspectives
Summary Points Measures must be taken to improve the QOL of Indian diabetic subjects bearing in mind
the rising prevalence, genetic predisposition, propensity to develop complications and financial burden it places both due to complications and chronic nature of disease in them. Adoption of a healthy diet and lifestyle and stress reduction techniques will in itself serve to improve QOL in a majority of diabetic subjects. Women have been found to have poor QOL, this calls for setting up of more women’s support groups. Educating the subjects about the disease, its complications and treatment will go a long way in improving QOL. The role of exercise and psychosocial adjustment must also be taught so as to improve QOL. Advices given in local language of patient is found to be very effective in improving compliance and thus improving QOL. Poor sleep, nueropathic pain and bladder complaints significantly worsened QOL. Smoking and alcoholism will further worsen QOL calling for the role of education in preventing the same. Poor social relationships and friendships, increased time for management of disease, feeling of weakness, fear of developing diabetic complications and poor attitude of physician were found to be the most significant factors accounting for a poor QOL. Measures must be taken to address each of these issues to improve the QOL of diabetic subjects. Renal complications, sexual dysfunction, neuropathy, visual problems, foot ulcers and depression were the major complications and comorbidities accounting for considerable worsening of QOL in diabetic subjects. The right balance has to be struck for optimal control of blood glucose as aggressive control of blood glucose may lead to worsening of QOL. The first modality of treatment in early stages must therefore be diet control and exercise. Newer modalities of treatment like inhaled insulin and insulin pens will improve patient compliance and thus serve to improve QOL. Yoga is also found to be beneficial in many studies. Oral insulin and role of camel milk for better compliance in younger type 1 diabetic subjects may significantly improve their QOL.
Acknowledgment We are grateful to Dr P.R. Venketaraman, St Albert’s College, Ernakulam, Kerala, India for his valuable suggestions and criticisms which have helped us in improving this article.
Appendix-1 The ADDQOL questionnaire which has been found to be ideal for measuring the QOL of Asian diabetic subjects (see > Figure 121-4–121-6; > Table 121-4)
Quality of Life Among Diabetic Subjects: Indian Perspectives
121
. Figure 121‐4 (a) Summary of the 13 domain-specific items used and their response options. ‘‘If I did not have diabetes . . . would (be) . . ..’’ (b) Format of the condition-specific domains showing the scores assigned
2089
2090
121
Quality of Life Among Diabetic Subjects: Indian Perspectives
. Figure 121‐5 Two overview items; (a) In general, my present quality of life is, (b) If I did not have diabetes, my quality of life would be: (With kind permission from Springer Science + Business Media: Bradley C, Todd C, Gorton T, Symonds E, Martin A, Plowright R (1991). Qual Life Res. 8: 79–81)
. Figure 121-6 ADDQOL: scoring specific domains (With kind permission from Springer Science + Business Media: Bradley C, Todd C, Gorton T, Symonds E, Martin A, Plowright R (1991). Qual Life Res. 8: 79–81)
Quality of Life Among Diabetic Subjects: Indian Perspectives
121
. Table 121-4 Ranges of responses showing the influence of importance ratings Impact Scores unweighted Domain
Min.
Importance ratings
Max.
Min.
Max.
Impact scores weighted by importance Means
Medians
Employment/career
3
0
0
3
1.012
0
Social life
3
0
0
3
1.014
0
Family relationships
3
0
0
3
0.931
0
Friends
3
0
0
3
0.35
0
Sex life
3
0
0
3
1.164
0
Sport/leisure
3
0
0
3
1.088
0
Travel
3
2
0
3
1.204
0
Future (own)
3
3
0
3
2.159
1
Future of family
3
2
0
3
1.401
0
Motivation
3
1
0
3
0.97
0
Physical activities
3
1
0
3
1.705
0
Others fussing
3
2
0
3
0.928
0
Enjoyment of food
3
1
0
3
2.309
2
With kind permission from Springer Science + Business Media: Bradley C, Todd C, Gorton T, Symonds E, Martin A, Plowright R (1991). Qual Life Res. 8: 79–81
References Aaron V. (2004). J Clin Endocrinol Metabol. 90: 4936–4945. Adachi H, Oyamada T. (2006). QLR J. A-68, Abstract #1294. Adepu R, Rasheed A, Nagavi BG. (2007). Indian J Pharm Sci. 69: 519–524. Agrawal RP, Aradhana, Hussain Sabir, Beniwal R, Sabir M, Kochar DK, Kothari RP. (2003). Int J Diab Dev Countries. 23: 130–134. Agrawal RP, Choudhary R, Sharma P, Sharma S, Beniwal R, Kaswan K, Kochar DK. (2007). Diabet Res Clin Pract. 77: 161–167. Alberti KGMM, Zimmet P, Shaw J. (2007). Diabet Med. 24: 451–463. doi: 10.1111/j.1464–5491.2007.02157. Alon Y. Avidan, Phyllis C. Zee (2006). Handbook of Sleep Medicine. Lippincott Williams and Wilkins, Philadelphia, PA. Baba CS, Alexander G, Kalyani B, Pandey R, Rastogi S, Pandey A, Choudhuri G. (2006). J Gastroenterol Hepatol. 21(1p1): 191–198. doi: 10.1111/j.1440– 1746.2005.04233.x.
Bakewell AB, Higgins RM, Edmunds ME. (2001). Nephrol Dial Transplant. 16: 1395–1401. Beydoun A, Kobetz SA, Carrazana EJ. (2004). Clin J Pain. 20: 174–178. Boulton AJM, Vileikyte L, Ragnarson-Tennvall G, Apelqvist J. (2005). Lancet. 366: 1719–1724. Brandle M, Zhou H, Smith BRK, Marriott D, Burke R, Tabaei BP, Brown MB, Herman WH. (2003). Diabetes Care. 26: 2300–2304. Ceglia L, Lau J, Pittas AG. (2006). Ann Intern Med. 145: 665–675. Chowdhury S. (2004). J Indian Med Assoc. 102: 554–556. Damodaran A, Malathi A, Patil N, Suryavansihi N, Marathe S. (2002). J Assoc Phys India. 50: 633. Das AK, Joshi SR. (2005). J Assoc Phys India. 53: 1. De Berardis G, Franciosi M, Belfiglio M, Di Nardo B, Greenfield S, Kaplan SH. (2002). Diabetes Care. 25: 284–291. de Visser CL, Bibo HJ, Groenier KH, de Visser W, Jong Meyboom-de B. (2002). Qual Life Res. 11: 249–261.
2091
2092
121
Quality of Life Among Diabetic Subjects: Indian Perspectives
Dowse GK, Gareeboo H, Zimmet PZ, Alberti KG, Tuomilehto J, Fareed D, Brissonnette LG, Finch CF. (1990). Diabetes. 39: 390–396. Epple C, Wright AL, Joish VN, Bauer M. (2003). Diabetes Care. 26: 2829–2834. Erdemoglu AK, Varlibas A. (2006). Neurol India. 54: 173–177. Ganesan VS, Balaji V, Seshaiah V. (1994). Int J Diab Dev Countries. 14: 119–128. Gayathri B. (2001). J Indian Acad Clin Med. 2: 270–275. Greenfeld S, Kaplan SH, Kahn R, Ninomiya J, Griffith JL. (2002). Ann Intern Med. 136: 111–121. Guleria S, Aggarwal S, Bansal VK, Varma MC, Kashyap L, Tandon N, Mahajan S, Misra MC. (2005). Natl Med J India. 18: 18–19. Gutierrez-Alvarez AM, Beltra´n-Rodrı´guez J, Moreno CB. (2007). J Pain Symptom Manag. 34: 201–208. Hoerger M, Brandle K, Hicks S, Sorensen P, Zhang RF, Hamman RT, Ackermann MM, Engelgau Ratner RE. (2005). Ann Intern Med. 142: 323–332. Hosler AS, Hinman TM, Juster HR. (2007). Diabetes Care. 30: 1883–1885. doi: 10.2337/dc07–0047. Hummer RA, Rogers RG, Nam CB, Leclere FB. (1999). Soc Sci Q. 80: 136–153. Hwee-Lin Wee, Yin-Bun Cheung, Shu-Chuen Li1, Kok-Yong Fong, Julian Thumboo. (2005). Health QualLifeOutcome.3: 2. doi:10.1186/1477–7525–3–2. HyeSook Park, YoungSun Hong, HyeJin Lee, EunHee Ha, YeonAh Sung. (2004). J Clin Epidemiol. 57: 978–984. Hypertension study group. (2001). Bull World Health Organ. 79: 490–500. Innes KE, Cheryl Bourguignon RN, Taylor AG. (2005). J Am Board Fam Pract. 18: 491–519. Jeffcoate WJ, Harding KG. (2003). Lancet. 361: 1545–1551. Joshi SR. (2003). J Assoc Physicians Ind. 51: 445–446. Kapur A. (2000). Cost of Diabetes in India – The CODI Study Paper Presented at the Novo Nordisk Diabetes Update, Bangalore, India. Kekre NS. (2006). Indian J Urol. 22: 9. King H, Aubert RE, Herman WH. (1998). Diabetes Care. 21: 1414–1431. Kochar DK, Rawat N, Agrawal RP, Vyas A, Beniwal R, Kochar SK, Garg P. (2004). Q J Med. 97: 33–38. Korytkowski M, Niskanen L, Asakura T. (2006). Clin Ther. 27(Suppl 2): S89–S100. Kumar S. (2004). J Indian Psychol. 22: 63–72. Kutty VR, Soman CR, Joseph A, Pisharody R, Vijayakumar K. (2000). Natl Med J India. 13: 287–292. Lau C, Qureshi AK, Scott SG. (2004). J Postgrad Med. 50: 189–194. Lloyd A, Sawyer W, Hopkinson P. (2001). Value Health. 4: 392–400.
Lori MB. (2003). Diabetes Care. 26: 3067–3073. Ma S, Cutter J, Tan CE, Chew SK, Tai ES. (2003). Am J Epidemiol. 158: 543–552. Mamidipudi PR, Vasavada AR, Merchant SV, Namboodiri VN, Ravilla TD. (2003). J Cataract Refract Surg. 29: 1143–1151. Mensing C, Boucher J, Cypress M, Weinger K, Mulcahy KB, Patricia, Hosey G, Kopher W, Lasichak A, Lamb B, Mangan M, Norman J, Tanja J, Yauk L, Wisdom K, Adams C. (2002). Diabetes Care. 25: S140–S147. doi: 10.2337/diacare.25.2007.S140. Mohan V, Madan Z, Jha R, Deepa R, Pradeepa R. (2004). Int J Diab Dev Countries, 24: 29–35. Mohan D, Raj D, Shanthirani CS, Datta M, Unwin NC, Kapur A, Mohan V. (2005). J Assoc Phys India. 53: 283–287. Mohan V, Shanthirani CS, Deepa M, Deepa R, Unnikrishnan RI, Datta M. (2006). J Assoc Physicians Ind. 54: 113–117. Mudaliar S. (2007). Indian J Med Res. 125: 275–296. Ng T, CheeChong L, AiZhen J, Shinfuku N. (2005). Qual Life Res. 14: 1755–1768. Nirmalan PK, Tielsch JM, Katz J, Thulasiraj RD, Krishnadas R, Ramakrishnan R, Robin AL. (2005). Invest Ophthalmol Vis Sci. 46: 2308–2312. Nutheti R, Shamanna BR, Nirmalan PK, Keeffe JE, Krishnaiah S, Rao GN, Thomas R. (2006). Invest Ophthalmol Vis Sci. 47: 4742–4748. Palaian S, Acharya LD, Rao PGM, Ravi Shankar P, Nair NM, Nair NP. (2006). P&T J. 31: 383–389. Piette JD, Richardson C, Valenstein M. (2004). Am J Manag Care. 10(Pt 2): 152–162. Pitale S, Kernan-Schroeder D, Emanuele N, Sawin C, Sacks J, Abraira C. (2005). J Diabetes Complicat. 19: 207–211. Prabhakaran D, Chaturvedi V, Ramakrishnan L, Jeemon P, Shah P, Snehi U, Reddy KS. (2007). Natl Med J India. 20: 4–10. Radha V, Vimaleswaran KS, Deepa R, Mohan V. (2003). Indian J Med Res. 117: 225–238. Raghupathy P, Antonisamy B, Fall CHD, Geethanjali FS, Leary SD, Saperia J, Priya G, Rajaratnam A, Richard J. (2007). Diabetes Res Clin PR. 77: 269–279. Ramachandran A, Snehalatha C, Kapur A, et al. (2001). Diabetologia. 44: 1094–1101. Ramachandran A, Snehalatha C, Mary S, Mukesh B, Bhaskar A, Vijay V. (2006). Diabetologia. 49: 289–297. Ramachandran A, Snehalatha C, Vijay V, King H. (2002). Diabet Med. 19: 130–135. Rayappa PH, Raju KNM, Kapur A, Bjork S, Sylvist C, Dilip Kumar KM. (1999). Int J Diab Dev Countries. 19: 7–16.
Quality of Life Among Diabetic Subjects: Indian Perspectives Rose M, Fliege H, Hildebrandt M, Schirop T, Klapp BF. (2002). Diabetes Care. 25: 35–42. doi: 10.2337/diacare.25.1.35. Rubin RR. (2000). Diab Spect. 13: 21. Rubin RR, Peyrot M. (1999). Diab Care. 22(3): 448–452. Sarah Wild, Gojka Roglic, Anders Green, Richard Sicree, Hilary King. (2004). Diabetes Care 27: 1047–1053. Sekar V, Sundaram A, Lakshmi B, Kalaivani Ajantha R, Mala S, Banupriya M, Maheswari U. (2006). Int J Diab Dev Countries. 26: 16–18. Sharma S, Brown GC, Brown MM, Hollands H, Shah GK. (2000). Curr Opin Ophthalmol. 11: 175–179. Sharma S, Hollands H, Brown GC, Brown MM, Shah GK. (2001). Curr Opin Ophthalmol. 12: 230–234. Sharma V. (2002).World Leisure J. 44: 16–23. Shobhana R, Rama Rao P, Lavanya A, Williams R, Padma C, Vijay V, Ramachandran A. (2002). Diabetes Res Clin Pract. 55: 45–48. Shobhana R, Rama Rao P, Lavanya A, Williams R, Vijay V, Ramachandran A. (2000). Diabetes Res Clin Pract. 48: 37–42. Sidhu SS, Nundy S, Tandon RK. (2001). Am J Gastroenterol. 96: 107–111. doi: 10.1111/j.1572– 0241.2001.03359.x. Singh H, Bradley C. (2006). Int J Diab Dev Countries. 26: 7–10. Singh JC, Devasia A, Gnanaraj L, Chacko KN. (2005). Natl Med J India. 18: 139–143. Singh N, Armstrong DG, Lipsky BA. (2005). J Am Med Assoc. 293: 217–228. Singh P, Bhandari M. (2004). Transplant Proc. 36: 2013–2014. Sprangers MA, de Regt EB, Andries F, van Agt HM, Bijl RV, de Boer JB, Foets M, Hoeymans N, Jacobs AE, Kempen GI, Miedema HS, Tijhuis MA, de Haes HC. (2000). J Clin Epidemiol. 53: 895–907. Sridhar GR, Madhu K. (1994). Diabetes Res Clin Pract. 23: 183–186. Sridhar GR, Madhu K, Veena S, Madhavi R, Sangeetha BS, Rani A. (2007). Diabetes Metab Syndr Clin Res Rev. 1: 181–187.
121
Sridhar GR, Venkata P. (2006). Diabetes Res Clin Pract. 72: 209–211. Sudhir PM, Kumaraiah V, Munichoopadda C. (2003). J Clin Psychol. 59: 715–722. Tan SMK, Shafiee Z, Wu LL, Rizal AM, Rey JM. (2005). Int J Psychiatr Med. 35: 123–136. Thakurdesai PA, Kole PL, Pareek RP. (2004). Diabetes Mellitus. 53: 309–313. Thumboo J, Fong KY, Machin D, Chan SP, Soh CH, Leong KH, Feng PH, Thio ST, Boey ML. (2003). Soc Sci Med. 56: 1761–1772. Translating Research into Practice. Fact Sheet (2000). Publication No. 00-P018. AHRQ Research on Diabetes Care. Available at: http://www.ahrq.gov/ research/tripdiab.htm. Varghese RT, Salini R, Abraham P, Reeshma KK, Vijayakumar K. (2007). Diabetes Metab Syndr Clin Res Rev. 1: 173–179. doi: 10.1016/j.dsx.2007.05.005. Vijayakumar K, Sivan YS, Das RR, Kutty VR. (1994). Health Policy Plan. 9: 331–336. Vohra P. (2006). Dissertation. Department of Psychology, Central Connecticut State University, New Britain, CT. Wagner TH, Hu TW, Bentkover JD, LeBlanc K, Stewart W, Corey R, et al. (2002). Am J Manag Care. 8: S598–S607. Wee HL, Li SC, Cheung YB, Fong KY, Thumboo J. (2006a). J Diabetes Complicat. 20: 170–178. Wee H-L, Tan C-E, Goh S-Y, Li S-C. (2006b). Pharmacoeconomics. 24: 673–682. Wei M, Gibbons LW, Mitchell TL, Kampert JB, Blair SN. (2000). Diabetes Care. 23: 18–22. WHO Expert Consultation. (2004). Lancet. 363: 157–163. Wilson E, Wardle EV, Chandel P, Walford S. (1993). Diabet Med. 10: 177–180. World Bank. World development report 1993: investing in health. World development indicators. Oxford. Oxford University Press, 1993; 213–214. Zelman DC, Brandenburg NA, Gore M. (2006). Clin J Pain. 22: 681–685.
2093
122 Healthy Lifestyle Habits and Health-Related Quality of Life in Diabetes C. Li . E. S. Ford 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2096
2 2.1 2.2 2.3 2.4 2.5
Correlates (Predictors or Determinants) of HRQOL in Diabetes . . . . . . . . . . . . . . . . 2097 Demographic and Socio-Economic Correlates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2097 Duration of Diabetes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2100 Diabetic Complications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2101 Insulin Therapy in Type 2 Diabetes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2101 Lifestyle, Behavioral, and Psychological Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2102
3 3.1 3.2 3.3 3.4 3.5
An Illustration of Original Results on HLHs and HRQOL in Diabetes . . . . . . . . . 2102 HRQOL Measures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2102 Healthy Lifestyle Habits (HLHs) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2103 Diabetes and Diabetes-Related Complications and Chronic Conditions . . . . . . . . . . 2103 Statistical Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2103 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2104
4
Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2110
5
Application of the Findings to Other Diseases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2112
6
Key Facts of Healthy Lifestyle Habits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2112 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2112 Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2113
Disclaimer: The findings and conclusions in this report are those of the author(s) and do not necessarily represent the official position of the Centers for Disease Control and Prevention. #
Springer Science+Business Media LLC 2010 (USA)
2096
122
Healthy Lifestyle Habits and Health-Related Quality of Life in Diabetes
Abstract: The prevalence of diabetes has continued to rise in the Unites States. Diabetes has significant impact on > health-related quality of life (HRQOL). Thus, improving HRQOL has been an important aspect of health care management in the diabetic population. In this chapter, we systematically reviewed 20 studies on the correlates of HRQOL in diabetes published between 1998 and 2008. Diabetes-related complications, older age, female sex, black or Native American race/ethnicity, longer > duration of diabetes, insulin therapy, obesity, smoking, and physical inactivity are associated with the impairment of HRQOL in diabetes. Micro- and macro-vascular complications appear to be the strongest correlates of HRQOL. Therefore, intervention strategies aimed at preventing or delaying the occurrence of these complications may lead to improvement of HRQOL. > Healthy lifestyle habits have been associated with improvement of HRQOL. Because people with diabetes are more likely to be non-smokers and to consume more fruits and vegetables but less likely to reach the recommended level of physical activity than those without diabetes, efforts are needed to promote the adoption of healthy lifestyle habits in order to improve HRQOL in diabetes. As an illustration, we provide updated prevalence estimates of impaired HRQOL among people with diabetes and demonstrate how the healthy lifestyle habits are associated with HRQOL using a large population-based sample. List of Abbreviations: BDI, Beck depression inventory; BMI, body mass index; BRFSS, risk factor surveillance system; CDC, Centers for Disease Control and Prevention; CI, confidence interval; CVD, cardiovascular disease; DM, > diabetes mellitus; DSQOLS, diabetes specific qualify of life scale; D-QOL, diabetes-quality of life; EQ-5D, euroQOL-5 dimensions; EQ-VAS, euroQOL-visual analogue scale; EuroQOL, European quality of life scale; FVC, fruit and vegetable consumption; HLHs, healthy lifestyle habits; HRQOL, healthrelated quality of life; HUI-3, health utility index; IDDM, insulin dependent diabetes mellitus; LTPA, leisure-time physical activity; NIDDM, non-insulin dependent diabetes mellitus; NSMK, not smoking; PR, prevalence ratio; QOL, quality of life; QWB-SA, quality of wellbeing index-self administered version; SE, standard error; SF-20, short form-20; SF-36, short form-36; SUDAAN, survey data analysis; SWED-QUAL, Swedish health-related quality of life scale; T1DM, type 1 diabetes mellitus; T2DM, type 2 diabetes mellitus; U.K., United Kingdom; U.S., United States; WHO, World Health Organization > behavioral
1
Introduction
The prevalence of diabetes mellitus has continued to rise in the United States (Mokdad et al., 2003). Diabetes affected approximately 20.6 million or 9.6% of the U.S. adults aged ≥20 years in 2005 (Centers for Disease Control and Prevention, 2005). People with diabetes mellitus have an increased risk of developing coronary heart disease and stroke and dying from these comorbidities (Haffner et al., 1998; Idris et al., 2006). Health-related quality of life (HRQOL) is one of the key indicators for national health in Healthy People 2010 (US Department of Health and Human Services, 2000), and it is a useful tool for assessing people’s perceived health and health burden (Testa and Simonson, 1996). Because of the chronic nature of diabetes, the goal of medical treatment and lifestyle modifications is not only to prolong life but also to maintain a high level of HRQOL. There is evidence showing that diabetes has significant adverse effects on HRQOL (Norris, 2005; Rubin and Peyrot, 1999). Thus, improving HRQOL is a critical component of clinical management and public health services for people with diabetes.
Healthy Lifestyle Habits and Health-Related Quality of Life in Diabetes
122
The goals of this chapter are as follows: (1) to conduct an overview of the correlates (predictors or determinants) of HRQOL in diabetes published between 1998 and 2008 through a systematic literature search of PubMed; (2) to illustrate the impairment of HRQOL and to provide an updated prevalence estimate of impaired HRQOL among people with diabetes; and (3) to demonstrate the association of healthy lifestyle habits (HLHs) with HRQOL in diabetes. Original data from the 2007 Behavioral Risk Factor Surveillance System (BRFSS), a large population-based sample in the United States, were analyzed for the illustration.
2
Correlates (Predictors or Determinants) of HRQOL in Diabetes
Rubin and Peyrot (1999) reviewed empirical studies published up to 1998 in English language using the keywords ‘‘quality of life’’ and ‘‘diabetes’’ through a systematic literature search of Medline and PsychLit and indicated that > diabetic complications were the most important factors of QOL. The studies reviewed in this paper represented a wide range of research on quality of life and diabetes because they covered both descriptive studies and randomized clinical trials. In addition, Norris (2005) reviewed some large-scale studies published up to 2003 on HRQOL and diabetes with a focus on the effects of medical treatments and educational and behavioral interventions on HRQOL. In recent years, studies on the association between HRQOL and diabetes have been growing. To review studies during the most recent 10 years, we identified studies published between 1998 and 2008 with a focus on correlates or predictors or determinants of HRQOL in diabetes. We conducted a literature search in PubMed using the following keywords in the title: (‘‘diabetes’’ or ‘‘diabetic’’ or ‘‘diabetics’’ or ‘‘NIDDM’’ or ‘‘IDDM’’) and (‘‘health-related quality of life’’ or ‘‘quality of life’’ or ‘‘health’’) and (‘‘correlates’’ or ‘‘predictors’’ or ‘‘determinants’’ or ‘‘factors’’) and found 20 empirical studies (> Table 122-1) (Camacho et al., 2002; Chyun et al., 2006; Coffey et al., 2002; Ghanbari et al., 2005; Hanninen et al., 1998; Hart et al., 2005; Holmes et al., 2000; Huang and Hung, 2007; Li et al., 2007; Lloyd et al., 2001; Maddigan et al., 2006; Misra and Lager, 2008; Nicolucci et al., 2008; Papadopoulos et al., 2007; Rejeski et al., 2006; Shiu et al., 2008; Tang et al., 2006; UK Prospective Diabetes Study Group, 1999; Valensi et al., 2005; Wexler et al., 2006). These studies represent a full range of research on the correlates (predictors or determinants) of HRQOL in diabetes published in English language worldwide in the past decade.
2.1
Demographic and Socio-Economic Correlates
Age. The majority of studies reviewed have shown that older age was associated with decreased physical and social functioning, whereas younger age was associated with decreased mental components of the HRQOL as measured by short form-36 (SF-36), health utility index-3 (HUI-3), or diabetes-quality of life (D-QOL). Thus, efforts in improving the HRQOL for people with diabetes need to take age into account. Sex. The studies reviewed have consistently shown that females have a lower level of HRQOL than males in all domains as measured by SF-36. However, one study (Chyun et al., 2006) showed that males had only significantly higher scores than females in the domains of physical functioning and general health after controlling for all other
2097
2098
122
Healthy Lifestyle Habits and Health-Related Quality of Life in Diabetes
. Table 122-1 Summary of studies on the correlates or predictors of health-related quality of life in diabetes Researcher (year)
Site
Target population
Hanninen et al. Finland (1998)
T2DM, aged diabetic neuropathy, > diabetic nephropathy, renal failure, nephritic syndrome, renal transplant, dialysis, microalbuminuria, renal insufficiency, amputation, foot ulcer, gastroparesis, diabetic diarrhea, impotence, postural hypotension, and vitreous hemorrhage) and > macrovascular complications (e.g., angina pectoris, myocardial infarction, stroke, revascularization procedures, transient ischemic attacks, intermittent claudication, and peripheral vascular disease). In most of the studies reviewed, the major types of microvascular complication such as retinopathy, neuropathy, nephropathy, and foot and leg ulcer and major types of macrovascular complications such as myocardial infarction, heart failure, and stroke were based on a patient’s selfreport. Only in two studies were medical records used to define a comprehensive list of complications (Nicolucci et al., 2008; Wexler et al., 2006). Nicolucci et al. (2008) indicated that microvascular and macrovascular complications strongly correlated with all domains of the SF-36 QOL instrument, particularly with physical components. Wexler et al. (2006) found that symptomatic conditions such as microvascular complications and heart failure were more likely to be significantly associated with decreased HRQOL than unsymptomatic conditions after controlling for other diabetes-related conditions. In addition, according to the results of a study in the United Kingdom, macrovascular complications were associated with decreased mobility, usual activities, general health, and vigor (UK Prospective Diabetes Study Group, 1999). In a longitudinal study, high blood pressure, nephropathy, and intermittent claudication had the most negative effects on changes in HRQOL (Hart et al., 2005). However, one study did not detect significant associations between diabetic complications and HRQOL (Hanninen et al., 1998). As noticed by the authors, the younger age of the study population ( Table 122-2). In particular, people who had a BMI ≥30 kg/m2, who had type 2 diabetes and used insulin, who had a foot sore, who had retinopathy, and who had 2, 3, or ≥4 chronic conditions had a higher prevalence of all four impaired HRQOL indicators compared to their counterparts (all P values PSA, probabilistic sensitivity analysis; > QALY, quality-adjusted life year; QOL, quality of life; RCT, randomized controlled trial; > SF-36, short form-36; TCS, topical corticosteroids; > VAS, visual analog scale; > WTP, willingness-to-pay
1
Introduction – Atopic Dermatitis and Health Related Quality of Life (HRQOL)
The burden of atopic dermatitis (AD) is quite significant on the patient, the patient’s family and society. The causes of this burden are multidimensional and may include: the chronic nature of the disease, its effect on physical appearance, the disruption of physical and psychosocial development for children, the impact on lifestyle and economic factors (> Table 126‐1). Patients with AD have a lower quality of life (QOL) than that of the general population due to decreased social functioning, worse mental health, increased anxiety and stress, increased incidence of psychiatric disorders, behavioral problems and decreased quantity and quality of
Quality of Life and Costs in Atopic Dermatitis
126
. Table 126-1 Features of atopic dermatitis (Williams et al., 2000) Diagnosis of AD includes the presence of an itchy skin condition in past 12 months plus At least 3 criteria Onset before 2 years of age History of flexural involvement History of dry skin Personal history of other atopic diseases Visible flexural dermatitis
sleep (Carroll et al., 2005; Chamlin, 2006). Families and caregivers of patients with AD are also affected by the disease and report lower quality of life as well. Not only do parents report more stress, but familial relationships are negatively impacted as a result of the extra time and effort that goes into caring for the patient with AD (Carroll et al., 2005; Chamlin, 2006). In addition, sleep loss for the patient extends to the family and can be a factor leading to increased stress. This increased stress leads to poor functioning and coping skills both at work and at home. Parents may actually need to take time away from work to care for their child with AD, which can translate into financial burden (Carroll et al., 2005). Costs associated with caring for the child with AD only adds to this burden. Direct and > indirect costs for AD have been measured in various countries and are substantial from both a patient and a societal perspective. The direct costs have been reported to range from US $71 to 2,559 per patient per year (Carroll et al., 2005). This variation in cost is due to differences in study methodology as well as differences in health care systems of the various countries. The majority of the costs of AD consists of indirect costs associated with time lost from work, lifestyle changes and non-traditional or over-the-counter treatments for AD (Carroll et al., 2005). The financial burden on the health care system and on society is expected to grow because the prevalence of the disease is increasing.
2
HRQOL Instruments Used in AD
Given the considerable impact that AD has on the patient, the caregiver and the family, the need for health-related quality of life (HRQOL) measures and QOL studies becomes imperative to assess the burden of AD, identify areas of need in disease management and monitor outcomes of treatment. Currently, assessment of QOL in standard clinical practice relies heavily on the clinician’s overall impression rather than on results of a QOL instrument. Although the use of these instruments is not often adopted in clinical practice, there are several validated questionnaires, both generic and disease-specific, that are available for use (> Table 126-2). The selection of which instrument to utilize depends on the subject of interest (patient, parent or family), the age of the patient (infant, child or adult), specific outcomes of interest, the availability of the instrument in various languages and the mode of administration (selfadministered, proxy-rated, interview-administered, telephone-administered).
2165
Dermatology-Specific Quality of Life (DSQL)
Dermatology-specific instruments
Short Form-36 (SF-36)
Generic instruments
35
36
Number of items
Not reported
0–100 (Higher score indicates better QOL)
Scoring
Self-perceptions
Work and school
Self-care activities
Social functioning
Psychological well-being
Symptoms
General health perceptions
Vitality
General mental health
Social functioning
Adults
Adults
General health Bodily pain
Adolescents
Applicable age group
Physical functioning
Measures
(Anderson and Rajagopalan, 1997)
(Ware and Sherbourne, 1992)
References
126
Name of questionnaire
. Table 126-2 HRQOL instruments used in atopic dermatitis
2166 Quality of Life and Costs in Atopic Dermatitis
10
10
Dermatology Life Quality Index (DLQI)
Children’s Dermatology Life Quality Index (CDLQI)
0–30 (Higher score indicates poorer QOL)
21–30 = extremely large effect on patient’s life
11–20 = very large effect on patient’s life
Treatment
Personal relationships
School or holidays
Leisure
Sleep
Symptoms and feelings
6–10 = moderate Treatment effect on patient’s life
2–5 = small effect Work and school on patient’s life Personal relationships
Leisure
0–30 Symptoms and 0–1 = no effect on feelings patient’s life Daily activities
5–16 years of age
>16 years of age
(Lewis-Jones and Finlay, 1995)
(Finlay and Khan, 1994)
Quality of Life and Costs in Atopic Dermatitis
126 2167
Disease severity: 1
Infant’s Dermatology Quality of Life Index (IDQOL)
Life quality index: 10
16
16, 29 or 61
Number of items
Social support
Illness cognition
Disease-related impact
Psychological functioning
Stressors
Physical functioning
Cognitive effects
Physical functioning
Social functioning
Emotions
Symptoms
Measures
0–30 Higher score Treatment indicates poorer QOL
Daily activities
Leisure
0–4 Higher score Symptoms and indicates feelings increased severity Sleep
Not reported
0–100 (Higher score indicates poorer QOL)
Scoring
16 years of age
Adults
Applicable age group
(Beattie and Lewis-Jones, 2006)
(Evers et al., 2008)
(Chren et al., 1996)
References
126
Impact of Chronic Skin Disease on Daily Life (ISDL)
Skindex
Name of questionnaire
. Table 126-2 (continued)
2168 Quality of Life and Costs in Atopic Dermatitis
10
Family Dermatology Life Quality Index (FDLQI)
45
28
Childhood Atopic Dermatitis Impact Scale (CADIS)
Parent’s Index of Quality of Life in Atopic Dermatitis (PIQOL-AD)
Disease-specific instruments
10
Dermatitis Family Impact Score (DFI)
Parental emotions
Parental sleep
Family/social interactions
Limitations on activity
Symptoms
0–28 Higher score Psychological indicates poorer impact QOL Physical functioning
0–180 Higher score indicates poorer QOL
Caretaking responsibilities
Personal relationships
Work or school
0–30 Higher score Emotional health indicates poorer Physical health QOL Leisure
Daily activities
Personal relationships
Leisure
Sleep
0–30 Higher score Emotional health indicates poorer Physical health QOL Vitality
Parents of children 16 years of age (family members or partners of patients)
Adults
(McKenna et al., 2005; Meads et al., 2005)
(Chamlin et al., 2007)
(Basra et al., 2007)
(Beattie and Lewis-Jones, 2006)
Quality of Life and Costs in Atopic Dermatitis
126 2169
Parents of Children with Atopic Dermatitis (PQOL-AD)
Name of questionnaire 26
Number of items Higher score indicates better QOL
Scoring
Treatment
Effects on social life
Acceptance of disease
Emotions
Psychosomatic well-being
Measures
Applicable age group
(Staab et al., 2005)
References
126
. Table 126-2 (continued)
2170 Quality of Life and Costs in Atopic Dermatitis
Quality of Life and Costs in Atopic Dermatitis
3
126
Impact of AD on Patient HRQOL
Several recent studies have demonstrated that AD has a profound impact on the QOL of both children and adults. It has been estimated that more than 25% of children with AD may endure chronic eczema associated with moderate-to-severe HRQOL impairment (Emerson et al., 2000). There appears to be a positive correlation between severity of AD and QOL in children (Ben-Gashir et al., 2004; Fivenson et al., 2002). Among children, AD affects activities of daily living such as feeding during mealtimes, getting dressed, bathing and playing (Mozaffari et al., 2007). Children also suffer greater problems associated with treatment, pain and itchiness and it is the increase in itchiness that leads to sleep loss and daytime drowsiness (Mozaffari et al., 2007). Oral antihistamines, sometimes used to treat itch associated with AD in children, only exacerbates daytime drowsiness. Disruption in sleep pattern is a widespread problem among children with AD, with more than 60% of children experiencing sleep issues (Lewis-Jones, 2006). Sleep disturbances present as difficulty falling asleep, frequent night-time awakenings, reduced total sleep and difficulty waking (Lewis-Jones, 2006). This may result in poor performance at school, behavior and discipline problems and poor social functioning. The child with AD may find that school and social life are tremendously affected such that participation in sports is limited, absence from school may increase and concentration may be impaired due to sleep loss (Lewis-Jones, 2006). AD also affects physical appearance and a child may experience embarrassment, low self-esteem, social isolation, mood changes and depression as a result of physical stigmatizations (Lewis-Jones, 2006; Lewis-Jones and Finlay, 1995). Adults with AD experience similar negative effects in QOL as that of children. Namely, activities of daily living such as getting dressed were reported as more difficult (Mozaffari et al., 2007). As with children, there also appears to be a positive correlation between disease severity and QOL in patients with AD (Coghi et al., 2007; Fivenson et al., 2002). Adults with AD report excessive itchiness, embarrassment, and treatment problems compared to the general population (Fivenson et al., 2002). In addition, studies utilizing the SF-36 questionnaire, a generic HRQOL instrument, report that AD patients have significantly lower QOL scores in the areas of vitality, social functioning and mental health (Fivenson et al., 2002; Kiebert et al., 2002). One study also found that adults with AD experience higher levels of anxiety, which can potentially serve as a trigger of the itch-scratch cycle, thereby aggravating or flaring up the patient’s AD (Carroll et al., 2005) (> Table 126-3). The psychosocial burden is particularly problematic for adult patients. Many adults have been suffering from AD since childhood and this, in combination with the physical stigmatization of the disease, lead to a lack of ability to cope with their condition, lack of adherence to treatment, and self-esteem and self-image issues (Wittkowski et al., 2004). In fact, it is thought that AD-related perceptions of stigma are stronger predictors of QOL than more general psychological factors like depression (Wittkowski et al., 2004).
4
Impact of AD on Caregiver and Family HRQOL
The burden of AD reaches beyond the patient and extends to caretakers and parents of children with AD, families of patients with AD and partners of adults with AD. Parents of children with AD undergo higher levels of stress and increased caregiver burden than parents of healthy children (Carroll et al., 2005). Furthermore, parents or caregivers with a child undergoing disease flares report lower QOL versus those without disease flares (Arnold et al., 2007).
2171
2172
126
Quality of Life and Costs in Atopic Dermatitis
. Table 126-3 Results of QOL studies in children and adults with AD References
Study subjects
Mozaffari 86 cases et al. (2007) with AD >4 years old 98 controls without AD >4 years old
QOL instrument used
QOL score
CDLQI for 4–16-year CLDQI: olds Cases: 16.5 DLQI for >16-year olds
Controls: 0.44 DLQI:
Conclusion Children and adults with AD have poorer QOL than those without AD
Cases: 20.5 Controls: 1.15
Fivenson 298 patients CDLQI for 4–16-year CDLQI: et al. (2002) with AD olds Patients with AD: 5.8 DLQI for >16-year Patients without AD: 0.38 olds DLQI:
SF-36 for adults
Children and adults with AD have poorer QOL than those without AD
Patients with AD: 6.6 Patients without AD: 0.5 SF-36: Statistically significant differences for the following domains: vitality, social functioning, mental health
Compounding this stress is the lack of sleep that parents experience as a result of caring for their child in distress (Lewis-Jones, 2005). Therefore, it is not unexpected to hear that parents feel helpless, frustrated, hopeless, anxious, depressed, angry and guilty as a result of their child’s symptoms (Carroll et al., 2005; Chamlin et al., 2004; Lawson et al., 1998). The child with AD requires more resources, both time-wise and financially. Su et al. reports that parents spend about 2–3 h per day caring for the child with AD and Arnold et al. found that patients with AD consume nearly two additional unscheduled visits per year (Arnold et al., 2007; Su et al., 1997). This translates into time away from other family members as well as time away from work. AD affects the entire family and it is evident that a patient with AD may lower the QOL of the family. Indeed, families of children with moderate to severe eczema have equal or significantly greater impact on HRQOL than families of diabetic or asthmatic children (Su et al., 1997). Not only does a positive correlation exist between QOL and AD severity in patients, but this relationship holds true for family life as well (Ben-Gashir et al., 2002). Several aspects of family life are affected, such as sleep, housework, emotions and finances (Ben-Gashir et al., 2002; Lawson et al., 1998; Ricci et al., 2007). Disturbed sleep has been reported to be the greatest burden on family members and often leads to irritability due to increased tiredness (Ricci et al., 2007). A study by Lawson et al. revealed that 38% of siblings of children with AD had problems with sleep (Lawson et al., 1998). Interestingly, in the case of a partner of an adult with AD, sleep quality was not affected (Misery et al., 2007). Furthermore, the QOL of partners did not seem to be as affected as seen with family life. However, the impact on sexual desire was significant for the partner as well as for the patient (Misery et al., 2007).
Quality of Life and Costs in Atopic Dermatitis
5
126
Effect of Treatment on HRQOL in AD
While AD has been shown to negatively impact HRQOL of both patients and their caregivers, there is evidence that effective treatment can significantly improve HRQOL. A number of clinical trials involving immunomodulators have shown a positive impact in pediatric and adult populations. In a study by Whalley et al. conducted on 403 children between the ages of 2 and 17, the benefits of pimecrolimus on parents’ QOL were realized based on PIQOL-AD scores (Whalley et al., 2002). These findings were supported and further elaborated upon in a 2005 study of 196 children aged 3–23 months by Staab et al. using the more detailed PQOL-AD survey (Staab et al., 2005). All five subcategories (psychosomatic well-being, effects on social life, confidence in medical treatment, emotional coping, and acceptance of disease) showed statistically significant increases in parents’ QOL. However, 12 weeks was not a sufficient amount of time to allow for the measurement of the full extent of improvements. Another immunomodulator, tacrolimus, was also studied by various groups, with the focus squarely on the effects on patient QOL. Kawashima et al. used both DLQI and World Health Organization QOL-26, a generic QOL survey, which showed statistically significant improvements in QOL from baseline 12 to weeks later (Kawashima, 2006). Results from a study by Drake et al. confirm these findings. Using DLQI, CDLQI, and a toddler QOL survey to assess QOL in 579 adults, 178 children, and 145 toddlers, respectively, Drake showed that drug treatment was successful in improving the patients’ QOL (Drake et al., 2001). What makes Drake’s study different is that he uses three arms: vehicle only; 0.03% tacrolimus ointment and 0.1% tacrolimus ointment. Though both strengths of tacrolimus showed efficacy, there was no statistically significant difference between the two. It should be noted that the stated trials involve the use of immunomodulators, only one of several options in the treatment of AD; other treatment modalities that are available but are not included in these trials are corticosteroids, methotrexate, cyclosporine and phototherapy. Despite the exclusion of these other treatment options, the immunomodulators alone show that treatments are undoubtedly available that will positively impact the patient’s and/or caregiver’s QOL.
6
Cost Burden and Cost-Effectiveness Analyses Associated with Atopic Dermatitis
In addition to its considerable impact on patient QOL, the monetary burden of AD is substantial (Abramovits et al., 2005; Delea et al., 2007; Ehlken et al., 2005; Ellis et al., 2002; Emerson et al., 2001; Fivenson et al., 2002; Green et al., 2004; Green et al., 2005; Jenner et al., 2004; Kemp, 2003; Verboom et al., 2002). In 2002, Ellis and colleagues estimated the thirdparty payer cost-of-illness for AD in the US to be in a range of US $0.9–3.8 billion for persons younger than 65 years old from 1997 to 1998 records (Ellis et al., 2002). In the U.K., in 2001, the cost of AD was estimated at £47 million. In Australia, the ISAAC study estimated the total cost at US $A316.7 million per year (Kemp, 2003). Indeed, studies in the past 7 years have demonstrated direct costs ranging from US $150 (Ehlken et al., 2005) (using approximate US$ equivalent in 2005) to US $580 (Ellis et al., 2002) per patient per year, with differences varying due to different cost-accounting methods. > Table 126-4 lists numerous references in which US$ (or equivalent) per patient burden of illness was calculated.
2173
2174
126
Quality of Life and Costs in Atopic Dermatitis
. Table 126-4 Selected AD burden-of-illness studies References
Year
Direct (in US$)
Indirect (in US$)
Ehlken et al. (2005)
2005
150b
1,589
Ellis et al. (2002)
2002
580
Not measured
Ellis et al. (2002)
2002
1,250
Not measured
Perspective (Payer) Societal
1,739
Private insurer Medicaid
Fivenson et al. (2002)
2002
167
147
Emerson et al. (2001)
2001
732
42
Jenner et al. (2004)
2004
2,812
Patient
Ricci et al. (2006)
2006
1,540
Patient
Verboom et al. (2002)
2002
71
Totala (in US$)
Health plan
609
Societal
115
Country
a
If both direct and indirect costs are calculated b US$ equivalent for 2005 calculated using www.gocurrency.com historic EU to US$ converter
Typically, outpatient visits and medications comprised the majority of direct costs,(Fivenson et al., 2002; Verboom et al., 2002) ranging from approximately 62–>90%.(Fivenson et al., 2002) In those studies that examined indirect costs (e.g., the patient out-of-pocket costs for co-pays, medications, household items, loss of productivity) in addition to direct costs, indirect costs comprised substantial percentages of the total, e.g., 36% (Emerson et al., 2001) or 73% (Fivenson et al., 2002). Several studies showed increasing costs with worsening disease severity in adults. Fivenson, Arnold and colleagues > Table 126-5 reported an average annual per patient direct cost ranging from US $435 in mild patients to US $3229 in severe patients. Indirect costs also increased by worsening disease severity – by more than two-fold to three-fold (Jenner et al., 2004) to as much as almost ten-fold (Fivenson et al., 2002). Similarly, Ehlken and co-authors showed a greater than 2-fold increase in total (both direct and indirect) costs for patients with mild versus severe disease.
7
Therapy-Specific Cost and Cost-Effectiveness
There have been several studies comparing cost and cost-effectiveness of different uses of topical corticosteroids (TCS), of TCS versus topical immunomodulators (i.e., pimecrolimus and tacrolimus) and of the topical immunomodulators against each other. Some of these are detailed below.
7.1
Topical Corticosteroids
Green and colleagues undertook a systematic review of 10 randomized controlled trials (RCTs) and one review article of the comparative cost-effectiveness of once- versus more frequent use of same potency corticosteroids in patients with AD (Green et al., 2004; Green et al., 2005). Their literature search at the time revealed no published studies of this nature. Given the
Quality of Life and Costs in Atopic Dermatitis
126
similar effectiveness of once- versus more frequent use of same potency corticosteroids in this population, the cost-effectiveness analysis distilled down to a cost-minimization analysis, favoring the lowest cost alternative. The authors noted a wide variation in price and product availability, with the lowest price being generic hydrocortisone ((£0.60 (approximately US $1.09)) to the highest at that time being mometasone furoate (Elocon) of £4.88 (approximate US $8.80 equivalent). Six of the RCT studies favored the once-daily option as the lowest cost treatment and four favored a twice-daily option, with successful outcome being defined by overall response to treatment, relapse/flareup rate, adverse effects, compliance, tolerability, patient preference measures and impact on QOL. One of the twice-daily-favored studies achieved a greater benefit (number of successful treatment responders) at a greater cost. However, it was felt that this greater cost would still likely be very cost-effective, given the relatively low prices of TCS. The limitations noted in the review were that of potentially limited generalizability due to 80% of the RCTs referring to potent TCS in patients with moderate-to-severe disease, whereas the majority of patients with AD have mild disease, and lack of information on quantity of product usage.
7.2
Topical Immunomodulators
Clinical data show that topical immunomodulators are effective in AD, yet do not cause the significant adverse effects associated with TCS (Abramovits et al., 2005). Delea and colleagues (Delea et al., 2007) retrospectively compared 157 pimecrolimus patients with 157 tacrolimus patients previously receiving TCS in a large claims database of managed care patients in terms of resource utilization (concomitant medications) and AD-related follow-up costs. They used propensity matching to control for differences between the groups in baseline demographic and clinical characteristics and utilization of AD-related services prior to assessment of disease severity. Patients in the pimecrolimus group had fewer pharmacy claims for TCS (mean 1.37 vs. 2.04, P = 0.021); this occurred primarily in the high-potency TCS category. Fewer patients in the pimecrolimus group also received antistaphylococcal antibiotics during the follow-up period (16% vs. 27%, P = 0.014) and total AD-related costs during this time were lower in this group than in the tacrolimus group (mean US $263 vs. US $361, P = 0.012). Ellis and colleagues used a > Markov model (a time-dependent simulation of the various probabilities and associated costs of patients as they might experience periods of remission and recurrence, treatment and response) to compare the cost-effectiveness of tacrolimus versus high-potency TCS in adults with moderate-to-severe AD who had failed mid-potency topical steroids (Ellis et al., 2003). In the model, following initial use of high-potency TCS (HPTC) or tacrolimus, patients could be ‘‘distributed’’ during 2- and then 4-week intervals to one of the following ‘‘health states’’: secondary therapy (oral antibiotic with TCS) or disease controlled. Probabilities and costs of clinical outcomes, consisting of continuing with HPTC, disease controlled, or continuing tacrolimus therapy (with or without secondary therapy) were derived from published literature and Phase III clinical trials. Drug cost was derived from published average wholesale price (AWP) and cost of physician visits as the median physician charge for Current Procedural Terminology code 99213 for an established patient office visit. Effectiveness was defined as disease-controlled days (DCD), the number of days on which patients experienced a greater than 75% improvement in their illness. Resource use and management of AD were based on physician panel opinion. The analysis was conducted
2175
0
1,873
Emergency room
Medications
1,024
0
Medication copays
Household Items
Child Care
9,983
9,983
23,944
Days lost from work
Subtotal
TOTAL
Productivity
6,035
316
Medications
Subtotal
380
3,740
Visit copays
575
7,926
0
Practitioner visits
Indirect Costs
Subtotal
Phototherapy
24
4,947
Outpatient
Mild (N = 55)
435.35 ± 156.40
181.51 ± 120.62
181.51 ± 120.62
109.72 ± 35.34
0
18.62 ± 10.37
5.74 ± 0.99
68.00 ± 25.83
6.91 ± 0.59
10.45 ± 6.75
144.13 ± 23.97
0
0.44 ± 0.44
34.06 ± 9.86
0
89.95 ± 7.54
19.68 ± 19.68
Mean per patient $ ± SE
17,941
8,705
8,705
3,803
0
1,863
254
1,306
280
100
5,433
63
0
1,765
0
3,605
0
Total $
578.79 ± 131.27
280.82 ± 113.89
280.82 ± 113.89
122.70 ± 34.49
0
60.10 ± 29.52
8.21 ± 1.73
42.13 ± 13.52
9.03 ± 1.28
3.23 ± 2.29
175.27 ± 25.84
2.03 ± 1.44
0
56.95 ± 14.14
0
116.29 ± 15.70
0
Mean per patient $ ± SE
Moderate (N = 31)
9,684
6,476
6,476
1,135
0
620
88
357
70
0
2,073
0
0
1,189
0
884
0
Total $
3,229.05 ± 1,306.96
2,159.65 ± 1,033.12
2,159.65 ± 1,033.12
378.33 ± 244.31
0
206.67 ± 206.67
29.33 ± 19.06
119.00 ± 55.32
23.33 ± 4.41
0
691.07 ± 389.36
0
0
396.40 ± 349.59
0
294.67 ± 49.67
0
Mean per patient $ ± SE
Severe (N = 3)
10,819
4,138
4,138
4,511
0
1,000
83
1,808
110
1,510
2,170
0
0
764
0
1,406
0
Total $
601.06 ± 137.26
229.90 ± 93.63
229.90 ± 93.63
250.61 ± 84.65
0
55.56 ± 25.66
4.61 ± 1.25
100.44 ± 48.73
6.11 ± 0.86
83.89 ± 58.39
120.55 ± 17.08
0
0
42.44 ± 11.57
0
78.11 ± 11.56
0
Mean per patient $ ± SE
Unknown (N = 18)
126
Labs
1,082
Inpatient
Direct Costs
Total $
. Table 126-5 Total Annual Costs for Adults by Provider-assessed Severity
2176 Quality of Life and Costs in Atopic Dermatitis
Quality of Life and Costs in Atopic Dermatitis
126
from the third-party payer perspective over a 1-year time period. Average cost-effectiveness in the base case was as seen in > Table 126-6. The average cost-effectiveness was best (lowest) with tacrolimus. Incremental costeffectiveness ratios (ICERs) were not meaningful in the base case because none of the therapies provided additional DCDs at additional cost; each regimen that provided additional DCDs did so at a lesser cost. In sensitivity analyses, in which values of key variables were varied within clinically achievable norms, cost-effectiveness ratios were sensitive to the duration of continuous therapy allowed with HPTCs. Comparing 4-week HPTCs and tacrolimus, the model was sensitive to the relapse rate of patients being treated with tacrolimus ointment and the costs and efficacy rate of secondary treatment. Adjusting for DCDs that occurred during the primary treatment periods was slightly favorable to tacrolimus ointment cost-effectiveness. In a sensitivity analysis, a combination of low costs and high efficacy for secondary treatment caused topical corticosteroid treatment to be more cost-effective than tacrolimus. Conversely, expensive secondary therapies resulted in better cost-effectiveness for tacrolimus therapy because patients who received tacrolimus used less secondary treatment. It is for this reason, the potential for modification in ‘‘dominant’’ strategy with the availability of new or different comparator or concomitant therapies, that cost-effectiveness analyses should be continuously re-evaluated rather than static (Arnold, 2007). An NHS evaluation of this study concluded that the ‘‘failure of Ellis and colleagues to value potential credible differences in resource consumption between tacrolimus and TCS might explain the sensitivity of their results to changes in the treatment pathways, concluding that tacrolimus is dominant if TCS are used for 2 weeks and steroids are dominant if used for 4 weeks. The analysis has significant methodological flaws and is of limited relevance to the UK’’ (Garside et al., 2005). Similarly, Ellis and colleagues (Ellis et al., 2006) sought to determine the comparative costutility (i.e., patient-preference-weighted effectiveness) of 1% pimecrolimus cream versus traditional therapy of waiting until TCS are medically necessary in patients aged 2–17 years with mild-to-moderate AD over a 1-year time frame. A Markov model was again used to perform the evaluation and clinical efficacy data to inform the model were obtained from a 12-month, double-blind, multinational, multicenter, controlled clinical trial. The outcome measure was the cost per quality-adjusted life-year (QALY) gained. Health states were based on Investigator’s Global Assessment (IGA) scores to assess disease severity/flares as follows: IGA 0/1 (clear, almost clear), IGA 2 (mild disease), IGA 3 (moderate disease) and IGA 4/5 (severe or very severe disease) and patients could, as previously described, transition between the states with probabilities as . Table 126-6 Average cost-effectiveness of high-potency topical corticosteroids (HPTCs) and Tacrolimus ointment Initial treatment option
Average cost (in US$)
Average effectiveness (Number of DCDs)
Average cost-effectiveness (Cost per DCD) (in US$)
HPTCs in 2-week cycles
1,682
185
9.08
HPTC in 4-week cycles
1,317
194
6.80
tacrolimus ointment
1,323
190
6.97
2177
2178
126
Quality of Life and Costs in Atopic Dermatitis
obtained in the clinical study. Assumptions on the annual frequency of office visits were based on IGA score. Utility values for each health state were previously determined from an independent parent survey in which they were asked to assign a visual analog scale (VAS) score to each. These values were then transformed to utility values using the power function from Torrance. Only direct medical costs were considered, and deductibles and copay (coinsurance) were not taken into consideration. Adjusting costs to 2004 US dollars using the Consumer Price Index, the incremental cost-utility ratio for treating patients with pimecrolimus versus conventional therapy was calculated as US $38,231 per QALY gained, indicating that pimecrolimus is cost-effective by the commonly-used threshold of Figure 126-1 reproduced below shows that, in general, as the willingness-to-pay (WTP) threshold (that is, the maximum a decision maker is prepared to pay for a gain of one QALY in health outcome) increased, the more likely that pimecrolimus was cost-effective as first-line therapy and at usual levels of WTP. In these scenarios, first-line pimecrolimus treatment showed a cost of £11,909 per QALY gained for children and £16.856 for adults. Since the probability of pimecrolimus being cost-effective rarely exceeded 0.5 regardless of the willingness-to-pay/QALY, the authors concluded that it was not preferential to use pimecrolimus over emollients in any scenario. They cautioned, however, that ‘‘differences of
Quality of Life and Costs in Atopic Dermatitis
126
. Figure 126-1 Cost-effectiveness acceptability curve for scenarios of treatment for adult facial eczema where pimecrolimus (Pim) is used as first- or second-line treatment. QALY, quality-adjusted life year
treatment benefit (QALYs) between comparators in the model are generally very small and there is considerable uncertainty in underlying data.’’ Hjelmgren and colleagues (Hjelmgren et al., 2007) developed a Markov model to analyze the cost-effectiveness of treatment with tacrolimus ointment versus TCS in patients with moderate to severe AD. The model was specific to Swedish treatment practice and consisted of four health states: severe AD, first-line treatment; severe AD, second-line treatment; moderate AD; and, virtually cleared. The societal perspective was undertaken. Data sources included a randomized, double-blind RCT, a patient survey with questions about AD symptoms, QOL (using a VAS), AD-related healthcare resource utilization (including absenteeism) and unit prices of healthcare resources in Sweden. Costs were calculated in Swedish krona (SEK) at 2004 prices and transformed to pounds sterling (£). The QALY gained was the outcome measure. An ordinary least square regression model was used to estimate a relationship between a calculated disease severity index and the VAS score. Using resource use data from the RCT, the ICER (cost per QALY gained) of tacrolimus over TCS was £12,304 in severe AD and £8,269 in moderate AD. These values were £3,875 and £2,334 for severe and moderate AD, respectively, when using data from the patient survey. Therefore, use of tacrolimus would be considered cost-effective in this situation. Although patient productivity was reported as having been measured, no data of this nature were apparent in the article. It is unusual to take utility measurements directly from the VAS indicator, as indicated by the authors; typically rating scores are converted to utilities using a power function (Torrance et al., 1996). However, assuming that the higher QOL scores and greater QALYs calculated for tacrolimus were at least somewhat realistic and translated into greater productivity, this would have only reduced the difference between the treatment options, thereby improving the cost-effectiveness of tacrolimus.
2179
2180
126
Quality of Life and Costs in Atopic Dermatitis
Coyle and Barbeau (Coyle and Barbeau, 2004) evaluated the cost-utility of pimecrolimus versus emollients (TCS could be used by either group if disease flares occurred) primarily using the results of an RCT in both adults and children to inform a Microsoft Excel spreadsheet model. The model was evaluated from both government and societal perspectives, the latter including patient productivity. The four health states studied were clear/almost clear, mild, moderate, and severe/very severe disease. Data sources in addition to the RCT included a previously published Canadian burden-of-illness study (health state costs by disease severity, some resource use, productivity), database of provincial drug formulary claims (medication costs), an expert panel (resource use), a Duke Clinical Research Institute study (utility values; VAS converted to utilities using standard power function relationship (Torrance et al., 1996). The QALY gained was the outcome measure. The time frames were 169 days for adults and 360 days for children. The incremental cost per QALY for pimecrolimus was US $40,000 and 37,000 for children and adults, respectively, from the health care system perspective and US $38,000 and 35,000 for children and adults, respectively, from the societal perspective. Therefore, pimecrolimus appeared to be a cost-effective treatment in both groups.
7.3
Non-Corticosteroids
Salo and colleagues (Salo et al., 2004) evaluated the cost-effectiveness of cyclosporin A versus UVAB therapy in patients with AD based on the results of a prospective clinical trial of Finnish patients incorporating both direct and indirect costs. All costs were first estimated in Finnish marks (FIM) and converted to US dollars (USD 1~FIM 5.1). The effectiveness metric was the number of days in remission. As cyclosporin A was both less costly (US $5,438 vs. US $5,635) and more effective (191 vs. 123 remission days/patient/year) than UVAB therapy, an ICER calculation was unnecessary.
Summary Points The burden of AD is multidimensional and includes: the chronic nature of the disease, its
effect on physical appearance, the disruption of physical and psychosocial development for children, the impact on lifestyle and economic factors. AD has a considerable impact on the HRQOL of the patient, the parent, the caregiver, the family and the partner of the patient with AD. Adults with AD experience similar negative effects in QOL as that of children. There is a negative correlation between HRQOL and disease severity; that is, as disease severity increases, HRQOL tends to decrease. While AD has been shown to negatively impact HRQOL, there is evidence that effective treatment, especially immunomodulators, can significantly improve HRQOL. The direct costs of AD range anywhere from US $150 to 580 per patient per year. Outpatient visits and medications comprise the majority of direct costs of AD. Both the direct and indirect costs of AD increase with worsening disease severity in adults. In comparisons between once-daily and twice-daily applications of TCS, it remains debatable which regimen is more cost-effective. Studies have generally shown immunomodulators to be cost-effective versus TCS.
Quality of Life and Costs in Atopic Dermatitis
126
References Abramovits W, Boguniewicz M, Paller AS, Whitaker-Worth DL, Prendergast MM, Tokar M, et al. (2005). Pharmacoeconomics. 23(6): 543–566. Anderson RT, Rajagopalan R. (1997). J Am Acad Dermatol. 37(1): 41–50. Arnold RJ. (2007). Drug Discov Today. 12(21–22): 960–965. Arnold RJ, Donnelly A, Altieri L, Wong KS, Sung J. (2007). Manag Care Interface. 20(2): 18–23. Basra MK, Sue-Ho R, Finlay AY. (2007). Br J Dermatol. 156(3): 528–538. Beattie PE, Lewis-Jones MS. (2006). Br J Dermatol. 155 (6): 1249–1255. Ben-Gashir MA, Seed PT, Hay RJ. (2002). J Eur Acad Dermatol Venereol. 16(5): 455–462. Ben-Gashir MA, Seed PT, Hay RJ. (2004). Br J Dermatol. 150(2): 284–290. Carroll CL, Balkrishnan R, Feldman SR, Fleischer AB, Jr., Manuel JC. (2005). Pediatr Dermatol. 22(3): 192–199. Chamlin SL. (2006). Dermatol Ther. 19(2): 104–107. Chamlin SL, Frieden IJ, Williams ML, Chren MM. (2004). Pediatrics. 114(3): 607–611. Chamlin SL, Lai JS, Cella D, Frieden IJ, Williams ML, Mancini AJ, et al. (2007). Arch Dermatol. 143(6): 768–772. Chren MM, Lasek RJ, Quinn LM, Mostow EN, Zyzanski SJ. (1996). J Invest Dermatol. 107(5): 707–713. Coghi S, Bortoletto MC, Sampaio SA, Andrade Junior HF, Aoki V. (2007). Clinics. 62(3): 235–242. Coyle D, Barbeau M. (2004). J Cutan Med Surg. 8(6): 405–410. Delea TE, Gokhale M, Makin C, Hussein MA, Vanderpoel J, Sandman T, et al. (2007). J Manag Care Pharm. 13(4): 349–359. Drake L, Prendergast M, Maher R, Breneman D, Korman N, Satoi Y, et al. (2001). J Am Acad Dermatol. 44(1 Suppl.): S65–72. Ehlken B, Mohrenschlager M, Kugland B, Berger K, Quednau K, Ring J. (2005). Hautarzt. 56(12): 1144–1151. Ellis CN, Drake LA, Prendergast MM, Abramovits W, Boguniewicz M, Daniel CR, et al. (2003). J Am Acad Dermatol. 48(4): 553–563. Ellis CN, Drake LA, Prendergast MM, Abramovits W, Boguniewicz M, Daniel CR, et al. (2002). J Am Acad Dermatol. 46(3): 361–370. Ellis CN, Kahler KH, Grueger J, Chang J. (2006). Am J Clin Dermatol. 7(2): 133–139. Emerson RM, Charman CR, Williams HC. (2000). Br J Dermatol. 142(2): 288–297. Emerson RM, Williams HC, Allen BR. (2001). Br J Dermatol. 144(3): 514–522.
Evers AW, Duller P, van de Kerkhof PC, van der Valk PG, de Jong EM, Gerritsen MJ, et al. (2008). Br J Dermatol. 158(1): 101–108. Finlay AY, Khan GK. (1994). Clin Exp Dermatol. 19(3): 210–216. Fivenson D, Arnold RJ, Kaniecki DJ, Cohen JL, Frech F, Finlay AY. (2002). J Manag Care Pharm. 8(5): 333–342. Garside R, Stein K, Castelnuovo E, Pitt M, Ashcroft D, Dimmock P, et al. (2005). Health Technol Assess. 9 (29): iii, xi-xiii, 1–230. Green C, Colquitt JL, Kirby J, Davidson P. (2005). Br J Dermatol. 152(1): 130–141. Green C, Colquitt JL, Kirby J, Davidson P, Payne E. (2004). Health Technol Assess. 8(47): 1–120. Hjelmgren J, Svensson A, Jorgensen ET, LindemalmLundstam B, Ragnarson Tennvall G. (2007). Br J Dermatol. 156(5): 913–921. Jenner N, Campbell J, Marks R. (2004). Australas J Dermatol. 45(1): 16–22. Kawashima M. (2006). Int J Dermatol. 45(6): 731–736. Kemp AS. (2003). Pharmacoeconomics. 21(2): 105–113. Kiebert G, Sorensen SV, Revicki D, Fagan SC, Doyle JJ, Cohen J, et al. (2002). Int J Dermatol. 41(3): 151–158. Lawson V, Lewis-Jones MS, Finlay AY, Reid P, Owens RG. (1998). Br J Dermatol. 138(1): 107–113. Lewis-Jones MS, Finlay AY. (1995). Br J Dermatol. 132 (6): 942–949. Lewis-Jones S. (2005). J Invest Dermatol. 125(6): viii. Lewis-Jones S. (2006). Int J Clin Pract. 60(8): 984–992. McKenna SP, Whalley D, Dewar AL, Erdman RA, Kohlmann T, Niero M, et al. (2005). Qual Life Res. 14(1): 231–241. Meads DM, McKenna SP, Kahler K. (2005). Qual Life Res. 14(10): 2235–2245. Misery L, Finlay AY, Martin N, Boussetta S, Nguyen C, Myon E, et al. (2007). Dermatology. 215(2): 123–129. Mozaffari H, Pourpak Z, Pourseyed S, Farhoodi A, Aghamohammadi A, Movahadi M, et al. (2007). J Microbiol Immunol Infect. 40(3): 260–264. Pitt M, Garside R, Stein K. (2006). Br J Dermatol. 154(6): 1137–1146. Ricci G, Bendandi B, Bellini F, Patrizi A, Masi M. (2007). Pediatr Allergy Immunol. 18(3): 245–249. Ricci G, Bendandi B, Pagliara L, Patrizi A, Masi M. (2006). J Pediatr Health Care. 20(5): 311–315. Salo H, Pekurinen M, Granlund H, Nuutinen M, Erkko P, Reitamo S. (2004). Acta Derm Venereol. 84(2): 138–141. Staab D, Kaufmann R, Brautigam M, Wahn U. (2005). Pediatr Allergy Immunol. 16(6): 527–533.
2181
2182
126
Quality of Life and Costs in Atopic Dermatitis
Su JC, Kemp AS, Varigos GA, Nolan TM. (1997). Arch Dis Child. 76(2): 159–162. Torrance GW, Feeny DH, Furlong WJ, Barr RD, Zhang Y, Wang Q. (1996). Med Care. 34(7): 702–722. Verboom P, Hakkaart-Van L, Sturkenboom M, De Zeeuw R, Menke H, Rutten F. (2002). Br J Dermatol. 147(4): 716–724. Ware JE, Jr., Sherbourne CD. (1992). Med Care. 30(6): 473–483.
Whalley D, Huels J, McKenna SP, Van Assche D. (2002). Pediatrics. 110(6): 1133–1136. Williams HC. (2000). Epidemiology of atopic dermatitis. Clin Exp Dermatol. 25(7): 522–529. Wittkowski A, Richards HL, Griffiths CE, Main CJ (2004). J Psychosom Res. 57(2): 195–200.
127 Quality of Life in Crohn’s Disease S. D. Wexner . J. C. Frattini 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2184 2 Crohn’s Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2184 3 Health Related Quality of Life . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2185 4 Crohn’s and HRQOL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2186 5 HRQOL and Adolescence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2187 6 Health Related Quality of Life and Medical Management of Crohn’s Disease . . . . 2189 7 Health Related Quality of Life and Surgical Management of Crohn’s Disease . . . . 2190 8 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2191 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2191
#
Springer Science+Business Media LLC 2010 (USA)
2184
127
Quality of Life in Crohn’s Disease
Abstract: > Crohn’s disease is a chronic, relapsing disease with an often early age of onset that can affect the entire gastrointestinal tract. Crohn’s disease can have a debilitating effect on a patient’s social, educational, professional, and familial activities; thus having a profound effect on their > health related quality of life (HRQOL). Only recently has an effort been made to determine the impact Crohn’s disease has on health related quality of life. Health related quality of life is a patient’s understanding of and relating to their illness within the scope of their lifestyle and in the process of doing so uses physical, psychological, cultural, educational, professional, and disease associated factors. Two types of instruments can be used to measure health related quality of life, general and disease specific, both of which look at various domains of a person’s life and must be validated, > reproducible, and reliable. The health related quality of life of Crohn’s patients has been found to be significantly worse when compared to the healthy population. It is affected by various factors like > disease activity, gender, medical and surgical treatment. It significantly affects the health related quality of life of adolescents in addition to the educational, familial, and professional aspirations of patients. Data on health related quality of life could provide valuable information on how to develop research protocols and alter treatment algorithms with the resolve of not only improving patient outcomes but also possibly improving utilization of resources. List of Abbreviations: HRQOL, Health related quality of life; IBD, disease
1
> Inflammatory
bowel
Introduction
Crohn’s disease is a chronic, relapsing disease with an often early age of onset, which can affect the entire gastrointestinal tract. It presents with a variety of signs and symptoms, which, in addition to the morbidities associated with the treatment, either medical or surgical, can impact a patient’s life in a multitude of ways. Crohn’s disease significantly affects a patient’s social, educational, professional, and familial activities. Until recently most of the data relating to Crohn’s disease outcomes have detailed its morbidity, disease free interval, disease activity, and the need for additional surgery. Health related quality of life (HRQOL) is an important measure when dealing with Crohn’s disease because of its chronic nature and impact on all aspects of life. New data have been emerging on the HRQOL in Crohn’s disease patients and the potential value of using it to improve patient outcomes, treatment algorithms, and utilization of resources.
2
Crohn’s Disease
Crohn’s disease most commonly presents during the teenage years and is very prevalent in patients under 30 years of age. It has another peak incidence, although smaller, during the sixth decade of life. Crohn’s disease equally affects both sexes and is more common in whites of Jewish and Western descent (Northern Europe and northern part of Eastern Europe). The etiology of Crohn’s disease is still unknown with various theories on a genetic, autoimmune, and/or infectious component. Crohn’s disease causes a transmural inflammation leading to the many physical manifestations ranging from chronic abdominal pain, cramping, diarrhea, fistulae, stricture, bleeding, obstruction, and/or perineal sepsis. Crohn’s disease is also accompanied by many extra-intestinal manifestations that affect the integumentary, skeletal, and the hepatobiliary systems.
Quality of Life in Crohn’s Disease
127
There are various medical and surgical treatments of Crohn’s disease. Each treatment being accompanied by its own set of sometimes severe morbidities. More specifically, the significant side effects of prolonged steroid and/or immunomodulator use and the various morbidities associated with one or many surgical procedures.
3
Health Related Quality of Life
There are many definitions of HRQOL but the World Health Organization describes it as ‘‘an individual’s perception of their position in life in the context of the culture and value systems in which they live and in relation to their goals, expectations, standards and concerns’’ (WHO, 1998). Others have simply defined it as a global measure of the patient’s perceptions, illness experience, and functional status that incorporates social, cultural, psychological, and diseaserelated factors (Drossman, 1993). Before a discussion on Crohn’s disease and HRQOL can be held the components of HRQOL and the ways to measure it must be explained. The four major components of HRQOL are physical function, emotional/social function, ability to productively work, and the absence of specific disease related symptoms (Cohen, 2002). More specifically the domains usually measured are sexual activity, social activity, ability to work or attend school, sports and recreation, and body image (Maunder et al., 1995). In addition, HRQOL can be measured by either general instruments applicable to various chronic diseases or by disease specific instruments. The various instruments available are listed in > Tables 127-1 and > 127-2. There are advantages and disadvantages in using both types of instruments. Disease specific questionnaires are very good at evaluating, providing more information on, and identifying changes within the specific disease for which they were developed and validated. Unfortunately, they cannot be used to compare the specific disease to other diseases and they cannot be applied to other diseases. General instruments . Table 127-1 General health related quality of life questionnaires used for inflammatory bowel disease 1. Sickness Impact Profile (SIP) 2. Psychological General Well-being index (PGWBI) 3. Symptom Checklist-90 (SCL-90) 4. Global Symptom Index (GSI) 5. EuroQOL
. Table 127-2 Disease specific health related quality of life questionnaires used for inflammatory bowel disease 1. Cleveland Clinic Questionnaire 2. Inflammatory Bowel Disease Questionnaire-36 (IBDQ-36) 3. Short Inflammatory Bowel Disease Questionnaire (Short IBDQ) 4. Rating Form of Inflammatory Bowel Disease Patient Concerns (RFIPC) 5. McMaster Inflammatory Bowel Disease Questionnaire 6. IMPACT (pediatrics)
2185
2186
127
Quality of Life in Crohn’s Disease
can be used to evaluate a broad range of diseases and allow comparisons between them. However they do not have the ability to detect disease specific changes (Cohen, 2002; Maunder et al., 1995). In addition to the type of instrument used to measure HRQOL, each instrument must have the following criteria for accurate and optimal measurement: > validity, > reliability, reproducibility, and > responsiveness. Validity assures that the instrument is truly measuring what it is intended to measure. Reliability assures consistency of the results. Reproducibility assures that the results will be similar if others use the same instrument. Responsiveness allows the detection of small changes relating to a disease (Cohen, 2002; Maunder et al., 1995). The most commonly used instruments to assess HRQOL in Crohn’s are the Inflammatory Bowel Disease Questionnaire (IBDQ) (Guyatt et al., 1989), the Rating Form of Inflammatory Bowel Disease Patient Concerns (RFIPC) (Drossman et al., 1991), and the Cleveland Clinic Questionnaire (Farmer et al., 1992). All of these instruments were found to be valid, reliable, and reproducible. The IBDQ contains 32 questions that are administered by a clinician. The questions fall in one of four domains of gastrointestinal symptoms, systemic symptoms, emotional dysfunction, and social dysfunction. The IBDQ has been further modified to a 36 question self administered questionnaire which also has been validated (Love et al., 1992; Maunder et al., 1995). The RFIPC contains 25 questions that are selfadministered. The four domains detailed in the questions cover impact of disease, sexual intimacy, complications of disease, and body stigma. The clinician administered Cleveland Clinic questionnaire contains 18 questions covering the domains of functional/economic, social/recreational, affect/life, and general (Farmer et al., 1992; Maunder et al., 1995).
4
Crohn’s and HRQOL
Within the last 10–15 years HRQOL has become an important outcome measure in Crohn’s disease. Studies have been done comparing HRQOL in patients with Crohn’s disease and in healthy subjects; individuals with other chronic diseases; patients with active and inactive Crohn’s disease; and individuals treated both surgically and medically (Cohen, 2002). Also, pharmacoeconomic analyses, which consider both the cost and outcomes (HRQOL), have increased in number (Feagan, 1999). The measurement of HRQOL can have a direct impact on the improvement of patient care and thus improve patient outcomes and possibly improve allocation and utilization of resources. Cohen (2002) identified three studies that demonstrated that the HRQOL was lower in patients with Crohn’s disease than in healthy controls. Only one of the three studies used a general instrument. In addition Cohen’s review explained that HRQOL is worse in patients with Crohn’s disease than in patients with ulcerative colitis but similar to or better than in individuals with rheumatoid arthritis, chronic obstructive pulmonary disease, back pain, or renal failure. Cohen (2002) also reviewed 18 articles which measured HRQOL in relation to Crohn’s disease activity. In all but one of the articles was a correlation between increased disease activity and poor HRQOL. Another study identified that those patients with other markers of increased disease activity, the presence of extra intestinal manifestations and increased number of relapses per year had a statistically significantly lower HRQOL scores on the IBDQ scale (Lopez Blanco, 2005). Crohn’s disease is a chronic disease in which a patient may live a majority of their life dealing with its affliction. A study looking to determine if the HRQOL was different between patients newly diagnosed with Crohn’s disease and individuals carrying the diagnosis for
Quality of Life in Crohn’s Disease
127
greater than 20 years found no difference. In this study, one hundred fifty two patients having the disease greater than 20 years were compared to 69 having Crohn’s disease less than 10 years using the Rating Form for Inflammatory Bowel Disease Patients’. Although the HRQOL was not different, this study identified several significant patient concerns regarding their disease. They are the possible need for an ostomy, the uncertain nature of the disease, and the lack of energy that correlates with similar findings in several other studies (Canavan et al., 2006). These concerns underscore the constant burden that Crohn’s disease places on a patient’s every day life. In a longitudinal, prospective study of 231 patients with Crohn’s disease, BlondelKucharski et al. (2001) confirmed previous findings that HRQOL was poorer with female sex, tobacco use, and corticosteroid treatment. In addition, HRQOL was confirmed to be underestimated by physicians treating patients with Crohn’s disease. Crohn’s disease also adversely impacts life on family and professional level. Andersson et al. (2003) found that Crohn’s disease has a significantly negative impact on family and professional lifestyles. In a prospective study of 127 Crohn’s disease patients and 266 controls, 68% of Crohn’s patients were living with a significant other whereas 78.4% of the controls were. Only 67.7% of the Crohn’s disease patients had children whereas 78% of the controls did. All of these findings were statistically significant. In addition, a statistically significant difference was found in the patients with Crohn’s disease with regard to their amount of sick leave and the number of those patients with a full disability pension. These findings correlated with previous studies on the same categories (Binder, 1985; Sorensen et al., 1987). In a cohort of 573 Crohn’s disease patients treated with infliximab, patients who were deemed in > remission as defined by the Crohn’s Disease Activity Index had not only statistically significant better employment rates but also had fewer hospitalizations and surgical procedures. The SF36 questionnaire was used in this study to determine HRQOL and those in remission had scores close to the general population; whereas, those not in remission had a statistically significant lower HRQOL (Lichtenstein et al., 2004). Poor family and professional lifestyles may have negative impact on a person’s mood. Whether or not a depressive mood has an affect on the activity Crohn’s disease has been postulated. To determine if in fact this supposition is true, a study on the impact of a depressive mood on relapse rates of inflammatory bowel disease (IBD) was undertaken. Seventy eight percent of the study population had Crohn’s disease. At the initiation of the study, all of the patients were in remission after an acute attack. The results of this study, in which the IBDQ questionnaire was used, demonstrated that the patients with a depressive mood had higher anxiety and lower HRQOL which where risk factors for earlier recurrence and more active disease (Mittermaier et al., 2004). Another study of IBD patients, in which 65% of the cohort had Crohn’s disease, evaluated the affect of psychological disorders on HRQOL using the SF-36 questionnaire. The study demonstrated that disease severity/activity and depression both independently lowered HRQOL. Also, after adjusting for disease severity and gender, patients with psychological disorders persisted with a lower HRQOL (Guthrie et al., 2002).
5
HRQOL and Adolescence
As stated earlier, the most common age of presentation of Crohn’s disease is during the teenage years. There is a significant amount of literature detailing the HRQOL of adolescents with Crohn’s disease. The validated questionnaire used in children to evaluate their HRQOL is the IMPACT or IMPACT II (Griffiths et al., 1999; Otley et al., 2002). The IMPACT II is a
2187
2188
127
Quality of Life in Crohn’s Disease
simplification of the original IMPACT questionnaire. It is composed of 35 questions using visual analogue scales for use in children ages 10–18 years. The six domains covered are bowel, systemic, emotional, social/functional, body image, and test/treatments. The IMPACT II questionnaire was used to evaluate a cohort of 280 pediatric IBD patients (77% with Crohn’s disease). This study identified that HRQOL was the lowest at the time of initial diagnosis but significantly improved over the next year after diagnosis and treatment. In addition, HRQOL was worse in those with more severe disease (Otley et al., 2006). The early age of onset for Crohn’s disease can significantly affect various aspects of education and career development. Fifty young adults with Crohn’s disease were interviewed at a mean of 14 years after their diagnosis. Twenty eight of 50 patients with Crohn’s disease had taken time off of two months or more from school, and despite the time off, their examinations grades were similar to healthy, age matched students. Fifty percent these patients with Crohn’s disease pursued higher education and only four were unemployed. Those four patients attributed their unemployment directly to Crohn’s disease. Overall, a majority of these teenagers believed that Crohn’s disease had a negative impact on both their educational and employment aspirations (Ferguson et al., 1994). In an attempt to gain insight as to how children understand the meaning of inflammatory bowel disease and its affects on daily life, 80 children with IBD (61 with Crohn’s disease) ages 7–19 were interviewed. The questions covered areas of impact of illness on self, functional impact, events/activities importance, challenges/obstacles, and relationships/family/peer/school. Each response was then critiqued and analyzed for content and themes. Common themes were concerns relating to IBD symptoms and treatments (mostly during active disease and exacerbations); vulnerability and lack of control over their lives; perceiving themselves in a negative light and being viewed as different than their peers; finding benefits from social support in coping; and finding personal ways of coping with IBD (Nicholas et al., 2006). These themes underscore the importance of the need of clinicians to recognize and treat not only the physical manifestations of Crohn’s disease but the psychosocial affects as well. With new and increasing amounts of data suggesting that HRQOL improves with decreasing disease activity, some clinicians may begin relying more on HRQOL measures than actual mucosal healing to dictate treatment (> Table 127-3). A recently published study underscores the importance of treating the whole patient because an improved HRQOL may not reflect healing on a mucosal level. Twenty-six children
. Table 127-3 General and psychosocial factors associated with HRQOL General – HRQOL worse in Crohn’s disease than Ulcerative colitis – Decreased disease activity associated with improved HRQOL – Improved HRQOL not always associated with mucosal healing Psychosocial – Increase rate of those either unmarried or divorced – Those in remission have better employment rates and take less sick time – Crohn’s disease patients with depression have more active disease and lower HRQOL – Increased concerns of need for stoma, uncertainty of disease, need for future surgery
Quality of Life in Crohn’s Disease
127
with Crohn’s disease were exclusively fed with enteral nutrition for 8 weeks. Twenty-three of twenty-six children had improved HRQOL as measured by IMPACT II and had also achieved clinical remission. Despite an improved HRQOL, there was no correlation between this and the degree of mucosal healing as measured by histological or endoscopic scores (Afzal et al., 2004) (> Table 127-4). . Table 127-4 Factors associated with HRQOL and adolescence – Despite being absent from school more, grades are equivalent to healthy children – Feel vulnerable, feel a lack of control over their lives, have a negative self image – Feel social support is a benefit and they find ways to cope with their disease
6
Health Related Quality of Life and Medical Management of Crohn’s Disease
There are a multitude of medications, including corticosteroids, 5-ASA compounds, and > immunomodulators, used to treat Crohn’s disease with some being used to induce remission and other used for maintenance of remission. While many of these medications may improve HRQOL by decreasing disease activity by induction and maintenance of remission, the side effects of the medications themselves may have a negative impact on HRQOL. One way to improve HRQOL is to avoid the use of those medications associated with significant side effects and lower HRQOL. A study of 169 patients with Crohn’s disease found lower HRQOL in those treated with corticosteroids or the immunomodualtor, azathioprine, as compared to non-users (Bernkley et al., 2005).Another study in which 33 patients in Crohn’s remission, treated with either azathioprine or 6-mercaptopurine, were compared with both 66 healthy individuals and 14 with active Crohn’s disease found contradictory results. Using the SF-36 questionnaire and Hamilton anxiety and depression scales, the authors found that both HRQOL and psychological well being were restored to normal levels in patients with thiopurinic induced remission (Calvet et al., 2006). Other medications, mesalazine and controlled ileal release budesonide, have been found to be equivalent to or effective alternatives to glucocorticoids with less side effects (Campieri et al., 1997; Clemett and Markham, 2000; Prantera et al., 1999; Rutgeerts et al., 1994). In comparing the two, budesonide was found to be more effective than mesalazine in inducing Crohn’s remission (Thomsen et al., 1998). Thomsen et al. (2002) demonstrated in 182 patients with Crohn’s disease that controlled release ileal budesonide improved HRQOL more than mesalazine. A newer modality used to treat both fistulizing Crohn’s disease and luminal disease refractory to conventional therapy is Infliximab (Remicade1). Infliximab is a monoclonal antibody against tumor necrosis factor-alpha. A recent review of the use of Infliximab and its effect on HRQOL found four articles that met their inclusion criteria and were found suitable for evaluation. Two were randomized control trials, two were cohort studies, and one of the four demonstrated long-term results. All four demonstrated significant improvements in HRQOL. This review also demonstrated that there was a decrease in hospitalizations and medical and surgical procedures after treatment (Koelewijn et al., 2006). These results may suggest that these decreases can lead to improved HRQOL and improvement in utilizations of resources (> Table 127-5).
2189
2190
127
Quality of Life in Crohn’s Disease
. Table 127-5 Effects of certain medications on HRQOL Corticosteroids negative Thiopurinic positive/negative Mesalamine positive Budesonide positive Remicade positive
7
Health Related Quality of Life and Surgical Management of Crohn’s Disease
In addition to the medical management of Crohn’s disease, surgical treatment plays a significant role. Approximately 70–80% of those with Crohn’s disease will need at least one surgical procedure in their lifetime (Shorb, 1989). Surgery and the potential need for an ostomy are some of the biggest concerns for Crohn’s disease patients (Drossman et al., 1989, 1991). Multiple studies have demonstrated that surgery significantly improves HRQOL postoperatively (Delaney et al., 2003; Thirlby et al., 1998; Tillinger et al., 1999; Yazdanpanah et al., 1997). These studies looked at time points from within the first 30 days post operatively up to 24 months. Delaney et al. (2003) looked at the HRQOL within 30 days of surgery in 82 patients undergoing a variety of different surgical procedures. Quality of life was most improved in females and in those who did not experience postoperative complications. In addition, 80% of patients were pleased with the surgical outcome and said they would have their surgery performed again. The type of procedure, activity of disease, and previous history of surgery did not affect HRQOL. Thirlby et al. (1998) also found that surgery improved HRQOL regardless of disease site, extent, or history of previous surgery. A study that looked at 16 patients postoperatively for up to 2 years demonstrated improved HRQOL except in those who developed chronic active disease (Tillinger et al., 1999). Twenty-six patients undergoing ileocolonic resection again had improved HRQOL 3 months postoperatively but these patients continued to have significant concerns with having an ostomy, the uncertainty of their disease, and the possibility of more surgery in the future (Yazdanpanah et al., 1997). Most articles on the relationship of HRQOL and surgery compare pre-operative HRQOL to post-operative HRQOL. Few have compared those patients who have had surgery and are in remission to those in remission and have never had surgery. One such study found that HRQOL was essentially equivalent in those patients in which remission was induced by either medical or surgical management yet the HRQOL significantly improved when compared to those with active Crohn’s disease (Casellas et al., 2000) (> Table 127-6). . Table 127-6 Effects of surgery on HRQOL – Surgery improves HRQOL as early as 30 days and up to 2 years post operatively – Improved HRQOL not associated with type of surgery, disease activity, extent of disease, disease site, or previous surgery
Quality of Life in Crohn’s Disease
127
Psychosocial factors affect the HRQOL after a surgical procedure also. Maladaptive coping behaviors (self-control, self-blame, and escape) are independent predictors of a lower HRQOL postoperatively; whereas, patients who had more social support had higher HRQOL (Moskovitz et al., 2000).
8
Conclusion
Crohn’s disease is a chronic and debilitating disease the course of which cannot be predicted. The chronic, unpredictable nature and the vast impact on all aspects of life are major concerns for the patient with Crohn’s disease. It is clear that health related quality of life in addition to disease activity and disease-associated morbidity should be used to measure treatment outcomes in Crohn’s disease. The data presented raise several questions about the treatment of Crohn’s disease. Longterm results (>20 years) demonstrate that HRQOL does not always improve over time. Does this finding imply that our current treatment regiments are inadequate? In addition to treating the disease, do we need to better educate the patient on the disease? In doing so, will the patient be better prepared psychologically to deal with impact of Crohn’s disease and thus improve HRQOL? Should surgery be performed earlier in the course of disease if HRQOL is significantly improved postoperatively? Can patients with Crohn’s disease have less psychosocial, familial, and educational dysfunction if they are provided better social support and if their patient education is improved? In addition, if Crohn’s disease is exacerbated by a poor psychological status then by improving it physicians can decrease disease activity and improve HRQOL. Despite finding a multitude of factors which can improve HRQOL, this area of research has also found that patients still have significant concerns-the need of an ostomy, the uncertainty of the disease, and the need for future surgery. In addition, an improved HRQOL does not always reflect complete mucosal healing and should not be used as a sole measure of improvement. The goal in treating Crohn’s disease should be to decrease hospitalizations, surgical morbidity, and the overall negative impact on a patient’s life. A study on the natural course of Crohn’s disease has shown that 33% of patient’s with Crohn’s disease who have developed complications will need hospitalization or surgery within the first year and this number decreases to 13 and 3% in the subsequent years (Lichtenstein et al., 2004). This problem can create a significant economic burden. The costs of Crohn’s disease in the United States in 1990 were estimated at 1.0–1.2 billion dollars and the cost for lost productivity at work due to IBD was estimated at 0.4–0.8 billion dollars (Feagan, 1999; Hay and Hay, 1992). This impact on utilization of resources compounds the significant burden that is placed not only on all aspects of a patient’s life but also on those around them. HRQOL measures are just another tool in which the variables related to Crohn’s disease morbidities and outcomes can be determined and be used to modify treatment algorithms. They also can be used to guide the development of future research and treatment options to improve the outcome of Crohn’s disease and to improve the utilization of resources.
Summary Points Crohn’s disease is an inflammatory disease that may affect the entire gastrointestinal tract from mouth to anus. The morbidities of the disease, bleeding, obstruction, perforation, and/or fistulas, affect a patient’s well-being and lifestyle thus impacting quality of life.
2191
2192
127
Quality of Life in Crohn’s Disease
Health related quality of life could be assessed by questionnaires that have been developed
and accepted by the medical field. The various areas they measure are physical function, emotional/social function, ability to work productively, and the absence of specific disease related symptoms. Patients with Crohn’s disease have HRQOL that is worse than the general population. The activity of the disease and psychological state are important factors in its determination. Patients with Crohn’s disease have concerns regarding the possible need for an ostomy, the uncertain nature of the disease, and the lack of energy. During adolescence, Crohn’s disease has a negative impact on both educational and professional aspirations. In addition, adolescents feel more vulnerable with a lack of control over their lives, they perceive themselves in a negative light, and feel different than their peers; they find benefit from social support. The medications used to treat Crohn’s disease can cause significant side effects that can have a direct impact on quality of life. Newer, more effective and less morbid medications are available and have shown to improve HRQOL. Despite the possible morbidity and mortality associated with surgery, after surgery patients with Crohn’s disease have improved HRQOL regardless of the type of surgery and extent of disease. They also would have the procedure performed again if they had the chance. Developing research and treatment protocols involving HRQOL may have a positive impact on outcome. The goal should be to induce remission and decrease hospitalizations and surgical procedures; thereby inducing a reciprocal effect of improving HRQOL which can not only improve outcome but can decrease disease activity.
References Afzal N, Van Der Zaag-Loonen H, Arnaud-Battandier F, Davies S, Murch S, Derkx B, Heuschkel R, Fell J. (2004). Aliment Pharmacol Ther. 20: 167–172. Andersson P, Olaison G, Bendtsen P, Myrelid P, Sjodahl R. (2003). Colorectal Dis. 5: 56–62. Bernkley T, Jahnsen J, Schulz T, Sauar J, Lygren I, Henriksen M, Stray N, Kjellevold O, Aadland E, Vatn M, Moum B. (2005). Eur J Gastroenterol Hepatol. 17: 1037–1045. Binder V, Hendriksen C, Kreiner S. (1985). Gut. 26: 146–150. Blondel-Kucharski F, Chircop C, Marquis P, Cortot A, Baron F, Gendre JP, Colombel JF. (2001). Am J Gastroenterol. 96: 2915–2920. Calvet X, Gallardo O, Coronas R, Casellas F, Montserrat A, Torrejon A, Vergara M, Campo R, Brullet E. (2006). Inflamm Bowel Dis. 12: 692–696. Campieri M, Ferguson A, Doe W, Persson T, Nilsson LG. (1997). Gut. 41: 209–214. Canavan C, Abrams KR, Hawthorne B, Drossman D, Mayberry JF. (2006). Aliment Pharmacol Ther. 23: 377–385.
Casellas F, Lopez-Vivancos J, Badia X, Vilaseca J, Malagelada JR. (2000). Am J Gastroenterol. 95: 177–182. Clemett D, Markham A. (2000). Drugs. 59: 929–956. Cohen RD. (2002). Aliment Pharmacol Ther. 16: 1603–1609. Delaney C, Kiran R, Senagore A, O’brien-Ermlich B, Church J, Hull T, Remzi F, Fazio V. (2003). J Am Coll Surg. 196: 714–721. Drossman D. (1993). In: Sleisenger MH, Fordtran JS (eds.) Gastrointestinal Disease: Pathophysiology, Diagnosis, and Management. WB Saunders, New York. Drossman DA, Patrick DL, Mitchell CM, Zagami EA, Appelbaum MI. (1989). Dig Dis Sci. 34: 1379–1386. Drossman DA, Leserman J, Li ZM, Mitchell CM, Zagami EA, Patrick DL. (1991). Psychosom Med. 53: 701–712. Farmer RG, Easley KA, Farmer JM. (1992). Cleve Clini J Med. 59: 35–42. Feagan BG. (1999). Aliment Pharmacol Ther. 13 (suppl. 4): 29–37.
Quality of Life in Crohn’s Disease Ferguson A, Sedgwick DM, Drummond J. (1994). Gut. 35: 665–668. Griffiths AM, Nicholas D, Smith C, Munk M, Stephens D, Dumo C, Sherman PM. (1999). J Pediatr Gastroenterol Nutr. 28: S46–S52. Guthrie E, Jackson J, Shaffer J, Thompson D, Tomenson B, Creed F. (2002). Am J Gastroenterol. 97: 1994–1999. Guyatt G, Mitchell A, Irvine EJ, Singer J, Williams N, Goodacre R, Tompkins C. (1989). Gastroenterology. 96: 804–810. Hay AR, Hay JW. (1992). J Clin Gastroenterol. 14: 318–327. Koelewijn C, Schrijvers A, Oldenburg B. (2006). Netherlands J Med. 64: 212–218. Lichtenstein GR, Yan S, Bala M, Hanauer S. (2004). Am J Gastroenterol. 99: 91–96. Lopez Blanco B, Moreno-Jimenez B, Devesa Mugica JM, Rodriguez Munoz A. (2005). Rev Esp Enferm Dig. 97: 887–898. Love JR, Irvine EJ, Fedorak RN. (1992). J Clini Gastroenterol. 14: 15–19. Maunder RG, Cohen Z, McLeod RS, Greenberg GR. (1995). Dis. Colon Rectum. 38: 1147–1161. Mittermaier C, Dejaco C, Waldhoer T, OefferlbauerErnst A, Miehsler W, Beier M, Tillinger W, Gangi A, Moser G. (2004). Psychosom Med. 66: 79–84. Moskovitz DN, Maunder RG, Cohen Z, McLeod RS, MacRae H. (2000). Dis Colon Rectum. 43: 517–521. Nicholas D, Otley A, Smith C, Avolio J, Munk M, Griffiths A. (2007). Health Qual Life Outcomes. 5L: 28. Otley A, Smith C, Nicholas D, Munk M, Avolio J, Sherman PM, Griffiths AM. (2002). J Pediatr Gastroenterol Nutr. 35: 557–563.
127
Otley A, Griffiths A, Hale S, Kugathasan S, Pfefferkorn M, Mezoff A, Rosh J, Tolia V, Markowitz J, Mack D, Oliva-Hemker M Wyllie R, Rothbaum R, Bousvaros A, Fernando Del Rosario J, Evans J, Blanchard W, Hyams J. (2006). Inflamm Bowel Dis. 12: 684–691. Prantera C, Cottone M, Pallone F, Annese V, Franze A, Cerutti R, Bianchi Porro G. (1999). Gastroenterol. 116: 521–526. Rutgeerts P, Lofberg R, Malchow H, Lamers C, Olaison G, Jewell D, Danielsson A, Goebell H, Thomsen OO, Lorenz-Meyer H. (1994). N Engl J Med. 331: 842–845. Shorb PE. (1989). Gastroenerol Clin North Am. 18: 111–128. Sorensen VZ, Olsen BG, Binder V. (1987). Gut. 28: 382–385. Thirlby R, Land J, Fenster F, Lonborg R. (1998). Arch Surg. 133: 826–832. Thomsen OO, Cortot A, Jewell D, Wright JP, Winter T, Veloso FT, Vatn M, Persson T, Pettersson E. (1998). N Engl J Med. 339: 370–374. Thomsen OO, Cortot A, Jewell D, Wright JP, Winter T, Veloso FT, Vatn M, Persson T, Pettersson E. (2002). Am J Gastroenterol. 97: 649–653. Tillinger W, Mittermaier C, Lochs H, Moser G. (1999). Dig Dis Sci. 44: 932–938. WHO. (1998). WHOQOL User Manual. Geneva: World Health Organization. Yazdanpanah Y, Klein O, Gambiez L, Baron P, Desreumaux P, Marquis P, Cortot A, Quandalle P, Colombel JF. (1997). Am J Gastroenterol. 92: 1897–1900.
2193
128 Impact of Self-Perceived Bothersomeness, Quality of Life and Overactive Bladder Ja Hyeon Ku . Soo Woong Kim 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2196
2
Prevalence of an Overactive Bladder . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2196
3 3.1 3.2
Assessment of the Quality of Life with an Overactive Bladder . . . . . . . . . . . . . . . . 2198 Generic Quality of Life Questionnaires for Overactive Bladder . . . . . . . . . . . . . . . . . . 2198 Disease-Specific Quality of Life Questionnaires for Overactive Bladder . . . . . . . . . 2198
4 4.1 4.2 4.2.1 4.2.2 4.3 4.4 4.5
Measurement of Urgency with an Overactive Bladder . . . . . . . . . . . . . . . . . . . . . . . . . 2199 Patient’s Perception . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2200 Impact of Overactive Bladder on Health Related Quality of Life . . . . . . . . . . . . . . . . 2201 United States . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2204 Europe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2205 Coping Behavior and Overactive Bladder . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2206 Impact of an Overactive Bladder on Sexual Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2207 Seeking Medical Attention . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2208
5
Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2208 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2208
#
Springer Science+Business Media LLC 2010 (USA)
2196
128
Impact of Self-Perceived Bothersomeness, Quality of Life and Overactive Bladder
Abstract: > Overactive bladder (OAB) is a condition characterized by urinary > urgency, with or without urge incontinence, urinary > frequency, and > nocturia. An estimated 10– 20% of individuals living in the US and Europe have an OAB. Although an OAB is not a normal part of aging, its prevalence increases with age. The causes of OAB, as with many bladder disorders, are multifactorial and are not completely understood. Whereas the symptoms of an OAB are not life threatening, they may contribute to other disabling conditions and are clearly recognized to cause significant disruption and have significant adverse effects on the quality of life (QOL). Urgency is the principal symptom of an OAB and thus, assessing this symptom and its impact on the health-related QOL is important. Because an OAB and urinary incontinence in women are associated with sexual dysfunction, an evaluation of sexual function should be part of the routine examination of women who present with an OAB. An OAB can lead to coping behaviors that significantly affect patient well-being. In addition, patient-perceived irritating symptoms significantly influence the QOL in women with an OAB. Thus, future research should help identify the patients have irritating factors associated with an OAB. Since patients may regard their symptoms as embarrassing, untreatable or an inevitable part of getting older, they may not report OAB symptoms despite significant discomfort. It is therefore of utmost importance that medical education on the symptoms of an OAB and other related problems improve, to help health care professionals identify and treat patients who will benefit from therapy. List of Abbreviations: FV chart, > frequency-volume chart; KHQ, King?0 s health questionnaire; OAB, overactive bladder; OAB-q, overactive bladder questionnaire; QOL, quality of life; SF-36, the medical outcomes study short form
1
Introduction
Overactive bladder (OAB) is characterized by urinary urgency, with or without > urge urinary incontinence, usually with frequency and nocturia (Abrams et al., 2002). Urgency is defined as a sudden compelling desire to urinate that is difficult to delay; urge incontinence is defined as involuntary leakage of urine accompanied by or immediately preceded by urgency (Abrams et al., 2002). This is a common and devastating condition. Apart from impairing physical health, an OAB may affect the psychological and social well-being of the patient with these symptoms. Generally, these symptoms and the coping strategies used to deal with them, have a negative effect on a patients’ health-related quality of life (QOL). It has been suggested that patients with an OAB often have greater QOL impairment than those suffering from > stress urinary incontinence (Simeonova et al., 1999). However, it is difficult to appreciate the extent of the disruption in the activities of daily living caused by the irritating symptoms of an OAB because the impact of an OAB varies depending on many different factors such as age and cultural beliefs. This chapter focuses on a number of contemporary issues that relate to the impact of the patients perception of the irritating symptoms associated with an OAB and the effect on the QOL.
2
Prevalence of an Overactive Bladder
OAB is a common health problem in both men and women. However, international studies have shown a wide-ranging variation in the prevalence of urinary symptoms between countries, making cross-country comparisons difficult (Hannestad et al., 2000). In addition, comparing
Impact of Self-Perceived Bothersomeness, Quality of Life and Overactive Bladder
128
the prevalence of incontinence among racial/ethnic groups ascertained in separate studies is problematic since the differences identified could be due to differences in the definition of incontinence, the selection of the study population or the study design. Milsom et al. (2000) randomly selected a population from six European countries. From this population, 17% of the respondents reported having OAB symptoms with 14% reporting frequency, 9% urgency and 6% urge incontinence (> Figure 128‐1) (Milsom et al., 2001). OAB affects approximately 16% of men and women in the United States (Stewart et al., . Figure 128‐1 Prevalence of different overactive bladder symptoms from a European population-based prevalence study. Frequency (85%) was the most commonly reported symptom among those with an overactive bladder, followed by urgency (54%) and urge incontinence (36%). The prevalence of individual symptoms occurring alone was small, as was the prevalence of respondents with all of the symptoms of an overactive bladder. Source: Milsom et al. (2001) reprinted with permission
2003). Unlike most women with OAB, most men with OAB do not experience incontinence (55 vs. 16%, respectively) (Stewart et al., 2003). In a questionnaire-based survey administered in 11 Asian countries, the overall prevalence of urge incontinence was 11.4% (Lapitan et al., 2001). In a study reported by Chen et al. (2003), the prevalence of an OAB was 18.6%. The perception and number of cases with an OAB was significantly increased among elderly women (over 65 years old, 39.3%). The prevalence of OAB increases with age (Milsom et al., 2001; Stewart et al., 2003). Milsom et al. (2001) reported that the prevalence of OAB increased from 3% in men aged 40–44 to 42% in those older than 75. In a study of women in Go¨teborg, Sweden, stress incontinence was more common than urge incontinence in women younger than 50 years of age, whereas in women 50 years of age or older, symptoms of mixed (concurrent stress and urge incontinence) or urge incontinence predominated (Simeonova et al., 1999).
2197
2198
128 3
Impact of Self-Perceived Bothersomeness, Quality of Life and Overactive Bladder
Assessment of the Quality of Life with an Overactive Bladder
The most useful way to assess the presence, severity and impact of a symptom or condition on a patient’s daily functioning and psychosocial well-being is by the use of a validated selfadministered questionnaire. Currently, there are two major types of QOL questionnaires, generic and disease-specific. To date, there are several instruments available to assess the severity of symptoms, perceived interference with activities of daily living and health-related QOL in patients with OAB.
3.1
Generic Quality of Life Questionnaires for Overactive Bladder
Generic QOL questionnaires (e.g., Medical Outcome Study Short Forms 36 [SF-36], European Quality of Life Scale and Sickness Impact Profile) offer some advantages because they are reliable, validated, readily available, and useful in assessing a broad range of populations and ages for many different disease states. However, the problem with using generic questionnaires is that the results are often insensitive to the specific condition measured and therefore fail to address many of the issues relevant to the disease.
3.2
Disease-Specific Quality of Life Questionnaires for Overactive Bladder
The King’s Health Questionnaire (KHQ) is a 21-item, condition-specific instrument to assess the QOL. It was developed to assess the health-related QOL in women with general symptoms of incontinence (Kelleher et al., 1997), but it is also a valid and reliable instrument for use in both men and women with OAB symptoms (Kobelt et al., 1999). The KHQ contains seven multi-domains: role limitations, physical limitations, social limitations, personal relationships, emotional problems, sleep/energy disturbances, severity measures, and two singleitem domains: general health perceptions and impact on life. The Urge-Incontinence Impact Questionnaire has seven domains (travel, activities, feelings, physical activities, relationships, sexual function, and nighttime bladder control) and has been demonstrated to be reliable and valid in patients with incontinence (Lubeck et al., 1999). This 32-item questionnaire was developed by adding and deleting items from the original questionnaire and by using focus groups, literature reviews and expert clinical opinion. The Overactive Bladder Questionnaire (OAB-q) is a 33-item, disease-specific measure designed to assess symptom bother and the impact of the OAB symptoms on the health-related QOL (Coyne et al., 2002). The OAB-q consists of a symptom bother scale and four health-related QOL domains: Concern, Coping, Sleep, and Social interactions. The Primary OAB Symptom Questionnaire, also called the OAB Bother Rating Scale is a recently developed 5-item questionnaire that assesses which symptoms of an OAB are the most bothersome to patients (Matza et al., 2005). On the first four items, patients rate how bothered they are by each OAB symptom (urinary urgency, urinary frequency, nocturia, and urge incontinence) during the past 2 weeks. The fifth item asks patients to indicate which of the four OAB symptoms bothers them the most. Recently, the Overactive Bladder Symptom Score has been validated to measure the overall symptom
Impact of Self-Perceived Bothersomeness, Quality of Life and Overactive Bladder
128
severity due to frequency, nocturia, urgency, and urge incontinence (Homma et al., 2006). Other instruments developed to assess the impact of an OAB on the health-related QOL are the Incontinence Quality of Life Index (Renck-Hooper et al., 1997), Urge Impact Scale (Debeau et al., 1999), and Urge Incontinence Impact Questionnaire (Brown et al., 1999). Importantly, when selecting a questionnaire, its items and overall performance in terms of reliability, validity and responsiveness must be considered.
4
Measurement of Urgency with an Overactive Bladder
Urgency is the principal symptom of an OAB and thus, assessing this symptom and its impact on the health-related QOL is important. A recent epidemiological survey revealed that the urinary urgency had a significant negative effect on the health-related QOL and increased the symptom bother, an effect that, in this community sample, was greater than for incontinence, frequency or nocturia (Coyne et al., 2004). Several instruments have been developed specifically to assess urinary urgency (> Table 128‐1). However, developing a standardized instrument that measures the experience of urgency across a spectrum or continuum, of severity is one of the more arduous challenges facing researchers currently investigating the OAB. The Urgency Rating Scale, recommended by the European Medicines Agency (2002), consists of a 5-point rating scale to be rated with every void, ranging from 1 (no urgency; I felt no need to empty my bladder but did so for other reasons) to 5 (urge incontinence; I leaked before arriving at the toilet). The Patient Perception of the Bladder Condition is a single-item, 6-point global scale that asks patients to rate their subjective impression of their current bladder problems (Coyne and Matza, 2002). The Indevus Urgency Severity Scale asks patients to rate their level of urgency on a 4-point scale, from 0 (no urgency) to 3 (extreme urgency discomfort that abruptly stops all activity/tasks) (Bowden et al., 2003). The Urgency Perception Scale was designed for use in clinical trials to evaluate patientperceived urgency (Cardozo et al., 2005). This instrument consists of a single question asking patients to describe their typical experience when they feel the need to urinate. The Overactive Bladder Symptom Composite Score point values were assigned to the corresponding IUSS value/void. This instrument is unique in that it seeks to capture the influence of daily frequency and daily urgency incontinence episodes, in addition to measuring the urgency severity as a surrogate measure of OAB severity (Zinner et al., 2005). The Urgency Perception Score as proposed by Blaivas et al. (2007) uses a single item question that was modified from the bladder sensation scoring system. It is scored on a 5 point scale ranging from voiding out of convenience (no urgency ¼ 0) to desperate urgency (score ¼ 4), and as such attempts to capture the perception of urgency on a continuum rather than urgency per se. With some of these scales, patients have the option of indicating that they had urge incontinence (an event) rather than the strongest feeling of urgency (a sensation) itself. In such cases, patients who have severe urgency, but not urge incontinence, do not have an option for endorsing the highest (worst) value, because they are not incontinent. Urgency severity scales that include the urge incontinence response option thus may be less useful than those that do not because such scales are attempting to measure two things at once, urgency and urge incontinence.
2199
2200
128 4.1
Impact of Self-Perceived Bothersomeness, Quality of Life and Overactive Bladder
Patient’s Perception
Women with severe incontinence are more bothered than those with mild incontinence (Hannestad et al., 2000). Recently, Oh and Ku (2007) found that statistically significant differences according to the patient-perceived severity for most domains of the SF-36 and the KHQ (> Tables 128‐2 and > 128‐3) (Oh and Ku, 2007). These findings suggest that patient-perceived bothersomeness significantly influences the QOL in women with OAB symptoms. Thus, future research should include the identification of the factors influencing the patient’s perception of bothersomeness. For example, a previous study indicated that . Table 128‐1 Scales utilized to assess/measure urgency Instruments Urgency Rating Scale
References
Scales
European 1 = no urgency Medicines Agency 2 = mild urgency (2002) 3 = moderate urgency 4 = severe urgency 5 = urge incontinence
Patient perception of bladder condition
Coyne and Matza (2002)
Does not cause me any problems at all Causes me some very minor problems Causes me some minor problems Causes me (some) moderate problems Causes me severe problems causes me many severe problems
Indevus Urgency Severity Scale
Bowden et al. (2003)
0: NONE – no urgency 1: MILD – awareness of urgency, but is easily tolerated and you can continue with your usual activity or task 2: MODERATE – enough urgency discomfort that it interferes with or shortens your usual activity or tasks 3: SEVERE – extreme urgency discomfort that abruptly stops all activity or tasks
Urgency Perception Scale
Cardozo et al. (2005)
1 = I am not usually able to hold urine (i.e. urge incontinence) 2 = I am not usually able to hold urine until I reach the toilet if I go immediately (i.e. urgency) 3 = I am usually able to finish what I am doing before going to the toilet (i.e. not urgency)
Overactive Bladder Symptom Composite Score
Zinner et al. (2005)
1 = no urgency 2 = mild urgency 3 = moderate urgency 4 = severe urgency 5 = urgency incontinence
Impact of Self-Perceived Bothersomeness, Quality of Life and Overactive Bladder
. Table 128‐1 (continued) Instruments Urgency Perception Score
References Blaivas et al. (2007)
128
Scales 0 = Out of convenience (no urge) 1 = Because I have a mild urge (but can delay urination for over an hour if I have to) 2 = Because I have a moderate urge (but can delay urination for more than 10 but less than 60 min if I have to) 3 = Because I have a severe urge (but can delay urination for less than 10 min) 4 = Because I have desperate urge (must stop what I am doing and go immediately)
The Urgency Rating Scale consists of a 5-point rating scale rated with every void; the Patient Perception of Bladder Condition is a single-item, 6-point global scale that asks patients to rate their subjective impression of their current bladder problems; the Indevus Urgency Severity Scale asks patients to rate their level of urgency on a 4-point scale; the Urgency Perception Scale consists of a single question asking patients to describe their typical experience when they feel the need to urinate; the Overactive Bladder Symptom Composite Score point values seeks to capture the influence of daily frequency and daily urgency incontinence episodes; the Urgency Perception Score attempts to capture the perception of urgency on a continuum rather than urgency per se. Source: Starkman JS, Dmochowski RR (2007) Neurourol. Urodyn. DOI 10.1002/nau. pp. 7–8 (reprinted with permission of Wiley-Liss, Inc., a subsidiary of John Wiley & Sons, Inc)
younger women appeared to be more bothered by their symptoms of stress urinary incontinence (Fultz et al., 2003). In addition, the clinical meaning of the patient’s perception of bothersomeness requires further definition. The extent to which women are bothered by their symptoms and report that their symptoms compromise their QOL is largely subjective. Patient-completed frequency-volume (FV) charts are commonly used in clinical trials as a primary tool for measuring subjective symptoms of the lower urinary tract. The FV chart has been considered as one of the most important tools for objectively assessing micturition patterns (Robinson et al., 1996). The FV charts allow the clinician to obtain information about voiding frequency, nocturia, the mean volume of urine passed, and provide documentation on voiding patterns to be established in the patient’s environment and during various daily activities. Therefore, the FV can provide important information on a patient’s voiding problem, which is not easily identified by other means. However, there has been no difference identified in the data from the FV charts based on the patient’s perceived severity of symptoms (Oh and Ku, 2007). Furthermore, the scores of the SF-36 and the KHQ domains do not correlate with the data of the FV charts (Oh and Ku, 2007). These findings suggest that since the objective clinical measures do not reflect the patient’s view, QOL measures should be included in clinical practice. In addition, this suggests that the instruments that measure physical symptoms do not measure the patient’s perceived degree of severity of the symptoms.
4.2
Impact of Overactive Bladder on Health Related Quality of Life
An OAB is irritating to patients and has negative effects on the health-related QOL (> Figure 128‐2) (Kobelt et al., 1999). In fact, all aspects of the QOL, including physical, social, psychological, occupational, domestic, and sexual functioning, can be impaired by OAB symptoms (> Table 128‐4).
2201
61.7 2.4
49.3 4.8
64.3 3.7
37.0 2.3
36.4 2.4
67.2 3.3
44.8 5.4
49.1 2.4
68.4 1.6
80.0 7.5
88.2 3.0
50.6 4.1
56.0 5.6
79.5 5.6
78.7 7.9
63.2 4.3
PF
RP
BP
GH
VT
SF
RE
MH
0.004
0.001
0.055
0.003
0.004
Figure 130-5) (Bohlke et al., . Figure 130-5 Comparison of mean SF-36 scores in tacrolimus-treated and tacrolimus-free immunosuppression in kidney transplant recipients. RP role limitations due to physical functioning; PF physical functioning; GH general health; VT vitality; BP bodily pain; RE role limitation due to emotional functioning; SF social functioning; MH mental health. The SF-36 scores were markedly better for patient on tacrolimus-based therapy, especially, for physical functioning scores. (Bohlke et al., 2006; Moons et al., 2003)
2006; Painter et al., 2005; Moons et al., 2003; Reimer et al., 2002). Sirolimus is a new molecule which is less nephrotoxic than calcineurin inhibitors and the use of sirolimus-based immunosuppression with early elimination of cyclosporine resulted in less fatigue and better vitality when compared with recipients on a combination of sirolimus, cyclosporine and steroids (Oberbauer et al., 2003). Tacrolimus, prednisolone and sirolimus are associated with an increased likelihood of gastrointestinal symptoms in renal transplant patients, and the overall increase in gastrointestinal symptoms may be a significant underlying reason for a reduction of HRQOL in renal transplant recipients (Ekberg et al., 2007).
4
Future Directions in HRQOL
Quality of life after kidney transplantation is good or excellent in the vast majority of recipients. Furthermore, the HRQOL of kidney transplant recipients is comparable to that of other chronic illness, yet HRQOL is not restored to the level of healthy population,
2235
2236
130
Quality of Life in Kidney Transplantation
especially for physical functioning scores. In this view, there are many applicable interventions to try to improve the HRQOL in kidney transplant recipients (> Figure 130-6). Treatment of side effects, especially the cutaneous ones, could provide a significant improvement of . Figure 130-6 Proposed interventions to improve the quality of life in kidney transplant recipients
HRQOL. The switch from mycophenolate mofetil to enteric-coated mycophenolic acid in renal transplant recipients with gastrointestinal side-effects may improve HRQOL, that may be comparable to recipients without gastrointestinal problems (Chan et al., 2006). Psychotherapy may have some degree of efficacy in the post-transplant setting. A formal psychosocial evaluation for transplant candidates should be recommended, and may determine the factors that may influence the decision to place an individual on the waiting list (Kasiske et al., 2002). Post-transplant psychotherapy should provide a multidisciplinary approach consisting of patient-initiated care to prevent post-transplant morbidities, an employment counseling and an enhanced social support (Dobbels et al., 2007). Intensive exercise training could provide a significant improvement in both physical and psychological functioning (Baines, 2002).
4.1
How to Improve the HRQOL in Kidney Transplant Recipients?
There are a number of open ended questions related to HRQOL. First, it is mandatory to achieve a universal accepted definition to better compare the QOL studies. Moreover, a kidney transplant may offer a long-term graft survival, so investigations should focus on such patients. It could be useful to determine the real impact on clinical outcome of HRQOL and further investigations should focus on those recipients not showing a beneficial HRQOL: in this group of patients there is increased incidence of side effects of the immunosuppressive drugs and post-transplant morbidities (infection, neoplasm, hypertension, cardiovascular disease). It may be hypothesized that a better HRQOL may improve some aspect of posttransplant functioning, such as the adherence to the treatment, finally resulting in an improvement of graft survival. HRQOL studies should focus on those groups of recipients who may theoretically have worst physical and psychological functioning. The increased use of expanded criteria donors
Quality of Life in Kidney Transplantation
130
to improve the donor pool, provides an opportunity of transplant to an increasing number of patients. However, patients receiving a kidney from an older donor or from a HCV-infected donor may expect a worse graft survival rate. In this group of patients an evaluation of HRQOL may validate the use of such donors. Finally, there is a clear need to use standardized methodologies for patient assessment by adopting a patient-centered intervention. This will lead to a better comprehension of patients’ care and improvement in the overall quality of life after kidney transplantation.
Summary Points Kidney transplantation is the best renal replacement therapy with a 1-year graft and patient survival that exceed 90%.
HRQOL is now considered as the best indicator of medical care in kidney transplant recipients.
HRQOL scores improves after kidney transplantation, except for physical domain, but kidney recipients scored lower than general population.
Psychological perception and issues related to sexuality, anxiety and even guilty may significantly affect the HRQOL.
Women, adolescent, unemployed recipients and cyclosporine-treated kidney recipients exhibit a worst HRQOL.
A better HRQOL may improve the clinical outcome of renal transplantation, by improving the adherence to the treatment.
A standardized, patient-centered medical and psychological support could improve the HRQOL in kidney transplant recipients.
References Aasebo W, Midtvedt K, Hartmann A, Stavem K. (2005). Clin Transplant. 19: 756–762. Akman B, Ozdemir FN, Sezer S, Micozkadioglu H, Haberal M. (2004). Transplant Proc. 36: 111–113. Baines LS, Joseph JT, Jindal RM. (2002). Clin Transplant. 6: 455–460. Bohlke M, Rocha M, Gomes RH, Marini SS, Ter Horst L, Barcellos FC, Hallal PC, Casarini D, Irigoyen MC. (2006). Clin Transplant. 20: 504–508. Cetingok M, Hathaway DK, Winsett RR. (2005). Prog Transplant. 15: 338–344. Chan L, Mulgaonkar S, Walker R, Arns W, Ambuhl P, Schiavelli R. (2006). Transplantation 81: 1290–1297. Christensen AJ, Raichle K, Ehlers SL, Bertolatus. (2002). Health Psychol. 21: 468–476. Cleemput I, Kesteloot K, De Geest S, Dobbels F, Vanrenterghem Y. (2003). Transplantation. 76: 176–182. Dew MA. (1998). J Psychosom Res. 44: 189–195. Dew MA, Switzer GE, Goycoolea JM, Allen AS, DiMartini A, Kormos, RL, Griffith BP. (1997). Transplantation. 64: 1261–1273.
Dobbels F, De Bleser L, De Geest S, Fine RN. (2007). Adv Chronic Kidney Dis. 4: 370–378. Ekberg H, Kyllonen L, Madsen S, Gisle G, Solbu D, Holdaas H. (2007). Transplantation. 83: 282–289. Fujisawa M, Ichikawa Y, Yoshya K, Isotani S, Higuchi A, Nagano S, Arakawa S, Hamami G, Matsumoto O, Kamidono S. (2000). Urology. 56: 201–206. Griva K, Ziegelmann JP, Thompson D, Jayasena D, Davenport A, Harrison M, Newman SP. (2002). Nephrol Dial Transplant. 17: 2204–2211. Gross CR, Kangas JR, Lemieux AM, Zehrer CL. (1995). Transplant Proc. 27: 3067–3068. Irwin M, McClintik J, Costlow C, Fortner M, White J, Gillin JC. (1996). FASEB J. 10: 643–653. Karam VH, Gasquet I, Delvart V, Hiesse C, Dorent R, Danet C, Samuel D, Charpentier B, Gandjbakhch I, Bismuth H, Castaing D. (2003). Transplantation. 12: 1699–1704. Kasiske BL, Cangro CB, Hariharan S, Hricik DE, Kerman RH, Roth D, Rush DN, Vazquez MA, Weir MR. (2002). Am J Transplant. 1(Suppl. 2): 1–95.
2237
2238
130
Quality of Life in Kidney Transplantation
Kauffman HM, Cherikh WS, McBride MA, Cheng Y, Hanto DW. (2006). Transplant Int. 19: 607–620. Kiley DJ, Lam CS, Pollak R. (1993). Transplantation. 55: 51–56. Kizilisik AT, Feurer ID, VanBuren DH, Wise P, Van Buren DV, Hopkins J, Ray J, Nylander W, Shaffer D, Heldermann JH, Langone AJ, Speroff T, Pinson CW. (2003). Am J Surg. 186: 535–539. Liem YS, Bosch JL, Arends LR, Heijenbrok-Kal MH, Hunink MGM. (2007). Value Health. 10: 390–397. Liu H, Feurer ID, Dwyer K, Speroff T, Shaffer D, Pinson CW. (2008). J Clin Nursing. 17: 82–89. Lumsdaine JA, Wray A, Power MJ, Jamieson NV, Akyol M, Bradley JA, Forsythe JLR, Wingmore SJ. (2005). Transplant Int. 18: 975–980. Metcalfe MS, Tweed A, White SA, Taylor R, Mullin E, Saunders RN, Waller JR, Wrigglesworth R, Nicholson ML. (2001). Transplant Proc. 33: 3403–3404. Moloney FJ, Keane S, O’Kelly P, Conlon PJ, Murphy GM. (2005). Br J Dermatol. 153: 574–578. Moons P, Vanrenterghem Y, Hoff JP, Squifflet JP, Margodt D, Mullens M, Thevissen I, De Geest S. (2003). Transplant Int. 16: 653–663. Muehrer RJ, Becker BN. (2005). Sem Dial. 18: 124–131. Neipp M, Karavul B, Jackobs, Vilsendorf AM, Richter N, Becker T, Schwarz A, Klempnauer J. (2006). Transplantation. 81: 1640–1644. Noohi S, Khaghani-Zadeh M, Javadipour M, Assarfi, S Najafi M, Ebrahiminia M, Poufarziani V. (2007). Transplant Proc. 39: 1074–1078. Oberbauer R, Hutchinson B, Eris J, Arias M, Claesson K, Mota A, Kreis H, Kleinman L, Wang F, Chen J, Revicki DA. (2003). Transplantation. 75: 1277–1285. Ojo AO. (2006). Transplantation. 82: 603–611. Overbeck I, Bartels M, Decker O, Harms J, Hauss J, Fangmann J. (2005). Transplant Proc. 37: 1618–1621. Painter PL, Topp KS, Krasnoff JB, Adey D, Strasner A, Tomlanovich S, Stock P. (2003). Kidney Int. 65: 2309–2316.
Parasuraman R, Yee J, Karthikeyan V, del Busto R. (2006). Adv Chronic Kidney Dis. 3: 280–294. Pinson CW, Feurer ID, Payne JL, Wise PE, Shockley S, Speroff T. (2000). Ann Surg. 232: 597–607. Pourmand G, Emamzadeh A, Moosavi S, Mehrsai A, Taherimahmoudi M, Nikoobakht M, Saraji A, Salem S. (2007). Transplant Proc. 39: 1029–1032. Rebollo P, Ortega F, Baltar JM, Badia X, Alvarez-Ude F, Diaz-Corte C, Naves M, Navascues RA, Urena A, Alvarez-Grande J. (2000). Clin Transplant. 14: 199–207. Reimer J, Franke GH, Philipp T, Heemann U. (2002). Clin Transplant. 16: 48–54. Rosenberger J, van Dijk JP, Nagyova I, Zezula I, Geckova AM, Roland R, van den Heuvel WJA, Groothoff JW. (2006). Transplantation. 81: 1306–1310. Sabbatini M, Crispo A, Pisani A, Gallo R, Cianciaruso B, Fuiano G, Federico S, Andreucci VE. (2005). Nephrol Dial Transplant. 20: 194–198. Slakey DP, Rosner M. (2007). Clin Transplant. 21: 224–228. Sundaram SS, Landgraf JM, Neighbors K, Cohn RA, Alonso EM. (2007). Am J Transplant. 7: 982–989. Tanriverdi N, Ozcurumez G, Colak T, et al. (2004). Transplant Proc. 36: 117–119. Valderrabano F, Jofre R, Lopez-Gomez JM. (2001). Am J Kidney Dis. 38: 443–464. Van Der Mei SF, Groothoff JW, van Sonderen ELP, van den Heuvel WJA, de Jong PE, van Son WJ. (2006). Transplantation. 82: 80–85. Virzı` A, Signorelli MS, Veroux M, Giammarresi G, Maugeri S, Nicoletti A, Veroux P. (2007). Transplant Proc. 39: 1791–1793. Wallace J, Yorgin PD, Carolan R, Moore H, Sanchez J, Belson A, Yorgin L, Major C, Granucci L, Alexander S, Arrington D. (2004). Pediatr Transplant. 8: 52–59.
131 Quality of Life in Liver Cirrhosis E. Kalaitzakis 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2240
2
Assessment of Quality of Life in Patients with Liver Cirrhosis . . . . . . . . . . . . . . . . . . 2241
3
Liver Cirrhosis and Health-Related QOL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2242
4
Etiology of Liver Cirrhosis and Health-Related QOL . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2242
5
The Impact of Severity and Complications of Liver Cirrhosis on Health-Related QOL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2243 5.1 Hepatic Encephalopathy and Impairment of Health-Related QOL . . . . . . . . . . . . . . . . 2244 5.2 Minimal Hepatic Encephalopathy and Fitness to Drive . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2246 6 The Impact of Cirrhosis-Related Symptoms on Health-Related QOL . . . . . . . . . . . 2247 6.1 Fatigue and Impairment of Health-Related QOL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2247 6.2 Gastrointestinal Symptoms and Impairment of Health-Related QOL . . . . . . . . . . . . . 2249 7
The Impact of Psychological Distress on Health-Related QOL in Liver Cirrhosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2249
8
The Impact of Medical Interventions and Liver Transplantation on Health-Related QOL in Liver Cirrhosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2250 8.1 Transjugular Intrahepatic Portosystemic Shunt (TIPS) and Health-Related QOL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2251 8.2 Liver Transplantation and Health-Related QOL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2251 9
Assessment of Utilities and Health-Related QOL in Liver Cirrhosis . . . . . . . . . . . . 2251
10
Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2252 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2252
#
Springer Science+Business Media LLC 2010 (USA)
2240
131
Quality of Life in Liver Cirrhosis
Abstract: Liver cirrhosis is a chronic condition imposing a considerable burden on patients, families, health care, and society. Assessment of health-related quality of life (QOL) is particularly important for these patients in view of the paucity of therapies substantially improving their survival except for liver transplantation. To date several investigators employing numerous generic and disease-specific QOL measurement instruments have reported poor QOL in these patients irrespective of the etiology of liver disease. Apart from the severity and complications of liver cirrhosis several disease-specific symptoms (such as pruritus, muscle cramps, sleep disturbance, sexual dysfunction, fatigue, and gastrointestinal symptoms) have been shown to be of importance in determining QOL in this patient group. Liver transplantation has been repeatedly shown to improve QOL. However, some of the few studies evaluating the impact of other medical interventions on QOL in cirrhotics have demonstrated that various treatment modalities may improve survival or decrease complication risk but they do not invariably improve patients’ QOL which stresses the need for rigorous selection of patients suitable for a specific type of treatment. Last, the value cirrhotic patients place on the state of their health differs from that assigned by physicians. Thus, in cost-utility analysis, which is based on quality-adjusted life years, utilities should be based on patient reports. List of Abbreviations: AUSQUAL, Austin quality of life scale; BDI, Beck depression inventory; BSI, brief symptom inventory; CC, compensated cirrhosis; CLDQ, chronic liver disease questionnaire; CLD-QOL, chronic liver disease quality of life questionnaire; DC, decompensated cirrhosis; GSRS, gastrointestinal symptom rating scale; HAD, hospital anxiety and depression index; LDQOL, liver disease quality of life instrument; LDSI, liver disease symptom index; LTX, liver transplantation; > MELD, model for end-stage liver disease; MFI-20, multidimensional fatigue index – 20; NC, non-cirrhotic control patients; NHP, Nottingham health profile; PGWBI, psychological general well-being index; QOL, quality of life; SF-36, 36-item short form health survey; SIP, sickness impact profile; TIPS, > transjugular intrahepatic portosystemic shunt
1
Introduction
Liver cirrhosis is defined histologically as a diffuse process with liver cell necrosis/apoptosis, fibrosis, and regenerative nodules. There are several causes of liver cirrhosis, the most common being high alcohol consumption, hepatitis C, hepatitis B, non-alcoholic steatohepatitis, primary biliary cirrhosis, primary sclerosing cholangitis, or autoimmune hepatitis (Sherlock and Dooley, 2002). Cirrhosis (apart from other features peculiar to the cause) results in two major events: hepatocellular failure and portal hypertension. Important complications of liver cirrhosis include, but are not limited to, esophageal varices, ascites, > hepatic encephalopathy, hepatic failure with jaundice, and hepatocellular cancer (Sherlock and Dooley, 2002). Cirrhosis has also been associated with varying degrees of malnutrition (Kalaitzakis et al., 2006; Sherlock and Dooley, 2002). Apart from liver transplantation, no specific cure exists for liver cirrhosis to date (> Table 131-1). Liver cirrhosis is a chronic condition imposing a considerable burden on families, health care, and society. Assessment of health-related quality of life (QOL), which is meant to give the patients’ perspective on the burden of disease, is particularly important for patients with liver cirrhosis because of the paucity of therapies substantially improving survival, other than liver transplantation. Hepatocellular failure and portal hypertension as well as their complications
Quality of Life in Liver Cirrhosis
131
. Table 131-1 Key facts of liver cirrhosis Liver cirrhosis is a diffuse histological process with liver cell necrosis/apoptosis and replacement of the hepatic parenchyma by fibrotic tissue and regenerative nodules Liver cirrhosis is mainly caused by high alcohol consumption, hepatitis C, hepatitis B, primary biliary cirrhosis, primary sclerosing cholangitis, autoimmune hepatitis, or non-alcoholic steatohepatitis Liver cirrhosis might lead to complications such as esophageal varices and variceal bleeding, ascites, hepatic failure with jaundice, hepatic encephalopathy, and hepatocellular cancer Apart from liver transplantation, no specific cure exists for liver cirrhosis to date The table summarizes key facts about definition, etiology, complications, and therapy of liver cirrhosis
and various symptoms (pruritus, fatigue, muscle cramps, gastrointestinal symptoms) that frequently accompany liver cirrhosis have, in addition to contributing to morbidity and mortality, adverse effects on the patients’ sense of well-being.
2
Assessment of Quality of Life in Patients with Liver Cirrhosis
Assessment of QOL is usually performed by using multi-item questionnaires which are completed by patients themselves thus reflecting their subjective experience of the impact of disease on daily activities and well-being (Borgaonkar and Irvine, 2000). Generic healthrelated QOL instruments may be used in any population irrespective of underlying disease, whereas disease-specific instruments are constructed for a particular disease. Combining generic and disease-specific instruments in usually recommended as it allows comparisons between diseases and within disease groups (Borgaonkar and Irvine, 2000). Generic and disease-specific instruments that have been used in different studies evaluating QOL in liver cirrhosis are shown in > Table 131-2. The most commonly used generic instrument is the 36-item short form health survey (SF-36), a 36-item self administered questionnaire encompassing eight physical and mental health domains and two physical and mental summary scales (Sullivan et al., 2002). The SF-36 has been thoroughly tested for validity and reliability in a variety of patient populations and it may be used to evaluate change in health status over time (Sullivan et al., 2002). To date, four health-related QOL instruments have been constructed for use in patients with chronic liver disease: the chronic liver disease questionnaire (CLDQ) (Younossi et al., 1999), the liver disease quality of life instrument in persons with advanced chronic liver disease (LDQOL) (Gralnek et al., 2000), the liver disease symptom index (LDSI) (van der Plas et al., 2004), and the chronic liver disease quality of life (CLD-QOL) questionnaire (Lee et al., 2008) (> Table 131-1). The CLDQ, LDSI, and LDQOL were developed in Western patient populations (Gralnek et al., 2000; van der Plas et al., 2004; Younossi et al., 1999) whereas the CLD-QOL was developed in an Asian patient population who live in different social and cultural environment compared to Western patients (Lee et al., 2008). Apart from their English versions, the CLDQ has been validated in German (Hauser et al., 2004b), Italian (Rucci et al., 2005), and Spanish-speaking patients (Ferrer et al., 2006) and the LDQOL in Spanish-speaking patients (Casanovas Taltavull et al., 2003).
2241
2242
131
Quality of Life in Liver Cirrhosis
. Table 131-2 Generic and disease-specific instruments used in different studies evaluating quality of life in liver cirrhosis Generic instruments Assessment of health-related quality of life The 36-item short-form health survey (SF-36) (Hauser et al., 2004a; Marchesini et al., 2001; Younossi et al., 2001b) Nottingham health profile (NHP) (Bianchi et al., 2003; Marchesini et al., 2001) Sickness impact profile (SIP) (Groeneweg et al., 1998; Prasad et al., 2007; Tarter et al., 1992) Austin quality of life scale (AUSQUAL) (Moore et al., 2000) Quality of life index (Gulberg et al., 2002) Assessment of mental health Brief symptom inventory (BSI) (De Bona et al., 2000) Beck Depression Inventory (BDI) (Bianchi et al., 2005; Singh et al., 1997) Hospital anxiety and depression (HAD) (Hauser et al., 2004a) Psychological general well-being index (PGWBI) (Bianchi et al., 2005) Assessment of gastrointestinal symptoms Gastrointestinal symptom rating scale (GSRS) (Kalaitzakis et al., 2006) Assessment of fatigue Multidimensional fatigue index-20 (MFI-20) (van der Plas et al., 2003) Assessment of sexual function International index of erectile function (Toda et al., 2005) Liver disease-specific instruments Chronic liver disease questionnaire (CLDQ) (Younossi et al., 1999) Liver disease quality of life instrument in chronic liver disease (LDQOL) (Gralnek et al., 2000) Liver disease symptom index (LDSI) (van der Plas et al., 2004) Chronic liver disease quality of life (CLD-QOL) questionnaire (Lee et al., 2008) A wide variety of generic and disease-specific questionnaires have been used in studies of health-related quality of life in liver cirrhosis. Relevant references provided in the table are indicative and not exhaustive
3
Liver Cirrhosis and Health-Related QOL
Several studies have shown that both physical and mental dimensions of health-related quality of life are affected in patients with chronic liver disease in general and liver cirrhosis in particular (Bianchi et al., 2005; Hauser et al., 2004a; Kalaitzakis et al., 2006; Marchesini et al., 2001; Singh et al., 1997; van der Plas et al., 2003; Younossi et al., 2001b) (> Figure 131-1).
4
Etiology of Liver Cirrhosis and Health-Related QOL
Most studies have not detected any difference in health-related QOL indices among patients with liver cirrhosis of different etiologies (Hauser et al., 2004a; Kalaitzakis et al., 2006; Marchesini et al., 2001). However, some controversy exists with some authors reporting that
Quality of Life in Liver Cirrhosis
131
. Figure 131-1 Health-related quality of life assessed as SF-36 domain and summary scores (means and 95% confidence intervals) in patients with liver cirrhosis (continuous line, n = 128) and healthy controls (dashed line, n = 299) (Kalaitzakis et al., 2006) (reproduced with permission from the publisher Taylor & Francis). Lower scores indicate poorer health-related quality of life. Patients with liver cirrhosis showed poorer health-related quality of life assessed by the 36-item shortform health survey (SF-36) compared to controls from the general population. All SF-36 indices and summary scores were significantly lower in patients with cirrhosis compared to controls except that for bodily pain. SF-36, the 36-item short form health survey; PF, physical functioning; RP, role limitations caused by physical health problems; BP, bodily pain; GH, general health perceptions; VT, vitality; SF, social functioning; RE, role limitations caused by emotional problems; MH, mental health; PCS, physical component summary; MCS, mental component summary. The PCS and MCS are summaries of the physical and mental SF-36 indices
both physical and mental dimensions of QOL are less impaired in patients with cholestatic disease than in those with hepatocellular disease (Younossi et al., 2001b). Also, chronic hepatitis C has been reported to have a negative impact on health-related QOL of end-stage liver disease patients (Kanwal et al., 2004) but patients with hepatitis C cirrhosis have not been found to differ in health-related QOL from patients with cirrhosis of other etiologies (Bjo¨rnsson and Kalaitzakis, unpublished data). Health-related QOL in hepatitis C is dealt with elsewhere in this book. Patients with alcoholic cirrhosis generally do not differ from patients with other cirrhosis etiologies but those with active alcohol abuse have been reported to have increased psychological distress and increased depression scores compared to abstainers (Bianchi et al., 2005).
5
The Impact of Severity and Complications of Liver Cirrhosis on Health-Related QOL
Factors that have been reported to be associated with impairment of QOL in these patients include cirrhosis severity (Bianchi et al., 2005; Kalaitzakis et al., 2006; Marchesini et al., 2001; van der Plas et al., 2003; Younossi et al., 2001b), complications of portal hypertension such as hepatic encephalopathy (Arguedas et al., 2003; Marchesini et al., 2001), ascites
2243
2244
131
Quality of Life in Liver Cirrhosis
(Marchesini et al., 2001), hepatocellular carcinoma (Bianchi et al., 2003), daily medications especially loop diuretics (Marchesini et al., 2001), comorbidities (Marchesini et al., 2001), and nutritional status impairment (Norman et al., 2006). A recent large study, in which patients with chronic liver disease (n = 489), > compensated liver cirrhosis (n = 391), > decompensated liver cirrhosis (n = 84), and transplanted patients (n = 186) were included, showed that all patients with chronic liver disease, including compensated and decompensated cirrhotics, have compromised health-related QOL compared to controls from the general population (van der Plas et al., 2003). However, when compared to patients with chronic liver disease without cirrhosis QOL was impaired mainly in patients with decompensated cirrhosis (> Figure 131-2). Furthermore, severity of liver cirrhosis expressed as the Child-Pugh and the > MELD (model for end-stage liver disease) score has been reported to be related to health-related QOL indices (Kalaitzakis et al., 2006; Kanwal et al., 2004; Saab et al., 2005). Interestingly the > Child-Pugh score has consistently been shown to be more strongly associated to health-related QOL than the MELD score (Kalaitzakis et al., 2006; Kanwal et al., 2004; Saab et al., 2005). This may be due to that the Child-Pugh score encompasses ascites and hepatic encephalopathy which have been shown to be independent predictors of QOL in cirrhosis (Arguedas et al., 2003; Marchesini et al., 2001) apart from biochemical parameters whereas the MELD score is derived solely from biochemical parameters.
5.1
Hepatic Encephalopathy and Impairment of Health-Related QOL
Patients with liver cirrhosis are prone to develop cognitive dysfunction termed hepatic encephalopathy. Clinical manifestations of hepatic encephalopathy range from subtle intellectual and personality changes to coma (Sherlock and Dooley, 2002). Hepatic encephalopathy is diagnosed according to certain clinical criteria. However, it is well-recognized that even patients without clinically overt hepatic encephalopathy may have subtle cognitive dysfunction identifiable only by means of psychometric tests (> minimal hepatic encephalopathy) (Sherlock and Dooley, 2002). Tarter and colleagues applied the sickness impact profile (SIP) and a psychometric test battery to 130 nonalcoholic patients whom they examined before and 3 years after liver transplantation (Tarter et al., 1992). They found that there was a substantial improvement from the pretransplant to the posttransplant periods across almost all dimensions of QOL. Psychometric test scores explained up to 20% of the variance in magnitude of change from pre- to post- surgery. Thus the investigators concluded that severity of hepatic encephalopathy is associated with posttransplantat improvement in QOL (Tarter et al., 1992). Hepatic encephalopathy has also been shown to be independently related to the physical functioning and role-emotional SF-36 domains in a large multicenter study in which 544 patients with cirrhosis were enrolled (Marchesini et al., 2001). In another study investigating the possible effect of hepatic encephalopathy on health-related QOL 160 consecutive patients with liver cirrhosis undergoing pretransplantation evaluation were included (Arguedas et al., 2003). Hepatic encephalopathy was assessed clinically as well as by means of a psychometric test and QOL was assessed by means of the SF-36 (Arguedas et al., 2003). Patients with hepatic encephalopathy (overt or minimal) had decreased physical and mental component summary scores compared to patients without encephalopathy (> Figure 131-3). Minimal hepatic encephalopathy has been shown to impair daily functioning as assessed by means of the sickness impact profile in 179 outpatients with liver cirrhosis (Groeneweg et al., 1998). One study by Schomerus et al investigated the potential role of hepatic encephalopathy
Quality of Life in Liver Cirrhosis
131
. Figure 131-2 Health-related quality of life assessed as SF-36 domain score differences (means and 95% confidence intervals) between non-cirrhotic patients with chronic liver disease (n = 489) (NC), and patients with compensated liver cirrhosis (n = 391) (CC), decompensated liver cirrhosis (n = 84) (DC), and transplanted patients (n = 186) (LTX). Differences are adjusted for gender, age, educational level, etiology, use of liver disease medication, use of psychopharmaca, and comorbidity (reproduced from van der Plas et al. (2003) with permission). Positive values indicate higher (better) and negative values poorer health-related quality of life SF-36 scores compared to controls (NC, non-cirrhotic patients with chronic liver disease). This figure illustrates that health-related quality of life is impaired mainly in patients with decompensated and not compensated cirrhosis compared to patients with chronic liver disease without cirrhosis (NC). Patients having received a liver transplant have better health-related quality of life in most dimensions compared to NC. * Scale score of subgroup is significantly different (p < 0.05) from scale score of controls (NC). SF-36, the 36-item short form health survey; NC, non-cirrhotic patients with chronic liver disease; CC, compensated cirrhosis; DC, decompensated cirrhosis; LTX, liver transplantation; PF, physical functioning; RP, role limitations caused by physical health problems; BP, bodily pain; GH, general health perceptions; VT, vitality; SF, social functioning; RE, role limitations caused by emotional problems; MH, mental health. @ 2003 van der Plas et al; licensee BioMed Central Ltd. Open Access Article: verbatim copying and redistribution of this article are permitted in all media for any purpose, provided this notice is preserved along with the article’s original URL (http://www.biomedcentral.com/1471–230X/3/33)
in working capacity in patients with cirrhosis (Schomerus and Hamster, 2001). A total of 110 outpatients with liver cirrhosis who were not willingly unemployed were enrolled in the study and underwent extensive psychometric testing. Forty-four percent were receiving disability pension. The authors found that although the working group and the group of patients on
2245
2246
131
Quality of Life in Liver Cirrhosis
. Figure 131-3 Health-related quality of life assessed as mean SF-36 domain score and summary scores in patients with liver cirrhosis without hepatic encephalopathy (dashed line, n = 23), with minimal hepatic encephalopathy (dotted line, n = 36), and with overt hepatic encephalopathy (continuous line, n = 89). Lower scores indicate poorer health-related quality of life. Patients with hepatic encephalopathy (overt and minimal) showed poorer health-related quality of life assessed by the 36-item short-form health survey (SF-36) compared to cirrhotic patients without hepatic encephalopathy. * Statistical significance by analysis of variance. The Child-Pugh score did not differ significantly among groups (p > 0.05). SF-36, the 36-item short form health survey; PF, physical functioning; RP, role limitations caused by physical health problems; BP, bodily pain; GH, general health perceptions; VT, vitality; SF, social functioning; RE, role limitations caused by emotional problems; MH, mental health; PCS, physical component summary; MCS, mental component summary. The PCS and MCS constitute summaries of the physical and mental SF-36 indices. Data from Arguedas et al. (2003)
disability pension did not differ in severity or complications of liver cirrhosis (including overt hepatic encephalopathy) the latter group scored worse in psychometric tests evaluating psychomotor function and personality (Schomerus and Hamster, 2001). It was concluded that minimal hepatic encephalopathy might be implicated in impaired working capability of patients with liver cirrhosis (Schomerus and Hamster, 2001).
5.2
Minimal Hepatic Encephalopathy and Fitness to Drive
Published data suggest that minimal hepatic encephalopathy impairs fitness to drive (Bajaj et al., 2008; Wein et al., 2004). A total of 48 patients with liver cirrhosis (34 with and 14 without minimal hepatic encephalopathy according to psychometric testing) were enrolled in a prospective study evaluating their ability to drive a car by means of a standardized on-road driving test (Wein et al., 2004). Patients with compared to those without minimal hepatic encephalopathy required a higher number of interventions by driving instructors to prevent
Quality of Life in Liver Cirrhosis
131
accidents (Wein et al., 2004). Patients with minimal hepatic encephalopathy have also been reported to have impaired navigations skills on a driving simulator which is correlated with impairment in response inhibition and attention (Bajaj et al., 2008). In view of these data it is not surprising that a study investigating the reported occurrence of traffic violations and motor vehicle accidents by means of an anonymous driving history and behavior questionnaire which was sent to 200 cirrhotics without overt hepatic encephalopathy and 100 age/ education matched controls found that cirrhotics have a higher self-reported occurrence of violations and accidents compared to controls (Bajaj et al., 2007). In the same study minimal hepatic encephalopathy was the only risk factor (odds ratios: 4.2:7.6) for violations and accidents selected in multivariate analysis (Bajaj et al., 2007).
6
The Impact of Cirrhosis-Related Symptoms on Health-Related QOL
Non-life threatening subjective symptoms of cirrhosis, frequently underscored by physicians, have been reported to be of concern in patients with liver cirrhosis. Thus, disease-specific instruments for assessment of health-related QOL include several symptoms affecting these patients (Gralnek et al., 2000; Lee et al., 2008; van der Plas et al., 2004; Younossi et al., 1999). In a multicenter study assessing QOL in 544 patients with liver cirrhosis, pruritus affected 26% and muscle cramps affected 36% of included patients and were found to be more closely associated with poor QOL than major, even life-threatening events (Marchesini et al., 2001). Sleep disturbance is also common in cirrhosis affecting 48% of these patients and it was found to be unrelated to clinical parameters and cognitive impairment in one study on 44 cirrhotic patients without clinically overt hepatic encephalopathy (Cordoba et al., 1998). Sexual dysfunction is another common complaint among cirrhotics and impotence affects 70% of alcoholic patients with cirrhosis and 25% of non-alcoholic cirrhotic patients (p < 0.05) (Cornely et al., 1984). In another study in which 53 male patients with liver cirrhosis were included erectile dysfunction was found to affect 92% of patients (Toda et al., 2005). To date no published study has particularly evaluated sexual dysfunction and its impact on dimensions of health-related QOL in both genders.
6.1
Fatigue and Impairment of Health-Related QOL
Fatigue is another common complaint in patients with chronic liver disease. It has been the object of extensive investigation in patients with primary biliary cirrhosis in whom it is thought to have a negative effect on health-related QOL (Goldblatt et al., 2002; Huet et al., 2000) although fatigue has not been found to be a specific symptom of primary biliary cirrhosis in all studies (Bjo¨rnsson et al., 2005). However, few of the patients included in these studies had frank liver cirrhosis and published data on fatigue in cirrhosis of other etiologies are scarce. A recent study performed in patients with chronic liver disease (n = 489), compensated liver cirrhosis (n = 391), decompensated liver cirrhosis (n = 84), and transplanted patients (n = 186) showed that all patients with chronic liver disease, including cirrhotic and transplanted patients, have increased fatigue indices compared to controls from the general population (van der Plas et al., 2003). However, when compared to patients with chronic liver disease without cirrhosis fatigue severity was increased mainly in
2247
2248
131
Quality of Life in Liver Cirrhosis
. Figure 131-4 Fatigue assessed as multidimensional fatigue impact-20 scale score differences (means and 95% confidence intervals) between non-cirrhotic patients with chronic liver disease (n = 489) (NC), and patients with compensated liver cirrhosis (n = 391) (CC), decompensated liver cirrhosis (n = 84) (DC), and transplanted liver patients (n = 186) (LTX). Differences are adjusted for gender, age, educational level, etiology, use of liver disease medication, use of psychopharmaca, and comorbidity (reproduced from van der Plas et al. (2003) with permission). Positive values indicate increased and negative values decreased fatigue scores compared to controls (NC, non-cirrhotic patients with chronic liver disease). This figure illustrates that fatigue is increased mainly in patients with decompensated and not compensated cirrhosis compared to patients with chronic liver disease without cirrhosis (NC). Patients having received a liver transplant have decreased fatigue scores compared to NC.* Scale score of subgroup is significantly different (p < 0.05) from scale score of controls (NC). NC, non-cirrhotic patients with chronic liver disease; CC, compensated cirrhosis; DC, decompensated cirrhosis; LTX, liver transplantation; GF, general fatigue; PHF, physical fatigue; RA, reduction in activity; RM, reduction in motivation; MF, mental fatigue. @ 2003 van der Plas et al; licensee BioMed Central Ltd. Open Access Article: verbatim copying and redistribution of this article are permitted in all media for any purpose, provided this notice is preserved along with the article’s original URL (http://www.biomedcentral.com/ 1471–230X/3/33)
patients with decompensated and not compensated cirrhosis (> Figure 131-4). In another recent study in which 83 consecutive patients with liver cirrhosis undergoing pretransplantation evaluation were included, fatigue (assessed by means of the fatigue impact scale) was found to be increased in cirrhotics compared to controls from the general population and it was also associated with impaired QOL in these patients (Kalaitzakis and Bjo¨rnsson, 2007). Mood disorders and low hemoglobin levels were reported to contribute to fatigue in cirrhotics
Quality of Life in Liver Cirrhosis
131
. Figure 131-5 Gastrointestinal symptom severity assessed as gastrointestinal symptom rating scale (GSRS) scores (means and 95% confidence intervals) in patients with liver cirrhosis (continuous line, n = 128) and healthy controls (dashed line, n = 2,162) (Kalaitzakis et al., 2006) (reproduced with permission from the publisher Taylor & Francis). Higher scores indicated increased gastrointestinal symptom severity. Patients with liver cirrhosis show increased severity of gastrointestinal symptoms as assessed by the gastrointestinal symptom rating scale (GSRS) compared to controls from the general population. GSRS, gastrointestinal symptom rating scale
in general whereas in male patients low testosterone was related to increased fatigue severity indices (Kalaitzakis and Bjo¨rnsson, 2007).
6.2
Gastrointestinal Symptoms and Impairment of Health-Related QOL
Gastrointestinal symptoms are also of concern in patients with liver cirrhosis. In one study (Kalaitzakis et al., 2006), gastrointestinal symptom severity was found to be increased in patients with liver cirrhosis compared to the general population (> Figure 131-5). Recent weight loss as well as physical and mental dimensions of health-related QOL were independently related to gastrointestinal symptoms (Kalaitzakis et al., 2006).
7
The Impact of Psychological Distress on Health-Related QOL in Liver Cirrhosis
Psychological distress and depression are common in patients with liver cirrhosis (Singh et al., 1997). In one study, the impact of depression on QOL and outcome was investigated in patients with liver cirrhosis undergoing pretransplantation evaluation (Singh et al., 1997). A total of 81 patients were assessed by means of the Beck depression inventory (BDI) and 64% were determined to be depressed according to the Beck depression scores. Depressed patients
2249
2250
131
Quality of Life in Liver Cirrhosis
did not differ from those without depression in any demographic variable, severity or complications of liver cirrhosis, or survival following liver transplantation. However, the depressed group had significantly poorer perceived QOL compared to non-depressed patients. Among patients not receiving a transplant depressed patients had shorter survival compared to those without depression (Singh et al., 1997). In another study the BDI and the psychological general well-being index (PGWBI) were applied to 156 consecutive patients with liver cirrhosis (Bianchi et al., 2005). Patients with cirrhosis showed significantly lower global and domain PGWBI scores compared to a reference population indicating psychological distress and poor sense of well-being in this group of patients (Bianchi et al., 2005). Also, 57% of cirrhotics had BDI scores suggestive of depression. Logistic regression analysis identified the presence of sleep disorders as the independent variable more frequently associated with low domains PGWBI scores and the severity of liver cirrhosis, expressed as the Child-Pugh score, as the sole variable related to depression as detected by the BDI (Bianchi et al., 2005). The etiology of liver disease was not found to be related to psychological distress or depression in this study. However, within the subgroup of patients with alcoholic cirrhosis active drinkers had higher BDI scores compared to abstainers (Bianchi et al., 2005). Thus, patients with liver cirrhosis have signs of psychological distress and depression which affects health-related quality of life. Treatment of depression in cirrhosis seems to be important.
8
The Impact of Medical Interventions and Liver Transplantation on Health-Related QOL in Liver Cirrhosis
Patients with liver cirrhosis frequently undergo medical interventions aiming to reduce morbidity and in certain cases improve survival. Apart from medical therapy, such as diuretics for ascites, beta-blockers as prophylaxis for variceal bleeding, and lactulose for hepatic encephalopathy, they may undergo paracentesis or receive transjugular intrahepatic portosystemic shunt (TIPS) for ascites. Furthermore, they may undergo liver transplantation. Some previous studies have evaluated the effect of some of these interventions on health-related QOL in this patients group. The number of daily medications as well as specific drugs have been implicated in the poor health-related QOL in patients with liver cirrhosis (Cordoba et al., 2003; Kalaitzakis et al., 2006; Marchesini et al., 2001). The number of daily medications and loop diuretics have been shown to affect QOL in cirrhotics of various etiologies (Marchesini et al., 2001) whereas betablockers and diuretics appear to have an important effect on QOL in cirrhotic outpatients with hepatitis C and prior decompensations (Cordoba et al., 2003). Also, daily lactulose use has been reported to be independently related to gastrointestinal symptom severity in patients with cirrhosis (Kalaitzakis et al., 2006). It is conceivable that medications may have been a surrogate marker for more severe liver cirrhosis with several complications in these studies but it cannot be excluded that certain side-effects of medications used might contribute to healthrelated QOL impairment in these patients. On the other hand, a recent randomized prospective study evaluating the effect of lactulose treatment on minimal hepatic encephalopathy and QOL in patients with cirrhosis showed that lactulose improves both cognitive function and QOL (Prasad et al., 2007). Improvement in health-related QOL in this study was related to the improvement in cognitive function as assessed by psychometric tests (Prasad et al., 2007). Furthermore, in patients with chronic hepatitis C and advanced fibrosis or cirrhosis
Quality of Life in Liver Cirrhosis
131
achievement of sustained virologic response after pegylated interferon and ribavirin therapy has been reported to improve health-related QOL and sexual health (Bonkovsky et al., 2007).
8.1
Transjugular Intrahepatic Portosystemic Shunt (TIPS) and Health-Related QOL
TIPS is often used as an alternative to medical therapy and large-volume paracentesis in patient with cirrhosis and refractory ascites (Sherlock and Dooley, 2002). A successful TIPS may minimize the requirement for diuretics and the need for large-volume paracentesis with patients experiencing symptom relief (e.g., less shortness of breath and early satiety). However, TIPS may fail to resolve ascites in up to 50% of cases and patients may experience deterioration in QOL mainly due to development of hepatic encephalopathy. A study, in which 21 cirrhotic patients receiving TIPS due to ascites were included, evaluated QOL prior and post TIPS in a cross-over manner (Gulberg et al., 2002). QOL showed significant improvement post TIPS which was more pronounced in patients with complete response to therapy (Gulberg et al., 2002). A randomized study comparing the effect of TIPS versus medical therapy on QOL in patients with cirrhosis and refractory ascites, however, showed that the two therapeutic modalities led to similar changes in QOL (Campbell et al., 2005). Competing effects of hepatic encephalopathy post TIPS and of requirement for repeated large-volume paracentesis and hospitalizations in the medical therapy group probably account for the similar changes in health-related QOL (Campbell et al., 2005). These data indicate that physicians responsible for the care of patients with liver cirrhosis ought to apply rigorous criteria for selection of patients suitable for specific types of medical treatment.
8.2
Liver Transplantation and Health-Related QOL
Liver transplantation, the only curative treatment for liver cirrhosis, has been repeatedly shown to improve not only survival but also health-related QOL (Belle et al., 1997; De Bona et al., 2000; Gross et al., 1999; Karam et al., 2003; Moore et al., 2000; Sherlock and Dooley, 2002; Tarter et al., 1991). Transplanted patients have been shown to have higher health-related QOL indices compared to patients with compared to patients with chronic liver disease (> Figures 131-2 and > 131-4). Specifically both physical and mental dimensions of QOL have been reported to improve posttransplant (Belle et al., 1997; De Bona et al., 2000; Gross et al., 1999; Karam et al., 2003; Moore et al., 2000) including anxiety and depression (De Bona et al., 2000; Gross et al., 1999; Moore et al., 2000), cognitive function (Moore et al., 2000), and disease-related symptoms (Belle et al., 1997; Gross et al., 1999; Karam et al., 2003; Moore et al., 2000). However, when compared to the general population not all studies have shown a complete return of health-related QOL to normal status at least as far as certain QOL dimensions are concerned (Karam et al., 2003; Tarter et al., 1991).
9
Assessment of Utilities and Health-Related QOL in Liver Cirrhosis
Cost-utility analysis is an important approach to economic analysis and is based on qualityadjusted life years. It is important to appropriately define utilities for a particular
2251
2252
131
Quality of Life in Liver Cirrhosis
disease before any meaningful cost-utility analysis can be performed. Economic analyses of liver disease have traditionally relied on utility estimates from experts which may, however, differ from patients’ direct experience and thus may be flawed. In one study, in which 120 patients with chronic liver disease (51% with cirrhosis) were enrolled, the validity of a widely used utility measure (Health Utility Index-2) was established and the decrement in health-related QOL associated with chronic liver disease was measured (Younossi et al., 2001a). Patients without cirrhosis and those with Child-Pugh A cirrhosis showed substantial decrement in utilities (0.82 and 0.83, respectively) in the range of patients surviving brain tumor. Those with Child-Pugh B and C cirrhosis showed greater decrement (0.67 and 0.56) that was in the range experienced by patients who survive a stroke (Younossi et al., 2001a). Another study compared physician-assigned and patientassigned utilities for six clinical scenarios in cirrhosis (1) compensated cirrhosis, (2) decompensated cirrhosis, (3) hepatic encephalopathy, (4) spontaneous bacterial peritonitis, (5) variceal bleeding, and (6) hepatocellular cancer (Wells et al., 2004). Although physicians and patients assigned similar rankings to each health state, physician-assigned utilities were significantly different from those assigned by patients (Wells et al., 2004). The authors of both studies conclude that utilities should be based on patient reports (Wells et al., 2004; Younossi et al., 2001a).
10
Conclusions
Patients with liver cirrhosis have poor health-related QOL which is related not only to severity and complications of cirrhosis but also to cirrhosis-specific symptoms. Liver transplantation improves QOL but other medical interventions (aimed to improve survival and/or reduce morbidity) in cirrhotics have been demonstrated to improve or compromise patients’ QOL. This suggests that there is a need for selection of patients most suitable for a specific type of treatment. Last, physician and patient-assigned utilities have been shown to differ significantly. Thus, in cost-utility analysis utilities should be based on patient reports.
Summary Points Liver cirrhosis is associated with poor health-related QOL. Severity and complications of liver cirrhosis (especially hepatic encephalopathy) are asso
ciated with poor health-related QOL whereas etiology of liver cirrhosis is not of major importance for QOL. Disease-specific symptoms such as pruritus, muscle cramps, sleep disturbance, fatigue and gastrointestinal symptoms have a negative impact on health-related QOL in patients with liver cirrhosis. Liver transplantation, the only curative treatment in liver cirrhosis, results in improvement of health-related QOL. Medical treatments aiming to improve survival or reduce mortality in patients with cirrhosis may not invariably improve health-related QOL which stresses the need for rigorous selection of patients suitable for specific therapies. Utilities for cost-utility analysis should be based on patient reports.
Quality of Life in Liver Cirrhosis
131
References Arguedas MR, DeLawrence TG, McGuire BM. (2003). Dig Dis Sci. 48: 1622–1626. Bajaj JS, Hafeezullah M, Hoffmann RG, Saeian K. (2007). Am J Gastroenterol. 102: 1903–1909. Bajaj JS, Hafeezullah M, Hoffmann RG, Varma RR, Franco J, Binion DG, Hammeke TA, Saeian K. (2008). Hepatology. 47: 596–604. Belle SH, Porayko MK, Hoofnagle JH, Lake JR, Zetterman RK. (1997). Liver Transpl Surg. 3: 93–104. Bianchi G, Loguercio C, Sgarbi D, Abbiati R, Brunetti N, De Simone T, Zoli M, Marchesini G. (2003). Dig Liver Dis. 35: 46–54. Bianchi G, Marchesini G, Nicolino F, Graziani R, Sgarbi D, Loguercio C, Abbiati R, Zoli M. (2005). Dig Liver Dis. 37: 593–600. Bjo¨rnsson E, Simren M, Olsson R, Chapman RW. (2005). Eur J Gastroenterol Hepatol. 17: 351–357. Bonkovsky HL, Snow KK, Malet PF, Back-Madruga C, Fontana RJ, Sterling RK, Kulig CC, Di Bisceglie AM, Morgan TR, Dienstag JL, Ghany MG, Gretch DR. (2007). J Hepatol. 46: 420–431. Borgaonkar MR, Irvine EJ. (2000). Gut. 47: 444–454. Campbell MS, Brensinger CM, Sanyal AJ, Gennings C, Wong F, Kowdley KV, McCashland T, Reddy KR. (2005). Hepatology. 42: 635–640. Casanovas Taltavull T, Vallejo Blanxart G, Herdman M, Verge Monedero JM, Tremosa Llurba G, Rodriguez Farina E, Ramos Rubio E, Baliellas Comellas C, Figueras Felip J, Menchon Magrina JM, Casais Alvarez LA. (2003). Gastroenterol Hepatol. 26: 234–244. Cordoba J, Cabrera J, Lataif L, Penev P, Zee P, Blei AT. (1998). Hepatology. 27: 339–345. Cordoba J, Flavia M, Jacas C, Sauleda S, Esteban JI, Vargas V, Esteban R, Guardia J. (2003). J Hepatol. 39: 231–238. Cornely CM, Schade RR, Van Thiel DH, Gavaler JS. (1984). Hepatology. 4: 1227–1230. De Bona M, Ponton P, Ermani M, Iemmolo RM, Feltrin A, Boccagni P, Gerunda G, Naccarato R, Rupolo G, Burra P. (2000). J Hepatol. 33: 609–615. Ferrer M, Cordoba J, Garin O, Olive G, Flavia M, Vargas V, Esteban R, Alonso J. (2006). Liver Transpl. 12: 95–104. Goldblatt J, Taylor PJ, Lipman T, Prince MI, Baragiotta A, Bassendine MF, James OF, Jones DE. (2002). Gastroenterology. 122: 1235–1241. Gralnek IM, Hays RD, Kilbourne A, Rosen HR, Keeffe EB, Artinian L, Kim S, Lazarovici D, Jensen DM, Busuttil RW, Martin P. (2000). Am J Gastroenterol. 95: 3552–3565. Groeneweg M, Quero JC, De Bruijn I, Hartmann IJ, Essink-bot ML, Hop WC, Schalm SW. (1998). Hepatology. 28: 45–49.
Gross CR, Malinchoc M, Kim WR, Evans RW, Wiesner RH, Petz JL, Crippin JS, Klintmalm GB, Levy MF, Ricci P, Therneau TM, Dickson ER. (1999). Hepatology. 29: 356–364. Gulberg V, Liss I, Bilzer M, Waggershauser T, Reiser M, Gerbes AL. (2002). Digestion. 66: 127–130. Hauser W, Holtmann G, Grandt D. (2004a). Clin Gastroenterol Hepatol. 2: 157–163. Hauser W, Schnur M, Steder-Neukamm U, Muthny FA, Grandt D. (2004b). Eur J Gastroenterol Hepatol. 16: 599–606. Huet PM, Deslauriers J, Tran A, Faucher C, Charbonneau J. (2000). Am J Gastroenterol. 95: 760–767. Kalaitzakis E, Bjo¨rnsson E. (2007). Gastroenterology. 132: A475. Kalaitzakis E, Simren M, Olsson R, Henfridsson P, Hugosson I, Bengtsson M, Bjornsson E. (2006). Scand J Gastroenterol. 41: 1464–1472. Kanwal F, Hays RD, Kilbourne AM, Dulai GS, Gralnek IM. (2004). Am J Gastroenterol. 99: 1726–1732. Karam V, Castaing D, Danet C, Delvart V, Gasquet I, Adam R, Azoulay D, Samuel D, Bismuth H. (2003). Liver Transpl. 9: 703–711. Lee EH, Cheong JY, Cho SW, Hahm KB, Kim HY, Park JJ, Lee DH, Kim SK, Choi SR, Lee ST, Moon SM. (2008). J Gastroenterol Hepatol. 23: 231–238. Marchesini G, Bianchi G, Amodio P, Salerno F, Merli M, Panella C, Loguercio C, Apolone G, Niero M, Abbiati R. (2001). Gastroenterology. 120: 170–178. Moore KA, Mc LJR, Burrows GD. (2000). Liver Transpl. 6: 633–642. Norman K, Kirchner H, Lochs H, Pirlich M. (2006). World J Gastroenterol. 12: 3380–3385. Prasad S, Dhiman RK, Duseja A, Chawla YK, Sharma A, Agarwal R. (2007). Hepatology. 45: 549–559. Rucci P, Taliani G, Cirrincione L, Alberti A, Bartolozzi D, Caporaso N, Colombo M, Coppola R, Chiaramonte M, Craxi A, De Sio I, Floreani AR, Gaeta GB, Persico M, Secchi G, Versace I, Mele A. (2005). Dig Liver Dis. 37: 850–860. Saab S, Ibrahim AB, Shpaner A, Younossi ZM, Lee C, Durazo F, Han S, Esrason K, Wu V, Hiatt J, Farmer DG, Ghobrial RM, Holt C, Yersiz H, Goldstein LI, Tong MJ, Busuttil RW. (2005). Liver Transpl. 11: 218–223. Schomerus H, Hamster W. (2001). Metab Brain Dis. 16: 37–41. Sherlock S, Dooley J. (2002). Sherlock S, Dooley J (ed.) “Diseases of the liver and biliary system”. Blackwell Publishing, Milan, pp. 365–380. Singh N, Gayowski T, Wagener MM, Marino IR. (1997). Dig Dis Sci. 42: 1421–1427.
2253
2254
131
Quality of Life in Liver Cirrhosis
Sullivan M, Karlsson J, Taft C, Ware JE. (2002). “SF-36 Health Survey: Swedish Manual and Interpretation Guide, 2nd ed.” Sahlgrenska University Hospital, Gothenburg. Tarter RE, Switala J, Arria A, Plail J, Van Thiel D. (1991). Arch Intern Med. 151: 1521–1526. Tarter RE, Switala J, Plail J, Havrilla J, Van Thiel DH. (1992). Arch Intern Med. 152: 2097–2101. Toda K, Miwa Y, Kuriyama S, Fukushima H, Shiraki M, Murakami N, Shimazaki M, Ito Y, Nakamura T, Sugihara J, Tomita E, Nagata C, Suzuki K, Moriwaki H. (2005). J Gastroenterol. 40: 894–900. Plas SM, Hansen BE, de Boer JB, Stijnen T, Passchier J, de Man RA, Schalm SW. (2003). BMC Gastroenterol. 3: 33. http://www.biomedcentral.com/1471–1230X/ 1473/1433.
van der Plas SM, Hansen BE, de Boer JB, Stijnen T, Passchier J, de Man RA, Schalm SW. (2004). Qual Life Res. 13: 1469–1481. Wein C, Koch H, Popp B, Oehler G, Schauder P. (2004). Hepatology. 39: 739–745. Wells CD, Murrill WB, Arguedas MR. (2004). Dig Dis Sci. 49: 453–458. Younossi ZM, Boparai N, McCormick M, Price LL, Guyatt G. (2001a). Am J Gastroenterol. 96: 579–583. Younossi ZM, Boparai N, Price LL, Kiwi ML, McCormick M, Guyatt G. (2001b). Am J Gastroenterol. 96: 2199–2205. Younossi ZM, Guyatt G, Kiwi M, Boparai N, King D. (1999). Gut. 45: 295–300.
3
Quality of Life Measures and Indices 3.2 Surgical
132 Quality of Life and Functional Outcome in Pediatric Patients Requiring Surgery: Italian Perspectives M. Castagnetti 1 1.1 1.2 1.3
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2258 From Survival to Quality of Life Assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2258 Quality of Life and Health-Related Quality of Life . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2259 The Pediatric Setting and the Pediatric Surgical Setting . . . . . . . . . . . . . . . . . . . . . . . . . . . 2259
2
Examples of HRQoL Assessment in Pediatric Surgical or Urological Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2261 2.1 Muscle-Sparing Thoracotomy for Benign Lung Conditions . . . . . . . . . . . . . . . . . . . . . . . . 2261 2.2 Esophageal Atresia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2261 2.3 Spina Bifida . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2265 3 Considerations Based on the Studies Described . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2266 3.1 Is There a Constant Correlation Between Functional Outcome and HRQoL? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2266 3.2 Can Improved Surgical Outcome Only be Reached at the Price of Poorer HRQoL? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2266 3.3 Can Assessment of HRQoL Help in the Decision-Making? . . . . . . . . . . . . . . . . . . . . . . . . 2267 3.4 Which Instruments should be Preferred in the Evaluation of HRQoL in Pediatric Surgical Patients? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2267 3.5 Can Assessment of HRQoL be Based on Proxy Report? . . . . . . . . . . . . . . . . . . . . . . . . . . . 2267 3.6 Final Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2268 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2268
#
Springer Science+Business Media LLC 2010 (USA)
2258
132
Quality of Life and Functional Outcome in Pediatric Patients Requiring Surgery
Abstract: Health related > quality of life (HRQoL) is a multidimensional concept that describes the impact of a condition and the related morbidity on a person’s feeling of leading a fulfilling life. Many studies in pediatric surgery and urology have so far failed to establish HRQoL and only focused on mortality and morbidity rates. By reviewing studies about HRQoL in three pediatric surgical or urological conditions, we discuss some of the most relevant points concerning the evaluation of HRQoL and its relationship with functional outcomes. The studies reviewed show that HRQoL does not necessarily parallel > functional results. Studies in survivors of esophageal atresia, for instance, show that an acceptable HRQoL can be achieved even despite significant morbidity. Studies in subjects born with > spina bifida, instead, show that, although surgery can achieve dramatic improvements in functional outcome in terms of urinary continence, this is not paralleled by any improvement in the HRQoL of these patients. HRQoL seems instead to improve in the caregivers. The studies reviewed also suggest that HRQoL is largely unrelated to the severity of the condition or the presence of associated anomalies. Consistently, improvement of survival of more severe neonatal conditions does not seem to be associated with a reduction in HRQoL. In conclusion, comprehending issues that influence HRQoL allows us to offer clear expectations of outcomes after surgery to patients, families, health care professionals and policy makers in a truly patient-oriented, evidence-based manner. List of Abbreviations: GIQLI, gastrointestinal quality of life index; HRQoL, health related quality of life; ICQ, illness cognition questionnaire; PedsQL, pediatric quality of life inventory 4.0; QoL, quality of life; RSRQLI, respiratory symptoms–related quality of life index; SF-36, short form – 36; SF-12, short form – 12™ health survey
1
Introduction
1.1
From Survival to Quality of Life Assessment
Children can be exposed to a variety of conditions requiring surgical treatment. Over the last decades, improvements in diagnosis, management, anesthesia, peri-operative support, and surgical techniques have been paralleled by an improvement in the results of surgical treatment of many congenital and acquired pediatric surgical conditions. For instance, before 1960 less than 10% of spina bifida patients survived infancy whereas today the number exceeds 85% (Rinck et al., 1989); during the same period survival for esophageal atresia increased from about 40% to more than 95% (Louhimo and Lindahl, 1983; Ure et al., 1998). This has caused a shift in the primary outcome measures considered for the evaluation of treatment from mere survival, to functional outcomes; and from the latter to quality of life (QoL). QoL is of special interest for chronic diseases, and for diseases associated with a relevant risk of long-term morbidity, or in which treatment may result in a mutilation (Eiser and Morse, 2001). Examples in children include congenital malformations such as esophageal atresia, anorectal malformations, congenital diaphragmatic hernia, or spina bifida; acquired chronic diseases such as cancer or inflammatory bowel diseases; and chronic organ failures (such as liver, kidney or bowel) requiring organ transplant (Stolk et al., 2000).
Quality of Life and Functional Outcome in Pediatric Patients Requiring Surgery
132
Comprehending issues that influence QoL allows us to offer clear expectations of outcomes after surgery to patients, families, health care professionals and policy makers in a truly patient-oriented, evidence-based manner.
1.2
Quality of Life and Health-Related Quality of Life
According to the World Health Organization (1995), QoL is a multidimensional concept encompassing three main domains: (1) the physical domain, which includes independence in activities of daily living and symptoms of disease; (2) the psychological domain, involving emotional, cognitive and behavioral status; and (3) the social domain, how people perceive their role and relationship with other people. Given its complexity, QoL is difficult to define and also to quantify. Therefore, a more practical approach is to restrict the assessment of QoL to the health related QoL (HRQoL), which can be defined as a multidimensional concept that includes the functional status, the psychological and social well being, the health perception and the disease and treatment-related symptoms (Guyatt et al., 1993). In other words, the impact of a chronic condition and the related morbidity on a person’s feeling of leading a fulfilling life. Initial attempts at evaluating HRQoL in pediatric surgical patients were done using ad hoc questionnaires focusing on multiple aspects of patient’s life such as educational level, living arrangements, employment, and sexual relationship (Bomalaski et al., 1995; Bouman et al., 1999). Nevertheless, standardized questionnaires should be preferred. The latter can be differentiated into generic or disease-specific instruments (Patrick and Deyo, 1989). Each of the two has advantages and drawbacks. A combination of disease-specific and generic instruments can also be used, but this drastically increases the number of items to be administered. Among the generic instruments, The Short Form – 36 (SF-36) is the most used instrument in the evaluation of adults treated as children for a congenital disease (Koivusalo et al., 2005). The pediatric quality of life inventory 4.0 (PedsQL), instead, is the most used for the evaluation of the pediatric population (Varni et al., 1999a, b, 2001, 2003). First described in 1999, it has been used in 122 studies up to January 2008 in a vast array of conditions throughout the world. Of note, of such 122 studies only 7 were in the field of pediatric surgery/urology (> Figure 132‐1). Indeed, it appears that many studies in pediatric surgery and urology have so far failed to establish HRQoL and only presented mortality rates and crude measures of childhood morbidity (Stolk et al., 2000).
1.3
The Pediatric Setting and the Pediatric Surgical Setting
HRQoL assessment in pediatric surgical patients is unique in several respects (Eiser and Morse, 2001). As for all the other pediatric conditions, surgical and non-surgical alike, it needs specific questionnaires devised for or adapted to children. Such questionnaires have also to be age-specific, as the needs of children change with age and because young children have limited notions of abstract concepts and language. For very young children, questionnaires need to be devised in formats for caregiver report, as child selfreport is impossible.
2259
2260
132
Quality of Life and Functional Outcome in Pediatric Patients Requiring Surgery
. Figure 132‐1 Percent of studies performed using the pediatric quality of life inventory 4.0 (PedsQL). Types of studies using the PedsQL based on a Medline/PubMed search made in January 2008. Most of the studies regard medical conditions or validation of the inventory. In only 6% of the studies, the PedsQL was used to investigate the health-related quality of life in pediatric patients with surgical or urological conditions
Particularly dealing with patients born with a congenital malformation, some other aspects need to be considered. To begin with, many of these patients have never experienced any previously healthy condition to compare their state of disease with. Many may have also developed > coping mechanisms for their handicap. Many of the most severe pediatric surgical patients are syndromic, which means that they have multiple concomitant malformations, which may have an impact on their HRQoL. Finally, surgical reconstruction is often anatomical and functional. It will definitely change the patient body image and the function patients have adjusted to. Therefore, although perception of treatment-related symptoms is considered key in every HRQoL assessment, it is particularly important in surgical patients because of the irreversibility of changes suddenly caused by surgery. Furthermore, considering surgical patients, it is apparent that sometimes the need for multiple procedures rather than the condition itself can affect the quality of life, or that the HRQoL might depend more on the type of reconstructive procedure used than on the condition itself. Herein we report the current research about HRQoL in three pediatric surgical or urological conditions, and discuss some of the most relevant points concerning the evaluation of HRQoL in pediatric surgical patients (> Table 132-1).
. Table 132‐1 Key questions concerning the evaluation of health related quality of life (HRQoL) in pediatric patients with surgical or urological conditions Is there a constant correlation between functional outcome and HRQoL? Can improved surgical outcome only be reached at the price of poorer HRQoL? Can assessment of HRQoL help in the decision-making? Which instruments should be preferred in the evaluation of HRQoL in pediatric surgical patients? Can assessment of HRQoL be based on proxy report?
Quality of Life and Functional Outcome in Pediatric Patients Requiring Surgery
132
2
Examples of HRQoL Assessment in Pediatric Surgical or Urological Conditions
2.1
Muscle-Sparing Thoracotomy for Benign Lung Conditions
Lung resection is required in children for a variety of congenital and acquired conditions and is classically performed via a postero-lateral thoracotomy. Musculoskeletal anomalies are the major long-term complications associated with such a procedure (Jaureguizar et al., 1985). > Muscle-sparing thoracotomy, video-assisted techniques, and use of mechanical stapling devices are alternatives or adjuncts proposed over time in order to reduce the invasiveness of lung resections (Mattioli et al., 1998; Rothenberg, 2000; Soucy et al., 1991). In order to evaluate the functional outcome and the HRQoL in children undergoing a stapled lung resection via a muscle-sparing thoracotomy, Mattioli et al. (2006) performed a clinical, radiological and functional evaluation, and an assessment of HRQoL in children undergoing this operation. Children older than 5 years and with at least 1 year of follow-up after surgery were included in the study. Presence of asymmetry of the chest wall, rib fusion, breast and pectoral muscle maldevelopment, abnormal rib cage dynamic, winged scapula, elevation or fixation of the shoulder, and/or scoliosis was assessed by a physician not previously involved in the care of the patients assessed. Oxygen saturation was evaluated at rest and during exercise. A chest X-ray was performed in all the cases, while a CT scan only if clinically indicated. A spirometry was also offered. An Italian version of the Short Form – 12™ Health Survey (SF-12) modified for children was used to assess HRQoL (Apolone et al., 2001). This was administered to children and their caregivers. The SF-12 is a simplified version of the SF-36, which has been proved to be reliable in the assessment of HRQoL in children. Nineteen of the initial 52 patients with histologically proved benign lung disorders treated by a stapled lung resection via a muscle-sparing thoracotomy were eventually available for the study. Musculoskeletal anomalies were observed in three such cases. All the patients had normal X-rays, but for the presence of a mild thickening in peri-bronchial vascular markers in three and a mild pleural thickening in two patients. Spirometry was normal in 10 patients, obstructive in 4, restrictive in the remaining 5. HRQoL was excellent or good in 17 out of 19 (89.5%) patients. An abnormal spirometric pattern resulted significantly more frequent in case with a poorer HRQoL (> Figure 132‐2) whereas the rate of symptomatic patients was not different (> Figure 132‐3). The type of resection did not affect the outcome with a proportion of patients with excellent or very good HRQoL not statistically different in patients undergoing anatomical versus wedge resections (> Figure 132‐4).
2.2
Esophageal Atresia
Long-term follow-up studies have shown that patients with esophageal atresia have morbidity from dysphagia, gastro-esophageal reflux, respiratory disorders, and problems related to associated anomalies (Anderson et al., 1992; Engum et al., 1995; Somppi et al., 1998; Ure et al., 1998).
2261
2262
132
Quality of Life and Functional Outcome in Pediatric Patients Requiring Surgery
. Figure 132‐2 Health related quality of life (HRQoL) in relation to spirometric findings. An excellent or very good HRQoL was statistically more common in patients with a normal spirometric pattern compared to patients with sufficient or poor HRQoL. Modified from Mattioli et al. Pediatr Surg Int 2006; 22: 491–495
. Figure 132‐3 Health related quality of life (HRQoL) in relation to symptoms. HRQoL was not correlated to the presence of symptoms. Modified from Mattioli et al. Pediatr Surg Int 2006; 22: 491–495
Four studies addressed the issue of HRQoL in patients with this condition using standardized and validated questionnaires (Deurloo et al., 2005; Koivusalo et al., 2005; Ure et al., 1995, 1998). Two such studies were from the same German group that initially assessed QoL together with the functional results from eight pediatric patients after colon interposition for long-gap esophageal atresia (Ure et al., 1995), than extended the same protocol to 58 patients reassessed more than 20 years after correction of their esophageal atresia (Ure et al., 1998). Fifty patients with primary anastomosis and eight surviving patients with colon interposition were studied. The mean age was 25.3 years (range, 20–31).
Quality of Life and Functional Outcome in Pediatric Patients Requiring Surgery
132
. Figure 132‐4 Health related quality of life (HRQoL) in relation to the type of resection. HRQoL was not correlated to the type of lung resection. Modified from Mattioli et al. Pediatr Surg Int 2006; 22: 491–495
Symptoms were evaluated by a standardized interview. HRQoL assessment was performed using a combination of disease-specific and generic instruments (Eypasch et al., 1995; Selby et al., 1984; Slim et al., 1999; Spitzer et al., 1981) (> Table 132‐2). Among patients undergoing primary anastomosis, 60% suffered from respiratory symptoms, mainly attacks of cough, frequent bronchitis, and short breath; 48% reported hold up, which was instead the most frequent gastrointestinal symptom after primary anastomosis; and 22% complained gastro-oesophageal reflux symptoms such as heartburn or regurgitation. In spite of such symptoms, HRQoL was unimpaired in these patients with a global score of 80 of 100 points, comparable to that of healthy individuals (Selby et al., 1984). Among patients undergoing colon interposition, symptoms were more frequent. All of these patients suffered from periods of short breath. The Spitzer Index and the Gastrointestinal Quality of Life Index (GIQLI) were significantly lower compared with . Table 132‐2 Instruments for the evaluation of HRQoL in studies on esophageal atresia patients Study
Generic instruments
Disease-specific Instruments
Ure et al., 1995, 1998
Visual analogue scale (Selby et al., 1984)
Spitzer Index (Spitzer et al., 1981); GIQLI (Eypasch et al., 1995)
Deurloo et al., 2005
SF-36 (Validated Dutch translation) (Aaronson et al., 1998); ICQ (Evers, 2001)
GIQLI (Eypasch et al., 1995; Slim et al., 1999); 19 of the 24 items of the Esophageal Cancer Module questionnaire (Blazeby, 2003)
Koivusalo et al., 2005
SF-36 (Aalto et al., 1999); Tests of psychosocial functioning (Nurmi et al., 1995)
RSRQLI (Koivusalo et al., 2005); GIQLI (Eypasch et al., 1995; Slim et al., 1999)
SF-36: Short Form – 36; GIQLI: Gastrointestinal Quality of Life Index; RSRQLI: Respiratory Symptoms–Related Quality of Life Index; ICQ: Illness Cognition Questionnaire
2263
2264
132
Quality of Life and Functional Outcome in Pediatric Patients Requiring Surgery
patients undergoing primary anastomosis, and the GIQLI also compared with healthy volunteers. However, the impairment in the GIQLI was exclusively caused by specific symptoms, which had no impact on physical functions, emotions, and social functions. The long-term HRQoL also in patients with colon interposition was acceptable. Besides suffering from specific symptoms these patients lead an otherwise normal life. Deurloo et al. (2005) analyzed 97 esophageal atresia survivors, 16–48-year old. They too used a combination of generic and disease-specific questionnaires (> Table 132‐2) (Aaronson et al., 1998; Blazeby et al., 2003; Evers et al., 2001; Eypasch et al., 1995; Slim et al., 1999). They found no differences in overall physical and mental health when comparing the generic HRQoL in patients who had had esophageal atresia with healthy subjects. Moreover, generic HRQoL did not appear to be influenced by the presence of concomitant congenital anomalies. Patients with concomitant congenital anomalies scored significantly lower only in the domain indigestion, which could probably reflect a focalization on the gastrointestinal system rather than the real presence of symptoms. Unfortunately, the number of patients with long-gap esophageal atresia in this study was too small to make a comparison between patients with and without long-gap esophageal atresia. The authors conclude that after esophageal atresia correction patients generally perceive their generic and > disease-specific HRQoL to be good. The presence of concomitant congenital anomalies did not influence generic HRQoL. However, a third of patients reported that the disease had had negative consequences. Finally, Koivusalo et al. from Filland (Koivusalo et al., 2005), compared 159 esophageal atresia survivors with 400 healthy children. They used a 5-part questionnaire incorporating a combination of generic and disease-specific questionnaires (> Table 132-2) (Aalto et al., 1999; Eypasch et al., 1995; Koivusalo et al., 2005; Nurmi et al., 1995; Slim et al., 1999). Median age was 38 (range, 24–54) years in the patients with esophageal atresia and 36 (range, 20–56) years in the control subjects (P = NS). Respiratory symptoms were significantly more frequent and more serious in patients with esophageal atresia. Mean Respiratory Symptoms–Related Quality of Life Index (RSRQLI) scores were significantly higher in control subjects than in patients with esophageal atresia. However, in both groups, the mean RSRQLI score level was high, suggesting a low overall incidence of significant respiratory symptoms. Mean GIQLI scores were not statistically different between patients with esophageal atresia and control subjects. Assessment of HRQoL with the SF-36 showed that the incidence of poor quality of life in patients with esophageal atresia – including both physical and mental domains – was 14.8% (19 patients), which is within the expected incidence of 16% of the general population. Moreover, the health problems that the patients with esophageal atresia with low HRQoL graded as most significant were related to acquired diseases (n = 11) – psychiatric problems (n = 5), acquired musculoskeletal problems (n = 3), hypertension (n = 2), and malignancy (n = 1) – or to congenital or esophageal atresia-related diseases (n = 8) – haircartilage hypoplasia (n = 1), functional gastrointestinal disorders (n = 2), gastro-oesophageal reflux (n = 2), respiratory problems (n = 1), vaginal atresia related vulvodynia (n = 1), and mental retardation (n = 1). HRQoL did not differ significantly between patients with short- and long-gap types of esophageal atresia as well as among different types of esophageal reconstruction. HRQoL
Quality of Life and Functional Outcome in Pediatric Patients Requiring Surgery
132
of patients with colon interposition was somewhat but not statistically significantly lower than in patients undergoing primary anastomosis. Functional and social problems of patients with esophageal atresia with gastric tube or colon interposition decreased significantly as the patients reached their 20s. The Authors concluded that most adult survivors of esophageal atresia repair have a normal HRQoL. Morbidity from esophageal functional disorders and respiratory disorders with or without acquired diseases impairs HRQoL in 15% of patients with esophageal atresia.
2.3
Spina Bifida
Spina Bifida is the second most common birth defect worldwide. Patients with spina bifida can experience a variety of health problems such as ambulatory problems, faecal and urinary incontinence. Several health specialists have subjected spina bifida patients to extensive study. Here we will focus on the role of continence surgery on the final HRQoL of these patients. Indeed, it might be assumed that reconstruction for incontinent spina bifida might improve HRQoL. MacNeily et al. (2005) performed a retrospective cohort study of 36 consecutive incontinent spina bifida cases undergoing surgery. The latter included augmentation, with or without creation of a > Mitrofanoff catheterizable conduit, bladder-neck reconstruction and cecostomy. These patients were compared with a group of patients not undergoing continence surgery, but otherwise matched for age, lesion level, parental marital status, ambulatory status and shunt status. A 5-point Likert questionnaire (> 5-point Likert Scale) was used for self-scoring of bladder and bowel continence. HRQoL was assessed by a validated disease-specific discriminative instrument (Parkin et al., 1997). The latter was also age specific and patients were stratified for ages, 12 years or less and 13 years or greater. After surgery, 78% of reconstructed cases achieved urinary continence for 3 h or more with equal or superior self-reported bladder and bowel continence compared to controls. This, however, was not paralleled by a similar improvement in HRQoL. The 2-sample t testing revealed no significant difference in mean HRQoL score between those who underwent reconstruction, both in children younger than 12 years than for those older than 13. The authors concluded that surgery may have no impact on HRQoL. The conclusion seems supported by further two studies. Parekh et al. (2006) assessed prospectively HRQoL in 10 spina bifida patients before and up to 6 moths after continence surgery using PedsQL 4.0 (Varni et al., 1999a, 2001). Although the results of surgery were excellent, they were not paralleled by any improvement in the HRQoL of the patients. Of note, caregivers’ HRQoL seems instead to increase significantly after surgery. The second study is a French cross-sectional multicentric study (Lemelle et al., 2006) attempting to determine the relationships between methods of management or urinary/faecal incontinence, methods of management, and HRQoL in 460 spina bifida patients cared for in six centers. HRQoL was evaluated by the SF-36 in adults and the VSP in children. Using both an univariate and multivariate analysis the authors found that urinary/faecal incontinence and their medical management may not play a determinant role in HRQoL of patients with spina bifida.
2265
2266
132 3
Quality of Life and Functional Outcome in Pediatric Patients Requiring Surgery
Considerations Based on the Studies Described
Evaluation of HRQoL allows us to give answer in a truly patient-oriented manner to many questions. Starting from the studies mentioned above, some key questions are addressed (> Table 132-1).
3.1
Is There a Constant Correlation Between Functional Outcome and HRQoL?
It might be hypothesized HRQoL and functional outcome/morbidity of surgery to be correlated to each other. Sometimes this is the case. In the study by Mattioli et al. (2006) about stapling lunge resections via a muscle sparing thoracotomy, the authors actually found that presence of an abnormal spirometric finding was statistically more common in patients with a poorer HRQoL. This observation however does not seem to be always true. The studies on HRQoL in patients operated on as newborns for esophageal atresia, for instance, show that although most of these patients may experience significant long-term morbidity or symptoms, their HRQoL can be expected to be normal or at least acceptable (Deurloo et al., 2005; Koivusalo et al., 2005; Ure et al., 1995, 1998). On the contrary, the studies on the HRQoL in spina bifida patients undergoing continence surgery show that these patients, in spite of dramatic improvements in the functional outcome in terms of continence after surgery, can still perceive their HRQoL as unchanged (Lemelle et al., 2006; Macneily et al., 2005; Parekh et al., 2006).
3.2
Can Improved Surgical Outcome Only be Reached at the Price of Poorer HRQoL?
This is a key question especially in patients with congenital disease where increased survival corresponds to a survival of an increased number of patients with more severe conditions or multiple associated anomalies. Deurloo et al. (2005) showed that this is not the case in patients with esophageal atresia. Although, a third of their patients reported that the disease had had negative consequences, generic and disease-specific QOL was generally perceived to be good, and this irrespective of the presence of concomitant congenital anomalies. It should be said, however, that although in this study, as well as in that by Koivusalo et al. (2005), the number of patients with associated anomalies was quite high, that of severe cases with long-gap atresias was instead quite limited. Indeed, both studies assessed the long-term quality of life in adult patients that survived an operation done even more than 20 years before, therefore in an age when the worse cases usually died. Both studies (Ure et al., 1995, 1998), however, addressed the issue of as to whether cases requiring more complex reconstructions, such as colon interposition, should be expected to have a worse HRQoL. Indeed, HRQoL was found to be acceptable also in these complex cases, although they suffered a greater long-term morbidity and scored slightly lower on the HRQoL tests. The fact that increased survival is not necessarily associated with a worse HRQoL does not seem to be peculiar of esophageal atresia. Poley et al. (2004) from Rotterdam reported similar results in patients with congenital diaphragmatic hernia and ano-rectal malformations. These
Quality of Life and Functional Outcome in Pediatric Patients Requiring Surgery
132
patients were found to experience considerable symptomatology, but the vast majority ultimately enjoyed healthy lives.
3.3
Can Assessment of HRQoL Help in the Decision-Making?
Assessment of HRQoL in pediatric surgical patients allows for a definitive assessment of the results of surgery. This is of interest for multiple reasons. First, it allows a more accurate comparison of treatments and therefore a choice of the best one among multiple possible alternatives. Second, the improvement in survival has also led to an increase in the survival of more severe cases. In turn, this has increased the costs of intensive care assistance and for the care of patients who can have chronic morbidity. The increasing budget constraints, has evoked the question of whether the effects of a given treatment are worth the costs. There is growing political interest in evidence-based, cost-effective medicine, including pediatric surgery. Third, as these patients are generally considered at greater risk of long-term morbidity, information on their long-term HRQoL is critical for parents’ counseling. While using data about HRQoL for the decision-making, a critical attitude is mandatory. For instance, data coming from the studies in spina bifida patients seem to suggest that surgery does not affect the HRQoL of these patients. Another possible interpretation however is that, since surgery causes a change in a chronic condition, if performed late, it does not cause immediate changes in the patient HRQoL. Therefore further research in needed to check as to whether surgery should be undertaken at a younger age or a wider interval be allowed before reassessing HRQoL.
3.4
Which Instruments should be Preferred in the Evaluation of HRQoL in Pediatric Surgical Patients?
It is controversial if disease-specific questionnaires should be preferred over generic ones in the assessment of patients’ HRQoL. It should be said that very few disease-specific instrument exists for pediatric patients and many studies actually adopt questionnaires devised for adult conditions (Eiser and Morse, 2001). For instance, all the mentioned studies about the HRQoL in patients operated on for esophageal atresia used as disease specific instrument a questionnaire originally devised for the evaluation of esophageal function in adult patients with esophageal cancer (Deurloo et al., 2005). It is of note that using such instrument, Ure et al. (1998) found that some esophageal atresia patients scored even better than healthy control subjects. This is an example of how coping mechanism can work well in patients born with a congenital malformation. On the other side, generic instruments allow for a more global evaluation of the impact of the condition on the patient life. They also enable comparison across different pediatric chronic and acute health conditions, as well as benchmarking with healthy population norms.
3.5
Can Assessment of HRQoL be Based on Proxy Report?
As mentioned before, studying pediatric conditions, because of the difficulties that small children have with notions of abstract concepts and language, it is often necessary to rely upon
2267
2268
132
Quality of Life and Functional Outcome in Pediatric Patients Requiring Surgery
proxies. It is known that a variety of factors can influence a parent’s rating of his or her child’s HRQoL and findings reported in the literature are indeed equivocal (Canning et al., 1992; Levi and Drotar, 1999; Waters et al., 2000). Nevertheless, recent research in children (Glaser et al., 1997; Theunissen et al., 1998; Barr et al., 2000) suggests that a parent is able to report appropriate information regarding his or her child’s HRQoL, especially concerning observable behaviors. There is as yet no clear evidence of whether the parents over- or under-estimate HRQoL. A number of studies indicate that parents tend to rate the child as having a poorer HRQoL than the child does him or herself, a tendency which would result in a conservative estimate of the HRQoL (Ennett et al., 1991). With regards to the studies mentioned above, Parekh et al. (2006) evaluated HRQoL in spina bifida patients and their parents. And while there was no change in patients’ HRQoL before and after surgery, they observed a significant difference in social functioning in parent report.
3.6
Final Considerations
The changing methodology of patient management should be accompanied by an increased awareness among medical providers toward patient HRQoL. In other words, patient-focused care is the key to improving HRQoL in pediatric surgery as well as other health services, and assessment of HRQoL is key for the development of a truly patient-oriented medicine.
Summary Points Health Related Quality of Life and functional outcomes in children with surgical conditions not necessarily coincide.
Health Related Quality of Life can be good or acceptable in spite of significant morbidity. Health Related Quality of Life can be independent from the severity of the condition and the presence of associated anomalies.
Improved survival in neonates with surgical conditions does not involve a worsening in Health Related Quality of Life.
Proxies are generally reliable in the assessment of Health Related Quality of Life but in some cases surgery might lead to an improvement in Health Related Quality of Life more in the caregivers that in the patients.
References Aalto AM, Aro AR, Teperi J. (1999). Rand-36 as a Measure of Health Related Quality of Life. Reliability and values in Finnish Population. STAKES research publication 101. Gummerrus, Helsinki pp. 1–78. Aaronson NK, Muller M, Cohen PD, Essink-Bot ML, Fekkes M, Sanderman R, Sprangers MA, te Velde A, Verrips E. (1998), J Clin Epidemiol. 51: 1055–1068. Anderson KD, Noblett H, Belsey R, Randolph JG. (1992). Surgery. 111: 131–136.
Apolone G, Mosconi P, Quattrociocchi L, Gianicolo EAL, Groth N, Ware JE Jr. (2001). Questionario sullo stato di salute SF-12. Versione italiana. Guerini e Associati Editore, Milano, pp. 1–85. Barr RD, Chalmers D, De Pauw S, Furlong W, Weitzman S, Feeny D. (2000). J Clin Oncol. 18: 3280–3287. Blazeby JM, Conroy T, Hammerlid E, Fayers P, Sezer O, Koller M, Arraras J, Bomalaski MD, Teague JL, Brooks B. (1995). J Urol . 154: 778–781.
Quality of Life and Functional Outcome in Pediatric Patients Requiring Surgery Blazeby JM, Conroy T, Bottomley A, Vickery C, Arraras J, Sezer O, Moore J, Koller M, Turhal NS, Stuart R, Van Cutsem E, D’haese S, Coens C (2003). Eur J Cancer 39: 1384–1394. Bouman NH, Koot HM, Hazebroek FW. (1999). J Pediatr Surg. 34: 399–404. Canning EH, Hanser SB, Shade KA, Boyce WT. (1992). Pediatrics 90: 692–696. Deurloo JA, Ekkelkamp S, Hartman EE, Sprangers MA, Aronson DC. (2005). Arch Surg. 140: 976–980. Eiser C, Morse R. (2001). Arch Dis Child. 84: 205–211. Engum SA, Grosfeld JL, West KW, Rescorla FJ, Scherer LR 3rd. (1995). Arch Surg. 130: 502–509. Ennett ST, DeVellis BM, Earp JA, Kredich D, Warren RW, Wilhelm CL. (1991). J Pediatr Psychol. 16: 557–568. Evers AW, Kraaimaat FW, van Lankveld W, Jongen PJ, Jacobs JW, Bijlsma JW. (2001). J Consult Clin Psychol. 69: 1026–1036. Eypasch E, Williams JI, Wood-Dauphinee S, Ure BM, Schmu¨lling C, Neugebauer E, Troidl H. (1995). Br J Surg. 82: 216–222. Glaser AW, Davies K, Walker D, Brazier D. (1997). Qual Life Res. 6: 43–53. Guyatt G, Feeny D, Patrick D. (1993). Ann Intert Med. 118: 622–629. Jaureguizar E, Vaizquez J, Murcia J, Diez Pardo JA. (1985). J Pediatr Surg. 20: 511–514. Koivusalo A, Pakarinen MP, Turunen P, Saarikoski H, Lindahl H, Rintala RJ. (2005). J Pediatr Surg. 40: 307–312. Lemelle JL, Guillemin F, Aubert D, Guys JM, Lottmann H, Lotart-Jacob S, Mouriquand P, Ruffion A, Moscovici J, Schmitt M. (2006). Qual Life Res. 15: 1481–1492. Levi RB, Drotar D. (1999). Int J Cancer. 83: 58–64. Louhimo I, Lindahl H. (1983). J Pediatr Surg. 18: 217–229. MacNeily EA, Morrell J, Secord S. (2005). J Urol. 174: 1637–1643. Mattioli G, Buffa P, Granata C, Fratino G, Rossi G, Ivani G, Jasonni V. (1998). Pediatr Surg Int. 13: 10–13. Mattioli G, Asquascaiti C, Castagnetti M, Bellodi S, Rossi G, Jasonni V. (2006). Pediatr Surg Int. 22: 491–495. Nurmi JE, Salmela-Aro K, Haavisto T. (1995). Eur J Psychol Assess. 11: 108–121.
132
Parekh AD, Trusler LA, Pietsch JB, Byrne DW, DeMarco RT, Pope JC, Adams MC, Deshpande JK, Brock JW. (2006). J Urol .176: 1878–1882. Parkin PC, Kirpalani HM, Rosenbaum PL, Fehlings DL, Van Nie A, Willan AR, King D. (1997). Qual Life Res. 6: 123–132. Patrick D, Deyo R. (1989). Med Care. 27: S217–S232. Poley MJ, Stolk EA, Tibboel D, Molenaar JC, Busschbach JJV. (2004). Arch Dis Child. 89: 836–841. Rinck C, Berg J, Hafeman C. (1989). Adolescence. 24: 699–710. Rothenberg SR. (2000). J Pediatr Surg. 35: 271–274. Selby PJ, Chapman JA, Etazadi-Amoli J, Dalley D, Boyd NF. (1984). Br J Cancer. 50: 13–22. Slim K, Bousquet J, Kwiatkowski F, Lescure G, Pezet D, Chipponi J. (1999). Gastroenterol Clin Biol. 23: 25–31. Somppi E, Tammela O, Ruuska T, Rahnasto J, Laitinen J, Turjanmaa V, Ja¨rnberg J. (1998). J Pediatr Surg. 33: 1341–1346. Soucy P, Bass J, Evans M. (1991). J Pediatr Surg. 26: 1323–1325. Spitzer WO, Dobson AJ, Hall J, Chesterman E, Levi J, Shepherd R, Battista RN, Catchlove BR. (1981). J Chronic Dis. 34: 585–597. Stolk EA, Post HA, Rutten FF, Molenaar JC, Busschbach JJ. (2000). J Pediatr Surg. 35: 588–592. Theunissen NC, Vogels TG, Koopman HM, Verrips GH, Zwinderman KA, Verloove-Vanhorick SP, Wit JM. (1998). Qual Life Res. 7: 387–397. Ure BM, Slany E, Eypasch EP, Gharib M, Holschneider AM, Troidl H. (1995). Eur J Pediatr Surg. 5: 206–210. Ure BM, Slany E, Eypasch EP, Weiler K, Troidl H, Holschneider AM. (1998). J Pediatr Surg. 33: 511–515. Varni JW, Seid M, Kurtin PS. (1999a). J Clin Outcomes Manage. 6: 33–40. Varni JW, Seid, M, Rode CA. (1999b). Med Care. 37: 126–139. Varni JW, Seid M, Kurtin PS. (2001). Med Care. 39: 800–812. Varni JW, Burwinkle TM, Seid M, Skarr D. (2003). Ambulat Pediatr. 3: 329–341. Waters E, Doyle J, Wolfe R, Wright M, Wake M, Salmon L. (2000). Pediatrics. 106: 1422–1428. World Health Organization. (1995). Soc Sci Med. 41: 1403–1409.
2269
133 Breast Reduction Surgery and Quality of Life and Clinical Outcomes A. Thoma . L. McKnight 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2272
2 2.1 2.2 2.3 2.4
Summary of Current Evidence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2273 Meta-analysis and Systematic Reviews . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2273 Randomized Controlled Trials (RCTs) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2274 Cohort Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2276 Case Control Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2278
3
Quality of Life Measurements Used in Breast Reduction . . . . . . . . . . . . . . . . . . . . . . . . 2279
4 Generic Scales . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2282 4.1 Short Form 36 (SF-36) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2282 5 Utility Measurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2282 5.1 European Quality of Life-5 Dimensions (EQ-5D) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2282 5.2 Health Utility Index Mark 2/3 (HUI 2/3) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2283 6 Condition-Specific Scales . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2283 6.1 The Multidimensional Body Self Relations Questionnaire (MBSRQ) . . . . . . . . . . . . . . 2283 7
Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2284 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2285
#
Springer Science+Business Media LLC 2010 (USA)
2272
133
Breast Reduction Surgery and Quality of Life and Clinical Outcomes
Abstract: > Breast hypertrophy is a common condition seen by plastic surgeons. Patients with this condition complain of upper back, neck and shoulder discomfort and sometimes chronic headaches. They also have trouble finding proper clothing and have difficulties participating in sport activities. Therefore, this condition carries important burdens in health-related quality of life (HRQL). > Breast reduction surgery is the solution to this problem. However, this procedure remains controversial in some geographic jurisdictions because third party payers refuse to pay for it. Also, rather arbitrarily, some plastic surgeons refuse to perform this surgery on overweight patients with this condition. This chapter provides an up-to-date review of the breast reduction studies in which quality of life was the primary outcome measure. The studies considered in this review covered the spectrum of the level of evidence, from case series to > systematic reviews. The majority of publications, however, fell into the lower levels of the evidence (i.e., cohort and > case control studies). All published studies, irrespective of study design demonstrated substantial improvements in quality of life in women who undergo breast reduction surgery. The recent evidence suggests that overweight patients with breast hypertrophy benefit from breast reduction just as much as thin patients with breast hypertrophy. Additionally, the mean > quality-adjusted life years (QALY) gained per patient because of the surgery was 0.12 during the 1-year follow-up period. The health-related quality-of-life (HRQL) effect of the surgery translates into an expected lifetime gain of 5.32 QALYs, which is equivalent to each patient living an additional 5.32 years in perfect health. List of Abbreviations: QOL, quality of life; HRQL, health related quality of life; RCT, > randomized controlled trial; HUI 2/3, health utilities index mark 2/3; SF-36, short form 36 health survey questionnaire; STAI, state–trait anxiety inventory; MBSRQ, multidimensional bodyself relations questionnaire; MPQ, McGill pain questionnaire; BRS, breast-related symptoms; GHQ12, general health questionnaire; STAI, state-trait anxiety index; RSE, Rosenberg selfesteem scale; SCS, self-consciousness scale; DAS-59, Derriford appearance scale 59; EQ-5D, European quality of life-5 dimensions; FPQ, Finnish pain questionnaire; FBAS, Finnish breastassociated symptoms questionnaire; 15D, 15D quality of life questionnaire; HADS, hospital anxiety and depression score; FANLT, functional assessment of non-life threatening conditions version 4; EPQ-R, Eysenck personality questionnaire-revised; HAQ-20, The Stanford health assessment questionnaire; DBPT, digital-body-photo-test; CAPT, color-a-person body dissatisfaction test; SQLP, subjective quality of life profile; NASS, The North American Spine Society Lumbar Spine Outcome Assessment Instrument; BSI, breast symptom inventory
1
Introduction
In addition to relieving clinical symptoms and prolonging survival, the primary objective of any health care intervention is the enhancement of quality of life and well-being (Berzon, 1998). The broader term of “quality of life” (QOL) can be defined as “the adequacy of people’s material circumstances and to their feelings about these circumstances” (McDowell, 2006). This encompasses indicators of life satisfaction, personal wealth and possessions, level of safety, level of freedom, spirituality, health perceptions, physical, psychological, social and cognitive well-being (McDowell, 2006). Health-related quality of life (HRQL), a sub-component of QOL, comprises all areas specific to health i.e., physical, emotional, psychological, social, cognitive, role functioning
Breast Reduction Surgery and Quality of Life and Clinical Outcomes
133
as well as abilities, relationships, perceptions, life satisfaction and well being, and refers to patients’ appraisals of their current level of functioning and satisfaction with it, compared to what they perceived to be ideal (Guyatt et al., 1993; Cella et al., 1990). Impairment in HRQL is a major reason why patients seek surgical care (Thoma et al., 2008a). Breast hypertrophy has been reported by patients to be associated with important burdens in HRQL, specifically pain, discomfort, and emotion. It also creates functional disabilities that adversely affect women because of disproportionate upper body weight. There is cumulative evidence from several studies that breast hypertrophy is associated with significant morbidity and reduced HRQL (Chadbourne et al., 2001; Jones and Bain, 2001; Chao et al., 2002; Collins et al., 2002; Miller et al., 2005; Thoma et al., 2007). Despite the growing evidence showing the salutary effect of reduction mammaplasty on women with breast hypertrophy, some insurers and government agencies set arbitrary body weight and/or tissue resection weight restrictions for coverage for reduction mammaplasty (Klassen et al., 1996; Collins et al., 2002; Kerrigan, 2005; Wagner and Alfonso, 2005; Schmitz, 2005; Thoma et al., 2007). Previous studies using a variety of instruments have reported that reduction mammaplasty had a substantial improvement in HRQL regardless of body weight or tissue resection weight. Collins and colleagues (2002) reported that weight loss was not an effective method of relieving the symptoms. In a recent retrospective chart review, Wagner and Alfonso (2005) found no significant difference among the various body mass index groups in terms of symptom relief or development of complication. Thoma and colleagues (2007) demonstrated that women with breast hypertrophy of all weights benefit from reduction surgery. Women having “small” reductions ( meta-analysis and systematic reviews of high quality randomized controlled trials (RCTs), RCTs, cohort studies, casecontrol studies, case series, expert opinions, and in vitro and animal studies (Sprague et al., 2008). This ranking has an evolutionary order, moving from simple observational methods at the bottom through to increasingly sophisticated and statistically refined study designs at the top level of evidence. Many of the publications in the breast reduction surgery literature fall into the lower levels of the evidence.
2.1
Meta-analysis and Systematic Reviews
In a well done systematic review the findings of all high quality studies pertaining to a particular clinical question are evaluated together to provide more valid information than any one study can (Haines et al., 2008). In a meta-analysis, the results of the primary studies that meet the standards for inclusion in a review are mathematically pooled to give a result that is more precise because of the overall increase in numbers of study participants
2273
2274
133
Breast Reduction Surgery and Quality of Life and Clinical Outcomes
contributing data (Haines et al., 2008). Systematic reviews and meta-analyses of the effectiveness of treatments can be performed based on RCTs and/or observational studies. However, RCTs are the traditional study design of choice for primary studies used in meta-analyses, as they are the most likely to be valid (Haines et al., 2008). There have been two systematic reviews performed on the breast hypertrophy literature (Chadbourne et al., 2001; Jones and Bain, 2001). Chadbourne and colleagues (2001) performed a systematic review and subsequent meta-analysis of the breast hypertrophy literature from 1985–1999. Twenty nine studies were identified that examined physical breast symptoms and QOL in reduction mammaplasty patients. The authors were able to pool data from 15 studies on physical symptoms (headache, neck, shoulder, lower back and breast pain, shoulder grooving, numbness in hand and intertrigo). They found an improvement in all physical symptoms post-operatively. Only four studies administered specific QOL scales to assess QOL. Due to a lack of data, only psychological and physical functioning domains could be analyzed, revealing a risk difference (95% confidence interval) of 0.46 (0.00–1.00) and 0.58 (0.44–0.71) respectively. Jones and Bain (2001) conducted a similar systematic review of breast hypertrophy articles from 1966–1997. Of the 17 publications identified, all reported physical breast symptoms and ten studies examined QOL (four using validated QOL scales) in reduction mammaplasty patients. Although the reporting outcomes varied greatly among studies, a substantial improvement in physical breast symptoms was reported in all studies. The measurement and reporting of quality of life was inconsistent among studies and was therefore difficult to analyze. However, all studies reported an improvement of psychological well-being after surgery. The authors were not able to perform a meta-analysis due to a lack of data on the subject. Neither systematic review identified any randomized controlled trials in the breast hypertrophy literature from 1966 to 1999. Randomized controlled trials are considered the most scientifically rigorous study design.
2.2
Randomized Controlled Trials (RCTs)
RCTs offer the maximum protection against bias and are generally regarded as the most scientifically rigorous study design to evaluate the effect of a surgical intervention (Sprague et al., 2008). This type of study offers the maximum protection against biases in the choice of treatment as it facilitates blinding and reduces selection bias, and it balances both known and unknown prognostic factors across treatment groups. Lack of randomization predisposes a study to potentially important imbalances in baseline characteristics between two study groups (Sprague et al., 2008). In breast reduction RCTs have evaluated the use of drains (Rayatt et al., 2005; Collis et al., 2005), lung function (Iwuagwu et al., 2006a), complication rates (Cruz-Korchin and Korchin, 2003), pain management (Bell et al., 2001; Culliford et al., 2007) and upper limb nerve conduction (Iwuagwu et al., 2005). However, very few RCTs in breast hypertrophy have examined QOL (> Table 133-1). Iwuagwu and colleagues (2006b,c) randomized patients to early surgery (surgery within 3 weeks of initial visit) or late surgery (surgery within 4–6 month of initial visit). QOL was measured using several validated patient reported questionnaires at the initial visit and 16 weeks after surgery (early surgery group, n = 36) and 16 weeks after initial visit (late surgery group, n = 37). Patients randomized to reduction mammaplasty had significant improvements in depression, pain, anxiety, extroversion and emotional stability.
Breast Reduction Surgery and Quality of Life and Clinical Outcomes
133
. Table 133-1 Randomized Controlled Trials (RCTs) examining quality of life (QOL) in breast hypertrophy patients Comparative intervention
Outcomes
Breast hypertrophy women not undergoing surgery, n = 42
Quality of 6 months life, pain, breast related symptoms
Statistically significant and clinically important improvements in all scores of the SF-36, FPQ, FBAS, 15D post-operatively
Iwuagwu Delayed et al., 2006b surgery (Surgery within 2 weeks of initial visit), n = 36
Early surgery (Surgery within 4–6 months of initial visit), n = 37
Anxiety 4–6 months and depression
Statically significant increase in the proportion of HADS normal scores in both anxiety and depression scores in the early treatment group
Iwaugwu Delayed et al., 2006c surgery (Surgery within 2 weeks of initial visit), n = 36
Early surgery (Surgery within 4–6 months of initial visit), n = 37
Quality of life and emotional stability
4–6 months
Early surgery groups experienced statistically significant improvements in: All domains of the FANLT and SF-36, Pain, anxiety and depression domains of the EuroQOL, Extroversion and emotional stability of EPQ-R
Freire et al., Delayed 2007 surgery (n = 50)
Early surgery (n = 50)
Pain and functional capacity
6 months
Early surgery groups experienced statistically significant improvements in the following HAQ-20 domains: getting dressed, getting up, walking, maintaining personal hygiene, reaching, and grasping objects Neck, shoulder and lower back pain using visual analogue pain scale
Authors Saariniemia et al., 2007
Intervention Breast reduction surgery, n = 40
Time horizon
Conclusions
This table summarizes the outcomes of RCTs published in the breast reduction literature. EQ-5D European quality of life-5 dimensions; SF-36 short form 36 health survey questionnaire; FPQ Finnish pain questionnaire; FBAS Finnish breast-associated symptoms questionnaire; 15D 15D quality of life questionnaire; HADS hospital anxiety and depression score; FANLT functional assessment of non-life threatening conditions version 4; EPQ-R Eysenck personality questionnaire-revised; HAQ-20 The Stanford health assessment questionnaire
2275
2276
133
Breast Reduction Surgery and Quality of Life and Clinical Outcomes
Freire et al., (2007) conducted a similar study. One hundred consecutive patients were randomized to early (immediate surgery, n = 50) or late surgery (surgery 6 months after the initial assessment, n = 50). Quality of life was measured using a > generic health scale and a visual analogue pain scale 6 months after the initial visit. Women who underwent surgery experienced significant improvements in neck, shoulder and lower back pain and an increase overall quality of life. Friere et al., (2007) used a lottery randomization system which is not an adequate method. The method of randomization is so crucial to the validity of the study that it needs to be done correctly and be transparent. Using even or odd birth year or alternate chart number or lottery are inadequately concealed and are prone to selection bias. The use random number tables or computer programs to generate the sequences are correct ways to randomize patients (Thoma et al., 2008b). A computer randomization system was used in a recent Finnish study by Saariniemia and colleagues (2007). Breast hypertrophy patients were randomized to receive breast reduction surgery (n = 40) or receive no surgical treatment (n = 42). Quality of life was measured using a combination of validated general health, pain specific and breast symptom questionnaires at the initial visit and 6 months later. Statistically significant and clinically important improvements in all quality of life, pain and breast related symptoms were observed in the surgical group post-operatively.
2.3
Cohort Studies
> Prospective
cohort studies involve the identification and follow-up of individual patients who have received a treatment of interest. Although, prospective cohort studies are considered lower level evidence, they have generated an abundance of important data on the HRQL in patients undergoing breast reduction surgery. Since the last systematic review of the literature (Chadbourne et al., 2001), several new prospective cohort studies have been published (> Table 133-2). In a recent study, we assessed and measured the HRQL experienced by breast reduction patients using four reliable and validated QOL measures (Thoma et al., 2005, 2007). Consecutive patients with breast hypertrophy completed self-reported outcome measurement tools at one week and one day pre-surgery and 1, 6, and 12 months post-surgery. We found an improvement in all health-related quality-of-life measures from before surgery to 1 month after surgery regardless of patient body mass index and tissue resection weight. The improvement from 1 month after surgery was maintained to 1 year after surgery for all health-related quality-of-life instruments. Spector and Karp, (2007) found similar results in a cohort of 171 women who underwent reduction mammaplasty. Quality of life was assessed using a custom-made 5 point likert scale per-operatively and at 1 and 3 years post-operatively. Patients reported improvements in buying clothes, participating in sports and running. These results were similar to an earlier studied performed by Blomqvist and colleagues (2000, 2004). Quality of life of 49 women was assessed using a validated scale and 3 custom made 10 point Likert scale questionnaires preoperatively and 1 and 3 years post-operatively. Statistically significant improvements in overall quality of life and pain were observed at 1 year and maintained 3 years post-operatively. A limitation of these studies was the use of non-validated instruments to assess quality of life.
Breast Reduction Surgery and Quality of Life and Clinical Outcomes
133
. Table 133-2 Cohort studies examining quality of life (QOL) in women undergoing breast reduction surgery Authors
Outcomes assessment scales
Population
Outcomes
Thoma et al., 2007
52 women undergoing reduction mammaplasty
HRQL, breast symptoms, body image and selfesteem
Spector and Karp, 2007
59 women, breast resection of case-control study is a type of observational study which begins with the identification of individuals who already have the outcome of interest, (referred to as the cases), and a suitable control group without the outcome event (referred to as the controls).
Breast Reduction Surgery and Quality of Life and Clinical Outcomes
133
Collins and colleagues (2002) evaluated the quality of life, breast related symptoms, body image, self esteem and pain in women who underwent breast reduction surgery and two control groups: (1) Women with bra size greater than D, “hypertrophy group” (2) Women with bra size less than D “normal group.” Quality of life was assessed using validated instruments at the initial visit and again in the surgical group 6–9 months post-operatively. Control subjects were matched for age; however, hypertrophy and operative groups had significantly higher BMI than that of normal controls. Women who underwent surgery rated their appearance significantly higher post-operatively compared to their initial assessment. Also, women in the surgical group experience significant improvements in quality of life and lower pain post-operatively. Pain post-operatively was similar to pain reported by both controls. Multiple regression analysis revealed improvement in outcomes was not associated with age, weight of resected tissue, BMI or bra cup size pre-operatively. Quality of life, body image and self-esteem were assessed in women suffering from breast hypertrophy (n = 71) and women who had underwent breast reduction surgery in the last two years (n = 94) (Hermans et al., 2005). Groups were matched for age and body mass index. Women in the operative group had significantly better quality of life, lower pain and physical disability and anxiety compared to the non-operative control groups. The non-operative control group demonstrated significantly higher insecurity, shame and unattractiveness (> Table 133-3).
3
Quality of Life Measurements Used in Breast Reduction
Traditionally QOL has been measured by complication rates, photographs and surgical assessments. However, these outcome measures are not sufficient to assess patient quality of life. Validated, reliable and responsive patient questionnaires specific to breast reduction surgery are the best method of measuring breast surgery outcomes. Validity refers to the ability of the questionnaire to measure what is intended to be measured. The ability of the questionnaire to produce consistent and reproducible results determines its reliability. The instrument must be sensitive enough to measure changes as a result of the surgical intervention, termed responsiveness. A wide variety HRQL instruments, generic and disease or > condition specific, have been applied to the area of breast hypertrophy and reduction mammaplasty (> Table 133-4). Generic HRQL instruments allow HRQL to be compared among patients with different types of diseases but may not be sensitive enough to detect small differences in patient groups with specific disabilities (Thoma et al., 2008a). Generic HRQL instruments provide an overall assessment of HRQL, with questions covering many health-related domains such as physical, social, emotional, and cognitive functioning, mental health, pain and general health. These include both descriptive health status questionnaires (i.e., Short Form 36 Health Survey Questionnaire (SF-36)), and health > utility measures (i.e., European Quality of Life-5 Dimensions (EQ-5D)). Utilities measures provide preference-weighted outcome measures that represent patients’ preferences for a given health state relative to death (represented by 0) or perfect health (represented by 1). There are various methods of measuring utilities including the visual analogue scale, the standard gamble, the time trade-off, and standardized questionnaires including the EQ-5D and the Health Utilities Index (HUI) (Thoma et al., 2008a). Disease-specific (condition-specific) HRQL measures consist of questions focusing on specific symptoms and impairments relevant to a particular disease state or surgical intervention (Thoma et al., 2008a). Evidence from other clinical settings has shown that the generic
2279
2280
133
Breast Reduction Surgery and Quality of Life and Clinical Outcomes
. Table 133-3 Cohort studies examining quality of life in breast hypertrophy patients Outcomes assessment scales
Time horizon
Authors
Population
Outcomes
Hermans et al., 2005
Non-operative breast hypertrophy control group (n = 71) and operative group who had breast reduction surgery 2 years prior (n = 94)
Quality of Life, Physical Appearance, Body Image and SelfEsteem
SF-36, EQ-5D, RSE, 2 years SCS, DAS-59, Visual Analogue Scale used to subjectively measure breast appearance
Non-operative group had scored significantly higher in DAS-59 domains: insecurity, pain, shame, and unattractiveness. Operative group had significantly higher self-esteem score (RSE), significantly lower anxiety (SCS), significantly higher SF-36 scores in 7/8 domains, improved pain and physical disability (EQ-5D)
Collins et al., 2002
Group 1: Hypertrophy control group with bra cup sizes > D (n = 88); Group 2: Control group with bra cup sizes Table 133-5).
Breast Reduction Surgery and Quality of Life and Clinical Outcomes
133
. Table 133-4 Generic scales used to assess the benefit of breast reduction surgery Instrument Short form 36 health survey questionnaire (SF-36)
Author, Year Thoma et al., 2007 Iwuagwu et al., 2006 Hermans et al., 2005 Miller et al., 2005 Thoma et al., 2005 Blomqvist and Brandberg, 2004 Collins et al., 2002 Behmand et al., 2000 Klassen et al., 1996
General health questionnaire
Chahraoui et al., 2006 Klassen et al., 1996
Stanford health assessment questionnaire
Freire et al., 2007
Functional assessment of non-LIFE threatening conditions version 4 (FANLT)
Iwuagwu et al., 2006
This table lists the generic scales that have been used to assess quality of life in breast hypertrophy patients
. Table 133-5 Instruments used to assess psychological functioning in breast reduction surgery patients Instrument Rosenberg self esteem scale
Author, Year Hermans et al., 2005 Miller et al., 2005 Klassen et al., 1996
Hospital anxiety and depression (HAD) scale
Iwuagwu et al., 2006
Eysenck personality questionnaire
Iwuagwu et al., 2006
State-trait anxiety inventory
Chahraoui et al., 2006
Derriford appearance scale 59
Hermans et al., 2005
This table lists the psychological functioning scales that have been used to assess quality of life in breast hypertrophy patients
A recent systematic review of quality of life instruments used in breast surgery conducted by Pusic et al., (2007) identified several breast surgery specific questionnaires including: The Multidimensional Body Self Relations Questionnaire (MBSRQ), and Breast Related Symptoms Questionnaire (BRS). Only the BRS has been developed specifically for breast reduction patients. However, very few studies have used breast specific scales to assess QOL after reduction mammaplasty.
2281
2282
133 4
Breast Reduction Surgery and Quality of Life and Clinical Outcomes
Generic Scales
Several different generic instruments have been used to assess the quality of life in breast reduction patients (> Table 133-4). The most commonly used instrument is the Short Form 36.
4.1
Short Form 36 (SF-36)
The SF-36 is a multipurpose, short form health survey with 36 questions consisting of eight domains: Physical function (10 items), role physical (4 items), bodily pain (2 items), general health (5 items), vitality (4 items), social functioning (2 items), role emotional (4 items), and mental health (5 items) (Ware, 1996). The two summary measures of the SF-36 are the physical component summary and the mental component summary. The scores for the multifunction item scales and the summary measures of the SF-36 vary from 0 to 100, with 100 being the best possible score and 0 being the lowest possible score. There is no principle for calculating the magnitude that constitutes a clinically important difference on the Short Form 36 subscales. A 10-point change in scores has been suggested as a rule of thumb with which to apply on 100-point quality-of-life scales (Thoma et al., 2007).
5
Utility Measurement
Utility scores of HRQL derived from responses to generic single index instruments such as the Health Utility Index (HUI) and European Quality of Life-5 Dimensions (EQ-5D) have the required measurement properties for calculating quality-adjusted life years (QALYs). QALYs are the measure of effectiveness in cost-utility analysis. This outcome measure incorporates both changes in quantity of life (i.e., reduction of mortality) and quality of life (i.e., reduction in morbidity) into a standard “metric.” QALY = (duration of health state) (utility of health state) + (future remaining life expectancy duration of health state) (utility of successful reconstruction) QALYs are used in economic analyses to calculate the Incremental Cost-Utility Ratio (ICUR). ICUR determines if the “novel” procedure is cost-effective, when compared to an “old procedure” or not (Thoma et al., 2008b). The ICUR represents the marginal cost per marginal unit of utility (effectiveness) and is calculated as follows: ICUR = △C/△U=(Mean Cost “intervention” – Mean Cost free “comparative intervention”)/ (Mean QALY “intervention” – Mean QALY “comparative intervention”) This ratio, which integrates costs and effectiveness, tells us whether we should adopt the novel procedure. The result is represented as cost per QALY. In simple words, it tells us how much it costs to prolong the life of a patient by one extra year in perfect health. The higher the ICUR, the greater the incremental cost for an additional healthy year of life.
5.1
European Quality of Life-5 Dimensions (EQ-5D)
The EQ-5D was created in 1990 by an international and interdisciplinary team (The EuroQOL Group, 1990). It consists of five dimensions of health (1) mobility, (2) self-care, (3) usual
Breast Reduction Surgery and Quality of Life and Clinical Outcomes
133
activities, (4) pain/discomfort, (5) anxiety/depression. Each dimension comprises three levels (1) no problems, (2) some/moderate problems, (3) extreme problems (> Table 133-6).
. Table 133-6 Utility measurement scales used to assess quality of life in breast reduction surgery patients Instrument EuroQOL
Author, Year Iwuagwu et al., 2006 Hermans et al., 2005 Collins et al., 2002
Health utilities index mark 2/3 (HUI 2/3)
Thoma et al., 2007
This table lists the utility scales that have been used to assess quality of life in breast hypertrophy patients
5.2
Health Utility Index Mark 2/3 (HUI 2/3)
The Health Utility Index is a family of comprehensive, reliable, responsive, and valid multiattribute utility instruments. The HUI is a well-known health status and quality of life assessment instrument developed as an indirect method of measuring utilities (preferences) in clinical trials and other studies (Torrance et al., 1996; Furlong et al., 2001; Feeny et al., 2002). The HUI is a comprehensive, reliable, responsive, and valid multi-attribute utility instrument. Responses to the questionnaire are converted using standard algorithms to levels of the Health Utilities Index Mark 2 (HUI2) and Mark 3 (HUI3) multi-attribute health status classification systems. The attribute levels are combined with published scoring functions to calculate utility scores of overall HRQL. The Health Utilities Index Mark 2 and Mark 3 health status classification systems are complementary. Together, they provide descriptive measures of ability or disability for health state attributes and descriptions of comprehensive health status. The minimum clinically important difference in Health Utility Index means is 0.03 for health-related quality of life and 0.05 for single-attribute utility scores (> Table 133-6).
6
Condition-Specific Scales
6.1
The Multidimensional Body Self Relations Questionnaire (MBSRQ)
The MBSRQ was designed to assess quality of life after breast reconstruction surgery. It is a well-validated self-report inventory for the assessment of self-attitudinal aspects of the bodyimage construct. The MBSRQ is a 69-item self-report inventory for the assessment of selfattitudinal aspects of the body-image constructs (Cash et al., 1990). The MBSRQ is intended for use with adults and adolescents over the age of 15 years (Cash et al., 1990). Two forms of the Multidimensional Body Self Relations Questionnaire are available, the full version and the Multidimensional Body Self Relations Questionnaire Appearance Scales. The full, 69-item version consists of seven factor subscales: (1) appearance evaluation, (2) appearance orientation, (3) fitness evaluation, (4) fitness orientation, (5) health evaluation, (6) health orientation, and
2283
2284
133
Breast Reduction Surgery and Quality of Life and Clinical Outcomes
(7) illness orientation. There are also three multiple item subscales: (1) the body areas satisfaction scale, (2) the overweight preoccupation scales, and (3) the self-classified weight scale. In breast reduction patients, the most relevant subscales are likely the appearance evaluation and the appearance orientation. Measuring body image Scores vary from 1 to 5. A high score indicates emphasis on one’s looks, attention to one’s appearance, and engaging in extensive grooming behaviors. A low score indicates apathy about one’s appearance, one’s looks are not especially important, and not expending much effort to “look good.” High scorers feel mostly positive and satisfied with their appearance; low scorers have a general unhappiness with their physical appearance. A recent systematic review by Pusic and colleagues (2007) identified questionnaires developed and validated for use in cosmetic and reconstructive breast surgery. Only seven questionnaires specific to breast surgery were found: 1. 2. 3. 4. 5. 6.
Dow Corning questionnaire (Cash et al., 2002) McGhan (McGhan Medical Corporation, 1995) Breast Implant Replacement Study (BIRS) (LipoMatrix, Inc., unpublished) Breast Evaluation Questionnaire (BEQ) (Anderson et al., 2006) Breast-Related Symptoms Questionnaire (BRSQ) (Anderson et al., 2006) Michigan Breast Reconstruction Outcomes Study – Satisfaction questionnaire (MBROS-S) (Kerrigan et al., 2001) 7. Michigan Breast Reconstruction Outcomes Study – Body Image Questionnaire (MBROS–BI) (Kerrigan et al., 2002) Of the seven, only one, the Breast-Related Symptoms Questionnaire (BRSQ), had undergone adequate development and validation in breast surgery patients. The BRSQ lists 13 breastrelated symptoms with five levels and the respondent indicates how much of the time she has the symptoms: upper back pain, difficulty finding bras and clothes, headaches, breast pain, lower back pain, rashes under breasts, bra strap grooves, difficulty participating in sports, neck pain, shoulder pain, hard time running, pain in hands, arm pain. From this questionnaire, two scores are derived. The first score is the breast symptom summary score (BSS score), which is calculated by taking the mean scores of all 13 items. The BSS score varies from 0 to 100, with a high score corresponding to fewer and less severe breast symptoms. For the second score, seven items of the 13-item scale are used to provide the physical symptom count.
7
Conclusions
All published studies, irrespective of study design, have demonstrated substantial improvements in QOL in women who undergo reduction mammaplasty. A plethora of instruments have been used to assess quality of life in breast reduction patients. However, very few studies have used breast specific scales, specifically the BSRQ. A combination of generic, breast specific and utility instruments should be used in future studies examining quality of life in these patients. High quality research evidence is needed now to compare the multitude of breast reduction techniques used (i.e., inferior pedicle, vertical scar techniques) and perform costeffectiveness analysis to identify which techniques are cost-effective (Thoma et al., 2008b).
Breast Reduction Surgery and Quality of Life and Clinical Outcomes
133
Summary Points Breast reduction surgery is the most common plastic surgery procedure for breast hypertrophy.
Some insurers and government agencies set arbitrary body weight and/or tissue resection weight restrictions for coverage for breast reduction surgery.
A wide variety health related quality of life instruments, generic and disease or condition specific, have been applied to the area of breast hypertrophy.
Various study designs have been used to assess quality of life in breast reduction surgery patients. Most falling into lower level research evidence (observational studies).
All published studies, irrespective of study design or quality of life measurement tool, have demonstrated substantial improvements in quality of life in women who undergo breast reduction surgery.
References Anderson RC, Cunningham B, Tafesse E, Lenderking WR. (2006). Plast Reconstr Surg. 118: 597. Behmand RA, Tang DH, Smith DJ, Jr. (2000). Ann Plast Surg. 45(6): 575–80. Bell RF, Sivertsen A, Mowinkel P, Vindenes H. (2001). Acta Anaesthesiol Scand. 45: 576–82. Berzon RA. (1998). Understanding and using healthrelated quality of life instruments within clinical research. In: Staquet MJ, Hays RD, Fayers PM (eds.) Quality of Life Assessment in Clinical Trials: Methods and Practice. Oxford University Press, Oxford. Blomqvist L, Eriksson A, Brandberg Y. (2000). Plast Reconstr Surg. 106: 991–997. Blomqvist L, Brandberg Y. (2004). Plast Reconstr Surg. 114: 49–54. Borkenhagen A, Ro¨hricht F, Preiss S, Schneider W, Bra¨hler E. (2007). Ann Plast Surg. 58: 364–370. Cash TF, Duel LA, Perkins LL. (2002). Plast Reconstr Surg. 109: 2112. Cash TF, Pruzinsky T. (1990). Body Images: Development, Deviance, and Change. The Guilford Press, New York. Cella DF, Tulsky DS. (1990). Oncology (Williston Park). 5: 29–38. Chadbourne EB, Zhang S, Gordon MJ, Ro EY, Ross SD, Schnur PL, Schneider-Redden PR. (2001). Mayo Clin Proc. 76: 503–510. Chahraoui K, Danino A, Benony H, Frachebois C, Clerc AS, Malka G. (2006). J Psychosom Res. 61: 801–806. Chao JD, Memmel HC, Redding JF, Egan L, Odom LC, Casas LA. (2002). Plast Reconstr Surg. 110: 1644–1652.
Collins ED, Kerrigan CL, Kim M, Lowery JC, Striplin DT, Cunningham B, Wilkins EG. (2002). Plast Reconstr Surg. 109: 1556–1566. Collis N, McGuiness CM, Batchelor AG. (2005). Br J Plast Surg. 58: 286–289. Cruz-Korchin N, Korchin L. (2003). Plast Reconstr Surg. 112: 1573–1578. Culliford AT, IV, Spector JA, Flores RL, Louie O, Choi M, Karp NS. (2007). Plast Reconstr Surg. 120: 840–844. Feeny D, Furlong W, Torrance GW, Goldsmith CH, Zhu Z, DePauw S, Denton M, Boyle M. (2002). Med Care. 40: 113–128. Freire M, Neto MS, Garcia EB, Quaresma MR, Ferreira LM. (2007). Plast Reconstr Surg. 119: 1149–1156. Furlong WJ, Feeny DH, Torrance GW, Barr RD. (2001). Ann Med. 33: 375–384. Guyatt GH, Feeny DH, Patrick DL. (1993). Ann Intern Med. 118: 622–629. Haines T, McKnight L, Duku E, PerryL, Thoma A. (2008). Clin Plast Surg. 35: 207–214. Hermans BJ, Boeckx WD, De Lorenzi F, van der Hulst RR. (2005). Ann Plast Surg. 55: 227–231. Iwuagwu OC, Bajalan AA, Platt AJ, Stanley PR, Drew PJ. (2005). Ann Plast Surg. 55: 445–448. Iwuagwu OC, Platt AJ, Stanley PW, Hart NB, Drew PJ. (2006a). Plast Reconstr Surg. 118: 1–6. Iwuagwu OC, Walker LG, Stanley PW, Hart NB, Platt AJ, Drew PJ. (2006b). Br J Surg. 93: 291–294. Iwuagwu OC, Stanley PW, Platt AJ, Drew PJ, Walker LG. (2006c). Scand J Plast Reconstr Surg Hand Surg. 40: 19–23. Jones SA, Bain JR. (2001). Plast Reconstr Surg. 108: 62–67.
2285
2286
133
Breast Reduction Surgery and Quality of Life and Clinical Outcomes
Kerrigan CL. (2005). Prospective study of outcomes after reduction mammaplasty (Discussion). Plast Reconstr Surg. 115: 1032. Kerrigan CL, Collins ED, Striplin D, et al. (2001). Plast Reconstr Surg. 108: 1591. Kerrigan CL, Collins ED, Kim HM, et al. (2002). Med Decis Making. 22: 208. Klassen A, Fitzpatrick R, Jenkinson C, et al. (1996). BMJ. 313: 454. McDowell I. (2006). Measuring Health: A Guide to Rating Scales and Questionnaires, 3rd ed. Oxford University Press, Oxford. McGhan Medical Corporation. (1995). Saline-Filled Mammary Implant Augmentation Clinical Study Protocol (Unpublished archival document). ASPS Archives, Arlington Heights, IL. Miller BJ, Morris SF, Sigurdson LL, Bendor-Samuel RL, Brennan M, Davis G, Paletz JL. (2005). Plast Reconstr Surg. 115: 1025–1031. Pusic AL, Chen CM, Cano S, Klassen A, McCarthy C, Collins ED, Cordeiro PG. (2007). Plast Reconstr Surg. 120: 823–837. Rayatt SS, Dancey AL, Jaffe W. (2005). Plast Reconstr Surg. 115: 1605–1608. Saariniemia KM, Sintonen H, Kuokkanen HO. (2008). Scand J Plast Reconstr Surg Hand Surg. 42: 194–198.
Schmitz D. (2005). Insurers tighten restrictions for reduction mammaplasty. Plast Surg News. 1. Spector JA, Karp NS. (2007). Plast Reconstr Surg. 120: 845–850. Spector JA, Rebecca K, Culliford AT, IV, Karp NS. (2006). Plast Reconstr Surg. 117: 374–381. Sprague S, McKay P, Thoma A. (2008). Clin Plast Surg. 35: 195–205. The EuroQOL Group. (1990). Euro QOL – a new facility for the measurement of health-related quality of life. Health Policy 16: 199–208. Thoma A, Cornacchi SD, Lovrics PJ, Goldsmith CH. (2008a). Can J Surg. 51: 215–224. Thoma A, Sprague S, Temple C, Archibald S. (2008b). Clin Plast Surg. 35: 275–284. Thoma A, Sprague S, Veltri K, Duku E, Furlong W. (2005). Health Qual Life Outcomes. 3: 44. Thoma A, Sprague S, Veltri K, Duku E, Furlong W. (2007). Plast Reconstr Surg. 120: 13–26. Torrance GW, Feeny DH, Furlong WJ, Barr RD, Zhang Y, Wang Q. (1996). Med Care, 34: 702–722. Wagner DS, Alfonso DR. (2005). Plast Reconstr Surg. 115: 1034. Ware JE, Jr. (1996). The SF-36 health survey. In: Spilker B. (ed.) Quality of Life and Pharmacoeconomics in Clinical Trials, 2nd ed. Lippincott-Raven Press, Philadelphia, PA, pp. 337–345.
134 Quality of Life and Postoperative Anesthesia in Gastrointestinal Surgery R. Kennelly . A. M. Hogan . J. F. Boylan . D. C. Winter 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2288
2 2.1 2.2 2.2.1 2.2.2
Post Operative Anesthetic Modalities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2289 Systemic Analgesia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2289 Regional Anesthesia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2290 Intraspinal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2290 Peripheral Nerve Blockade . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2291
3 3.1 3.2 3.3
Post Operative Pain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2291 Systemic Morphine (PCA) Versus Epidural Analgesia . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2291 Systemic Morphine (PCA) Versus Intrathecal Morphine . . . . . . . . . . . . . . . . . . . . . . . . . 2296 Peripheral Nerve Block in Gastrointestinal Surgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2296
4
Patient Outcome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2296
5
Quality of Life . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2297
6
Concluding Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2301 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2302
#
Springer Science+Business Media LLC 2010 (USA)
2288
134
Quality of Life and Postoperative Anesthesia in Gastrointestinal Surgery
Abstract: Gastrointestinal surgery impacts considerably on quality of life particularly in the post operative period. The magnitude of this effect is governed by the systemic response to surgery and minimally invasive techniques have allowed for considerable advances in this area. Traditionally, the role of the anaesthetist was limited to pain control. Adequate analgesia is essential for rapid rehabilitation after surgery and is a major factor contributing to patient satisfaction. However, > anesthesia may have a further influence by favorably modulating the systemic response, expediting recovery and improving quality of life. Various anesthetic regimens have been utilized in gastrointestinal surgery. Intraspinal techniques have shown promising results in comparison with more conventional intravenous analgesics but evidence of improved outcome is scarce. Multimodal post operative care has an undoubted positive impact on patient outcome however the influence of the anesthetic component is difficult to ascertain. Some recent trials have indicated that epidural anesthetic techniques improve QOL in the post operative period when compared to more traditional anesthetic techniques. While these results are encouraging, further clinical trials are needed assessing impact of anesthesia on quality of life before recommendations can be made. List of Abbreviations: CEI, continuous epidural infusion; CGQL, Cleveland global quality of life questionnaire; GI, gastrointestinal; IV, intravenous; MASTER, Multicentred Australian Study of Epidural Anesthesia; NSAID, > non-steroidal anti-inflammatory drugs; PCA, patient controlled analgesia; PCEA, patient controlled epidural analgesia; QOL, quality of life; RCT, randomized controlled trial; SF 8, Short form 8 questionnaire; SF 36, Short form 36 questionnaire; VAS, visual analogue scale
1
Introduction
Quality of Life (QOL) and other patient-outcome studies may reveal important differences between treatment options from the perspective of the patient (Coffey et al., 2002; Flynn et al., 2003; Kalbassi et al., 2003; Kell et al., 2003; Winter et al., 2004). These patient centered, evidence-based data are important for planning interventions, including post-operative analgesia. Major surgery imposes pain, physical, mental and physiological stresses that translate into diminished QOL (Wu et al., 2003). The focus on > perioperative management is driven by the concept that modulation of these stressors can impact favorably on postoperative morbidity and therefore improve QOL. It is known that the type of surgical intervention is a major determinant impacting on QOL (az De et al., 2003) and increased understanding of the role of the operation has given rise to interest in muscle sparing incisions and minimally invasive surgical techniques. This concept has formed the basis for numerous investigations into the modulation of inflammatory markers in open vs. laparoscopic procedures (Hill, 2006; Schietroma et al., 2004; Schietroma et al., 2007). The overall impression is that reduction of size of incision reduces the inflammatory stress response. Similarly the role of anesthesia as a purely analgesic intervention is now being questioned. Certainly pain control is essential and remains the most important endpoint when assessing anesthetic techniques however the impact of anesthesia on the metabolic and endocrine consequences of major surgery is now an area of active research (Holte and Kehlet, 2002b). It is important to remember that it is the integration of analgesic care with a perioperative recovery program which emphasizes > minimally invasive surgery, conservative IV fluid
Quality of Life and Postoperative Anesthesia in Gastrointestinal Surgery
134
titration, minimal use of nasogastric intubation and early oral intake and rapid mobility that has brought about the recent advancements in perioperative patient management. Bearing this in mind the anaesthetist has an integral role to play in perioperative patient care. There are numerous pharmaceutical combinations and drug delivery mechanisms present in the modern day anesthetic armentarium. All of these have been used to varying affect in the postoperative setting. The aim of this chapter is to examine the evidence in the published literature of the effectiveness of these modalities in gastrointestinal surgery with particular emphasis on patient outcome and quality of life.
2
Post Operative Anesthetic Modalities
2.1
Systemic Analgesia
Opioid analgesia has been the mainstay of post operative analgesia and indeed is the model against which all techniques are measured. The original method of delivering post operative analgesia involving nurse controlled administration has largely been replaced by patient controlled analgesia (PCA) pumps (> Figure 134-1). A PCA device provides as needed bolus analgesia to the patient. There is a timed lock out function preventing overdose. PCA has been available for almost 25 years and is accepted as the optimum method of systemic opioid delivery after major surgery. It provides good pain control and increases patient satisfaction (Nitschke et al., 1996). However, side effects such as sedation and nausea and vomiting are common and opiate induced respiratory depression can result in poor patient outcome. . Figure 134-1 Patient Controlled Analgesia Device. These devices allow the patient to self administer intravenous anesthetic agents such as opioid medication
2289
2290
134
Quality of Life and Postoperative Anesthesia in Gastrointestinal Surgery
2.2
Regional Anesthesia
2.2.1
Intraspinal
Intraspinal techniques involve the instillation of anesthetic agents (opioids or local anesthetic or a combination) into the epidural (extradural) space or the subarachnoid (intrathecal) space. Due to the anatomy of the spinal cord, intrathecal or spinal anesthesia is only possible below the level of the conus medullaris and therefore is only relevant to low colorectal and perineal procedures. Epidural anesthesia can be used as an anesthetic solution for upper and lower gastrointestinal surgery (> Figure 134-2). Intraspinal anesthesia is widely used
. Figure 134-2 Anatomy of Lumbar Spine, magnetic resonance image (MRI). Arrows identify anatomy of the spinal cord including the specific sites where intraspinal anesthesia is administered (a) Conus medullaris: end of spinal cord (b) Epidural space (c) Subarachnoid space (d) Point of entry for epidural catheter
Quality of Life and Postoperative Anesthesia in Gastrointestinal Surgery
134
and delivers excellent pain relief (Block et al., 2003; Gwirtz et al., 1999). Opioid related side effects tend to be less frequent due to the smaller dose required to provide an analgesic effect however > pruritus and nausea and vomiting, motor block and urinary retention can occur and can cause reduced patient satisfaction (Wu et al., 2001).
2.2.2
Peripheral Nerve Blockade
Paravertebral, intercostal nerve blockade and interpleural instillation of anesthetic agents can be employed to achieve local pain control in the post operative period following esophagectomy (Richardson and Cheema, 2006). More experimental mechanisms of peripheral blockade relevant to abdominal surgery have been described (McDonnell et al., 2007). Evidence of the impact of these techniques on patient outcome and quality of life will be addressed.
3
Post Operative Pain
Many studies have been performed comparing the analgesic effects of differing post operative anesthetic regimes. As a rule, visual analogue scales (VAS) are used to assess pain from the patient’s perspective. The patient is asked to mark their level of pain on a 100 mm line where 0 equals no pain and 100, the worst pain ever experienced (> Figure 134-3). It must be
. Figure 134-3 Visual Analogue Pain Score. This is a common device employed to assess pain from the patient’s perspective
remembered that although pain impacts on QOL it is not the only contributory factor. The measurement of QOL requires application of patient questionnaires that have been psychometrically validated allowing for the condition and the context i.e., the post operative setting. VAS is useful for measurement of trends of pain control only. To extrapolate this data as a surrogate marker of QOL is a gross over simplification only apparent when the VAS is compared with a typical validated QOL instrument (> Figure 134-4). All things considered however, pain control is an integral factor in maintaining QOL and therefore deserves attention.
3.1
Systemic Morphine (PCA) Versus Epidural Analgesia
The efficacy of epidural analgesia in comparison with patient controlled systemic analgesia was examined by > meta-analysis (Block et al., 2003). A total of 100 randomized controlled trials
2291
2292
134
Quality of Life and Postoperative Anesthesia in Gastrointestinal Surgery
. Figure 134-4 Short Form 36 questionnaire (SF-36). An instrument for assessment of quality of life, validated in the postoperative setting
Quality of Life and Postoperative Anesthesia in Gastrointestinal Surgery
. Figure 134-4 (continued)
134
2293
2294
134
Quality of Life and Postoperative Anesthesia in Gastrointestinal Surgery
. Figure 134-4 (continued)
Quality of Life and Postoperative Anesthesia in Gastrointestinal Surgery
. Figure 134-4 (continued)
134
2295
2296
134
Quality of Life and Postoperative Anesthesia in Gastrointestinal Surgery
were included in the study. Epidural techniques provided significantly better pain relief over parenteral delivery with a combination of opioid and local anesthetic providing the best results. A further meta-analysis comparing a patient controlled epidural (PCEA) and a continuous infusion (CEI) (Wu et al., 2005) showed a significant reduction in side effects in the PCEA group with comparable analgesia.
3.2
Systemic Morphine (PCA) Versus Intrathecal Morphine
Intrathecal anesthesia, unlike epidural, is administered as a once only dose. Trials comparing this technique to PCA have reported improved analgesia in the immediate post operative period in the intrathecal group and reduced parenteral morphine consumption (Beaussier et al., 2006; Devys et al., 2003). It is important to note that regardless of these findings there was no difference in incidence of adverse reactions and side effects, nor was there any difference in length of hospital stay. From these studies it can be concluded that intrathecal opioid improves pain scores in the first post operative day in gastrointestinal surgery. The lack of improvement in patient outcome and increased exposure to side effects associated with spinal analgesia would suggest that its effect on patient QOL is questionable.
3.3
Peripheral Nerve Block in Gastrointestinal Surgery
The mainstay of treatment for esophageal cancer is surgery. It is performed either through a combined > thoracotomy and upper abdominal incision or > laparotomy and > trans hiatal resection with a small proportion of centers performing minimally invasive procedures (Bottger et al., 2007; Luketich et al., 2003). There is much discussion regarding the best surgical approach and the role of minimally invasive techniques and there is no doubt that the surgical stress response is modulated to a large extent by these variables. Much research has been conducted into the optimum methods for achieving pain control and preliminary results are promising (McDonnell et al., 2007; Richardson et al., 1995, 1999) No comparison has yet been made with epidural analgesia or intrathecal techniques and so no conclusions can be drawn as to the comparative impact these may have on patient satisfaction. From the current literature it can be concluded that epidural analgesia provides the best pain control in gastrointestinal surgery. A combination of opioid and local anesthetic is the preferred anesthetic regimen for use via the epidural route. Peripheral nerve blocks and intrathecal administration provide good pain relief in the first postoperative day and can be used in conjunction with parenteral morphine. As previously discussed, pain control is important in maintaining patient quality of life however it is not the whole story. Patient morbidity and mortality (known collectively as patient outcome) have a large part to play.
4
Patient Outcome
Gastrointestinal surgery is associated with significant morbidity and mortality which impact on quality of life. Most complications are common to all procedures but their prevalence is
Quality of Life and Postoperative Anesthesia in Gastrointestinal Surgery
134
variable and depends on the site of surgery. Pulmonary complications can occur in any patient post operatively however they are more common in esophageal procedures. A recent survival analysis of a large cohort of esophagectomy patients found that the development of pneumonia was the single greatest factor impacting on mortality (Atkins et al., 2004). > Paralytic ileus occurs to some extent in all intra-abdominal procedures (Delaney et al., 2006). Many studies have looked at the effect of anesthetic regimes on rates of specific complications, the results are conflicting. A large meta-analysis at the start of the new millennium examined the use of epidural analgesia over the previous 30 years (Rodgers et al., 2000). It concluded that epidural analgesia significantly reduced both morbidity and mortality. Unfortunately this work related to a heterogeneous population undergoing all types of operations with widely varying postoperative management plans. These same categorical results have yet to be repeated in the field of gastrointestinal surgery. Two recent reviews (Ballantyne et al., 1998; Holte and Kehlet, 2002a) stated that epidural analgesia reduced pulmonary complications in major abdominal procedures however the majority of the data used to reach this conclusion in both reviews was at least twenty years old and so did not allow for the advances in perioperative management outside the control of the anesthetist. The MASTER anesthesia trial study group (Rigg et al., 2002), found no difference in major morbidity or mortality when comparing epidural analgesia to the more traditional parenteral opioid regime save for a small increase in respiratory failure in the parenteral group. Bearing this in mind, patient outcome as a primary end point has not provided us with very many answers regarding the best anesthetic regimen to use in the post operative setting in gastrointestinal surgery.
5
Quality of Life
Patient outcome is an objective measure of patient wellbeing. It is important to distinguish, however, between outcome and patient reported quality of life when scrutinizing the literature. There is a paucity of data examining the impact of post operative anesthesia on quality of life from the patient’s perspective. It is often assumed that reduction in postoperative complications correlates with improved QOL and logical as this may seem, evidence is needed to support this corollary. Sporadic inclusion of QOL as a secondary endpoint occurs but these studies are often underpowered to draw meaningful conclusions for lesser study objectives (Liu and Wu, 2007). Even as there is little in the published literature regarding the effect of anesthesia on QOL, there is far less concerning gastrointestinal surgery. Unlike traditional outcome measures, post operative quality of life and patient satisfaction can be influenced by such wide ranging factors as gender, age, patient expectations and anxiety preoperatively (Kalkman et al., 2003). QOL cannot be assessed retrospectively as is possible with morbidity or mortality (Wu et al., 2006) (due to recall bias) making QOL a difficult end point for any study demanding real time collection of data and close follow up. As previously mentioned, the heterogeneous nature of many studies and the wide variety of perioperative management strategies create great difficulty in ascertaining the contribution of anesthetic choice to patient quality of life. Even the timing of anesthetic interventions is a possible confounder in the comparison of the effect of different regimes on QOL (Ochroch et al., 2002). Zutshi et al. (2005) examined the differential effect of thoracic epidural vs. intravenous PCA analgesia on QOL post colonic resection. All patients were recruited to fast track
2297
2298
134
Quality of Life and Postoperative Anesthesia in Gastrointestinal Surgery
perioperative management with anesthetic regime representing the only test portion of the care pathway. QOL was assessed using the SF 36 questionnaire at 30 days. Interestingly, no difference in QOL was found between the two groups (> Table 134-1). These results must be qualified by the fact that the primary endpoint for this study was patient length of stay and is underpowered to draw conclusions regarding QOL. The reported QOL data is, therefore, subject to a type 1 error. Our own investigations have shown that post operative epidural analgesia improves QOL in patients undergoing > aero digestive surgery (Winter et al., 2007). Patients were randomized
. Table 134-1 Comparison of outcome for Short Form-36 component scores, Cleveland Global Quality of Life Scores (CGQL), satisfaction survey, and recovery of normal activity for PCA and epidural patients between PCA and epidural PCA Median (Q1, Q3)
Epidural Median (Q1, Q3)
Wilcoxon P value
Discharge Mental component scale
46.9 (38.0, 57.1)
53.6 (37.5, 59.5)
0.38
Physical component scale
41.1 (34.4, 45.8)
32.3 (27.3, 41.1)
0.06
CGQOL
0.63 (0.5, 0.8)
0.40 (0.4, 0.6)
0.043
Happiness
10.0 (9.0, 10.0)
10.0 (9.0, 10.0)
0.73
Hospital satisfaction
9.0 (8.0, 10.0)
9.0 (7.0, 10.0)
0.70
Surgery satisfaction
10.0 (8.0, 10.0)
10.0 (9.0, 10.0)
0.81
Activity percent
50.0 (25.0, 50.0)
40.0 (10.0, 50.0)
0.51
Mental component scale
52.8 (42.6, 58.3)
56.0 (43.6, 58.0)
0.53
Physical component scale
30.1 (25.0, 42.6)
32.0 (28.6, 36.9)
0.91
CGQOL
0.60 (0.5, 0.8)
Day 10
0.50 (0.3, 0.7)
0.15
Happiness
9.0 (8.0, 10.0)
8.0 (6.0, 9.0)
0.32
Hospital satisfaction
8.5 (4.0, 9.0)
8.0 (7.0, 10.0)
0.42
Surgery satisfaction
9.0 (9.0, 10.0)
9.0 (8.0, 10.0)
0.71
50.0 (30.0, 50.0)
40.0 (25.0, 50.0)
0.38
Mental component scale
49.2 (36.4, 60.2)
43.6 (31.9, 50.8)
0.16
Physical component scale
35.8 (30.1, 42.8)
37.5 (32.8, 42.6)
0.74
CGQOL
0.75 (0.6, 0.8)
0.60 (0.4, 0.8)
Activity percent Day 30
0.063
Happiness
9.0 (7.5, 10.0)
8.0 (7.0, 10.0)
0.59
Hospital satisfaction
9.0 (5.0, 9.5)
8.0 (5.5, 9.0)
0.57
9.0 (8.0, 9.0)
9.0 (7.0, 10.0)
0.86
70.0 (55.0, 80.0)
57.5 (35.0, 75.0)
0.35
Surgery satisfaction Activity percent
Reprinted from Zutshi M. et al., Am J Surg. 189. ((Copyright 2005), with permission from Elsevier, Figures are medians with 1st and 3rd quartiles). Patient cohort = 34. Control: n = 20. Test group: n = 14
Quality of Life and Postoperative Anesthesia in Gastrointestinal Surgery
134
to receive either PCA delivered systemic opioid or thoracic epidural infusion of local anesthetic/ opioid combination. Using the validated instruments short form questionnaire 8 (SF-8) and short form questionnaire 36 (SF-36), we showed that thoracic epidural provided superior QOL at all time points (> Figure 134-5). A questionnaire designed to assess the patient’s self caring ability from a nursing perspective showed that epidural analgesia allowed increased mobility and independence despite the extra equipment. Patient satisfaction was also assessed and this mirrored the QOL results. VAS was applied as a measure of pain as a secondary endpoint and pain scores closely correlated with patient reported QOL. A randomized trial of open colon resection patients found similar improvements in QOL as a result of epidural anesthesia (Carli et al., 2002). Using the SF-36, better scores were noted immediately post operatively and persisted for 6 weeks. There was a concomitant improvement in pain scores, mobility and a faster return of bowel function (> Table 134-2). Notwithstanding these findings, it would be simplistic to claim that anesthesia is the sole variable that has significant impact on QOL in the post operative patient. Simple interoperative variability such as differences in placement of intercostal sutures has been shown to have a significant impact on post operative pain. Differences in surgical access such as muscle sparing thoracotomy and minimally invasive techniques play a large role in post operative recovery by modulating the surgical stress response. How and ever, there is little doubt that anesthetic management can greatly influence the patient experience in the post operative setting (> Table 134-3). . Figure 134-5 Mean physical and mental quality of life scores for epidural and PCA groups. Data from Prof. Winter (unpublished results). Values represent means with 95% confidence intervals. Increasing scores indicate increasing quality of life. Total patient cohort = 60. Control group: n = 30. Test group: n = 30. PCS and MCS indicate physical and mental scores on the quality of life scoring instruments. { P value < 0.001. SF-8 Short form 8 questionnaire; SF-36 Short form 36 questionnaire
2299
2300
134
Quality of Life and Postoperative Anesthesia in Gastrointestinal Surgery
. Table 134-2 Difference from Baseline Values of Functional Exercise Capacity and Health-related Quality of Life at 3 and 6 Weeks after Surgery PCA Group (n = 31) Mean
Epidural Group (n = 32) SD
Mean
SD
Significant Effect
6-minute walking test (m) 3 wk
62.9
74.5
32.0
62.6
Time x group; P = 0.0005
6 wk
21.7
48.3
5.0
59.0
–
3 wk
13.5
10.0
11.2
11.3
Time x group; P = 0.0001
6 wk
5.5
9.7
3.8
10.4
–
3 wk
3.1
12.6
5.0
11.6
Group; P = 0.002
6 wk
3.3
13.9
8.7
12.7
–
3 wk
29.4
24.3
23.7
25.1
Time x group; P = 0.0001
6 wk
10.5
17.3
6.9
26.0
–
3 wk
63.4
44.9
36.3
58.4
Group; P = 0.0109
6 wk
50.0
54.3
9.7
51.5
Time x group; P = 0.0003
3 wk
22.62
50.56
6.45
51.94 Group; P = 0.0023
6 wk
26.44
54.47
19.36
44.54 –
3 wk
19.8
33.4
12.1
30.8
Time x group; P = 0.0001
6 wk
0.1
26.1
5.4
31.5
–
3 wk
4.2
18.5
2.9
19.8
Time x group; P = 0.0022
6 wk
0.1
18.7
3.9
17.1
–
3 wk
17.3
21.5
1.6
25.8
Group; P = 0.0093
6 wk
4.5
17.1
8.2
22.4
Time x group; P = 0.0001
Physical health
Mental health
SF-36 subscales Physical functioning
Role – physical
Role – emotional
Bodily pain
General health
Vitality
134
Quality of Life and Postoperative Anesthesia in Gastrointestinal Surgery
. Table 134-2 (continued) PCA Group (n = 31)
Epidural Group (n = 32)
Social functioning 3 wk
25.0
30.6
11.3
26.5
Group; P = 0.0223
6 wk
16.8
33.1
5.6
40.2
Time x group; P = 0.0078
3 wk
4.7
26.2
5.6
22.8
Group; P = 0.0496
6 wk
0.7
26.7
13.8
22.8
Time x group; P = 0.0084
Mental health index
Figures represent mean with standard deviation (SD). Increasing numbers indicate improved quality of life P values shown for significant results. Patient cohort = 63. Control group: n = 31. Test group: n = 32. PCA patient-controlled analgesia; SF-36 Short Form 36; Wk Week)
. Table 134-3 A summary of randomised controlled trials (RCTs) examining the effect of post operative anesthesia on Quality of Life (QOL) in gastrointestinal surgery
Author
n
Type of surgery
Carli et al Ref 7
63
Winter et al Ref 31 Zutshi et al Ref 37
Anesthetic regimes
Quality of Life Instrument(s)
Control
Test
Findings
Abdominal
IV PCA n = 31
Thoracic Epidural n = 32
SF-36 assessed preoperatively and at 3 and 6 weeks post surgery
Significant improvement in QOL in epidural group at all time points
60
Thoracic
IV PCA n = 30
Thoracic Epidural n = 30
SF-8 (24 hours post surgery) and SF-36 (day 7 post surgery)
Significant improvement in QOL inepidural group
34
Abdominal
IV PCA n = 20
Thoracic Epidural n = 14
SF-36 and CGQL assessed preoperatively and day 10 and day 30 post surgery
No difference detected
IV PCA Intravenous patient controlled analgesia; SF-36 Short form 36 questionnaire; SF-8 Short form 8 questionnaire; CGQL Cleveland global quality of life questionnaire
6
Concluding Remarks
Patient wellbeing relies on many factors including intraoperative, perioperative and genetic variables and anesthesia is central to these processes. While traditional calculation of morbidity may not reflect this reality, patient centered measures of QOL highlight the importance of
2301
2302
134
Quality of Life and Postoperative Anesthesia in Gastrointestinal Surgery
appropriate anesthesia in the postoperative period. Intraspinal anesthesia and particularly epidural delivery results in optimum pain relief and patient satisfaction. Large scale prospective trials will clarify the exact benefit of different modalities of anesthetic delivery on patient mobility, psychological well being and overall quality of life.
Summary Points The major factor impacting on patient outcome in gastrointestinal surgery is the systemic response to injury.
A multidisciplinary approach (anesthesia, surgery, nursing staff and physiotherapy) is needed to optimize patient outcome.
Pain, patient outcome and quality of life are discrete measurements of patient management and are not interchangeable.
Intraspinal analgesia provides good pain relief and can have an opioid sparing effect however, the side effect profile can be poor.
It is not definitively shown that intraspinal analgesia shortens inpatient stay or reduces complication rates or shortens length of hospital stay.
There are very little data regarding the impact of anesthesia on post operative quality of life.
References Atkins BZ, Shah AS, Hutcheson KA, Mangum JH, Pappas TN, Harpole DH Jr, D’Amico TA. (2004). Ann Thorac Surg. 78(4): 1170–1176. az De LA, Oteiza MF, Ciga MA, Aizcorbe M, Cobo F, Trujillo R. (2003). Br J Surg. 90(1):91–94. Ballantyne JC, Carr DB, deFerranti S, Suarez T, Lau J, Chalmers TC, Angelillo IF, Mosteller F. (1998). Anesth Analg. 86(3): 598–612. Beaussier M, Weickmans H, Parc Y, Delpierre E, Camus Y, Funck-Brentano C, Schiffer E, Delva E, Lienhart A. (2006). Reg Anesth Pain Med. 31(6): 531–538. Block BM, Liu SS, Rowlingson AJ, Cowan AR, Cowan JA Jr, Wu CL. (2003). JAMA. 290(18): 2455–2463. Bottger T, Terzic A, Muller M, Rodehorst A. (2007). Surg Endosc. 21(10): 1695–1700. Carli F, Mayo N, Klubien K, Schricker T, Trudel J, Belliveau P. (2002). Anesthesiology. 97(3): 540–549. Coffey JC, Winter DC, Neary P, Murphy A, Redmond HP, Kirwan WO. (2002). Dis Colon Rectum. 45(1): 30–38. Delaney CP, Senagore AJ, Viscusi ER, Wolff BG, Fort J Du, W, Techner L, Wallin B. (2006). Am J Surg. 191(3): 315–319. Devys JM, Mora A, Plaud B, Jayr C, Laplanche A, Raynard B, Lasser P, Debaene B. (2003). Can J Anaesth. 50(4): 355–361.
Flynn MJ, Winter DC, Breen P, O’Sullivan G, Shorten G, O’Connell D, O’Donnell A, Aherne T. (2003). Eur J Cardiothorac Surg. 24(4): 547–551. Gwirtz KH, Young JV, Byers RS, Alley C, Levin K, Walker SG, Stoelting RK. (1999). Anesth Analg, 88(3): 599–604. Hill AG. (2006). Br J Surg. 93(4): 504–505. Holte K, Kehlet H. (2002a). Minerva Anestesiol. 68(4): 157–161. Holte K, Kehlet H. (2002b), Clin Nutr. 21(3): 199–206. Kalbassi MR, Winter DC, Deasy JM. (2003). Dis Colon Rectum. 46(11): 1508–1512. Kalkman CJ, Visser K, Moen J, Bonsel GJ, Grobbee DE, Moons KG. (2003). Pain. 105(3): 415–423. Kell MR, Power K, Winter DC, Power C, Shields C, Kirwan WO, Redmond HP. (2003). Ir J Med Sci. 172(2): 63–65. Liu SS, Wu CL. (2007). Anesth Analg. 105(3): 789–808. Luketich JD, velo-Rivera M, Buenaventura PO, Christie NA, McCaughan JS, Litle VR, Schauer PR, Close JM, Fernando HC. (2003). Ann Surg. 238(4): 486–494. McDonnell JG, O’Donnell B, Curley G, Heffernan A, Power C, Laffey JG. (2007). Anesth Analg. 104(1): 193–197. Nitschke LF, Schlosser CT, Berg RL, Selthafner JV, Wengert TJ, Avecilla CS. (1996). Arch Surg. 131(4): 417–423.
Quality of Life and Postoperative Anesthesia in Gastrointestinal Surgery Ochroch EA,Gottschalk A, Augostides J, Carson KA, Kent L, Malayaman N, Kaiser LR, Aukburg SJ. (2002). Anesthesiology. 97(5): 1234–1244. Richardson J, Cheema S. (2006). Br J Anaesth. 96(4): 537. Richardson J, Sabanathan S, Jones J, Shah RD, Cheema S, Mearns AJ. (1999). Br J Anaesth. 83(3): 387–392. Richardson J, Sabanathan S, Mearns AJ, Shah RD, Goulden C. (1995). Br J Anaesth. 75(4): 405–408. Rigg JR, Jamrozik K, Myles PS, Silbert BS, Peyton PJ, Parsons RW, Collins KS. (2002). Lancet. 359(9314): 1276–1282. Rodgers A, Walker N, Schug S, McKee A, Kehlet H, van ZA, Sage D, Futter M, Saville G, Clark T, Macmahon S. (2000). BMJ. 321(7275): 1493. Schietroma M, Carlei F, Cappelli S, Pescosolido A, Lygidakis NJ, Amicucci G. (2007). Hepatogastroenterology. 54(74): 342–345. Schietroma M, Carlei F, Franchi L, Mazzotta C, Sozio A, Lygidakis NJ, Amicucci G. (2004). Hepatogastroenterology. 51(60): 1595–1599.
134
Winter DC, Dozois EJ, Pemberton JH. (2007). Curr Drug Ther. 2: 75–77. Winter DC, Murphy A, Kell MR, Shields CJ, Redmond HP, Kirwan WO. (2004). Dis Colon Rectum. 47(5): 697–703. Wu CL, Cohen SR, Richman JM, Rowlingson AJ, Courpas GE, Cheung K, Lin EE, Liu SS. (2005). Anesthesiology. 103(5): 1079–1088. Wu CL, Naqibuddin M, Fleisher LA. (2001). Reg Anesth Pain Med. 26(3): 196–208. Wu CL, Naqibuddin M, Rowlingson AJ, Lietman SA, Jermyn, RM, Fleisher LA. (2003). Anesth Analg. 97(4): 1078–1085. Wu CL, Rowlingson AJ, Herbert R, Richman JM, Andrews RA, Fleisher LA. (2006). J Clin Anesth. 18(8): 594–599. Zutshi M, Delaney CP, Senagore AJ, Mekhail N, Lewis B, Connor JT, Fazio VW. (2005). Am J Surg. 189(3): 268–272.
2303
135 Health-Related Quality of Life After Surgery for Crohn’s Disease M. Scarpa . I. Angriman 1 Surgery for Crohn’s Disease: for Whom, Why, When and How . . . . . . . . . . . . . . . . . . . 2306 2 Why Measure Quality of Life after Surgery for Crohn’s Disease . . . . . . . . . . . . . . . . . . 2308 3 Preoperative HRQL: Concerns about Surgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2310 4 Early Postoperative HRQL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2311 5 Long-Term Postoperative HRQL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2314 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2317
#
Springer Science+Business Media LLC 2010 (USA)
2306
135
Health-Related Quality of Life after Surgery for Crohn’s Disease
Abstract: Crohn’s disease cannot be healed but just taken into remission. Intestinal obstruction and fistulization, lack of response to medical management and perianal disease are the most frequent indication for intestinal surgery. Why is it necessary to measure quality of life after surgery for Crohn’s disease? Firstly, morbidity and mortality provide a partial and, very often, incomplete picture of outcome. Secondarily, nowadays, indications for surgery for Crohn’s disease are broader and not limited to life saving procedures but in many cases they include chronic conditions such as failure of medical therapy, or poor quality of life on it self. Finally, quality of life is a more patient orientated measure of outcome that can give the patients’ point of view about the procedure that is proposed. HRQL is a multi-dimensional concept which includes several dimensions based on biological and symptom variables. Disease-related worries and concerns about the disease on itself and its therapy reflect one of these dimensions and they are considered to be a major determinant of HRQL in patients with IBD. In fact, concerns about having surgery and having an ostomy bag have a relevant impact on HRQL of Crohn’s disease patients and having surgery increases concerns about body stigma. The early impact of surgery on HRQL is an important component of the patient’s decision regarding immediate and future surgery and understanding his or her recovery. Obviously, HRQL is expected to improve after operative procedures. In effect, in most of the studies, a significant improvement in HRQL early in the postoperative period was observed. Improvement, apparently, occurred irrespective of the disease activity measured with CDAI, the indication for surgery, type of procedure (abdominal or perineal), and history of previous surgery. On the contrary, the long-term impact of surgery on HRQL is more controversial. Some studies, mainly those performed with generic questionnaires, reported an improved HRQL while other (those performed with disease specific instruments) described a decreased HRQL. According to these authors, HRQL, apparently, depends mainly on the long-term disease activity.
1
Surgery for Crohn’s Disease: for Whom, Why, When and How
As well explained in the previous chapter, Crohn’s disease is a chronic, transmural inflammatory disease of the gastrointestinal tract of unknown ethiology that can involve any part of the alimentary tract from the mouth to the anus but most commonly affects the small intestine and colon. Its most common clinical manifestations are abdominal pain, diarrhoea, and weight loss and it can be complicated by intestinal obstruction or localized perforation with fistula formation. Nowadays, either medical therapy or surgical treatments are palliative: Crohn’s disease cannot be healed but just taken into remission; in fact, operative therapy can provide effective symptomatic relief for those patients with complications from Crohn’s disease and can produce a reasonable long-term benefit. The great part of patients with Crohn’s disease requires surgery some time during the course of their illness. In patients with more than 20 years of disease, the cumulative probability of surgery is estimated to be 78%. The indications for operation are limited to complications that include intestinal obstruction, intestinal perforation with fistula formation or abscess, free perforation, gastrointestinal
Health-Related Quality of Life after Surgery for Crohn’s Disease
135
bleeding, urologic complications, cancer, and perianal disease (Delaney and Fazio, 2001). Since there is no curative intent, surgery should be specifically directed to the complication and only the segment of bowel involved in the complicating process should be resected. In fact, wide resections give no further benefit and can lead to the short bowel syndrome. Intestinal obstruction and perforation are the main intestinal complications of Crohn’s disease and therefore the most frequent indication for intestinal surgery. Obstruction is usually caused by chronic fibrotic lesions, which eventually narrow the lumen of the bowel, producing partial or near-complete obstruction. Free perforations into the peritoneal cavity leading are a rare presentation in patients with Crohn’s disease. More commonly, fistulas occur between the sites of perforation and adjacent organs, such as loops of small and large intestine, the urinary bladder, the vagina, the stomach, and sometimes the skin, usually at the site of surgical scar such as the site of the previous appendectomy or laparotomy (Delaney and Fazio, 2001). Localized abscesses often occur near the sites of perforation. Rarely, Crohn’s colitis may result in toxic megacolon. Perianal disease (fissure, fistula, stricture, or abscess) can occur in up to 48% of patients and it can affect their sexual life. Perianal disease may be the sole presenting feature in 5% of patients and may precede the onset of intestinal disease. Extra intestinal manifestations of Crohn’s disease may be present in 30% of patients and the most common ones are skin lesions, such as erythema nodosum and pyoderma gangrenous, arthritis and joint pain, uveitis and iritis, hepatitis and pericholangitis, and aphthous stomatitis. These extra intestinal manifestations can heavily affect quality of life but they may persist after surgery. Finally, long-standing Crohn’s disease predisposes to cancer of both the small intestine and colon. Recent evidence indicates that the risk of cancer in Crohn’s disease of the colon is at least as great as that in ulcerative colitis and this could be a major concern for patients. Common indications for surgery in Crohn’s disease patients are shown in > Figure 135-1. Intestinal obstruction is the most common indication for surgical therapy in patients with Crohn’s disease. Surgery is required in case of complete obstruction and in patients with partial obstruction whose condition does not resolve with non operative management. The treatment of choice of intestinal obstruction caused by Crohn’s disease is segmental resection . Figure 135-1 Indications for surgery in patients with Crohn’s disease (personal data Scarpa and Angriman, 2008)
2307
2308
135
Health-Related Quality of Life after Surgery for Crohn’s Disease
of the involved segment with primary re-anastomosis. In selected patients with obstruction caused by strictures (either single or multiple), one option is to perform a strictureplasty that effectively widens the lumen but avoids intestinal resection (Delaney and Fazio, 2001). Strictureplasty has the most application in those patients in whom multiple short areas of narrowing are present over long segments of intestine, in those patients who have already had several previous resections of the small intestine, and when the areas of narrowing are due to fibrous obstruction rather than acute inflammation. This procedure preserves intestine and is associated with complication and recurrence rates comparable to resection and re-anastomosis. The role of laparoscopic surgery for patients with Crohn’s disease has not been clearly defined. In appropriately selected patients, for example those with single or multiple stenosis, localized abscesses, simple intra-abdominal fistulas, and peri-anastomotic recurrent disease, this technique appears feasible and safe and clearly improves body image of these patients. Indications for surgery also include a lack of response to medical management or complications of Crohn’s colitis, which include obstruction, hemorrhage, perforation, and toxic megacolon. Depending on the diseased segments, operations commonly include segmental colectomy with colo-colonic anastomosis, subtotal colectomy with ileo-rectal anastomosis, and in patients with extensive perianal and rectal disease, total proctocolectomy with Brooke ileostomy. A particularly troubling problem after proctocolectomy in patients with Crohn’s disease is delayed healing of the perineal wound. Although controversial, continence-preserving operations, such as ileoanal pouch anastomosis or continent ileostomies (Kock pouch) are proscribed for patients with Crohn’s colitis because of the high rate of recurrence of Crohn’s disease in the pouch. The treatment of perianal disease should be conservative. Even if surgery for Crohn’s disease provide patients with often significant symptomatic relief it is not curative and recurrence rates are reported as high in most series (Wolff, 1998). Endoscopic evidence of recurrence is detected in approximately 70% of patients within 1 year of surgery and reoperation, rates are 25–30% at 5 years and 40–50% at 20 years. Approximately 45% of patients will ultimately require a second operation, of which only 25% will require a third operation. Despite the risk of recurrence, many patients who have had surgery for Crohn’s disease wish that they had had their operation sooner. The overwhelming majority of such patients report relief of symptoms after surgery, restoration of a feeling of well-being and the ability to eat normally, and a reduction in the need for medical therapy.
2
Why Measure Quality of Life after Surgery for Crohn’s Disease
As exhaustively described in the previous chapter, Crohn’s disease affect quality of life since it is a chronic illness with many relapses and symptoms that interfere with daily activities of patients (Kirshner, 1980), who are often young and with a subsequently long life expectancy (Guassora et al., 2000). Patients with Crohn’s disease generally experience worse quality of life than those with ulcerative colitis (Casellas et al., 2001). As would be expected, quality of life is worse during relapse than remission, but even when in remission studies have found patients’ quality of life is poorer than that of healthy controls (Casellas et al., 2001; Drossman et al., 1989). Quality of life in Crohn’s disease is adversely affected by smoking and taking steroids and it is worse in women (Blondel-Kucharski et al., 2001). In such a scenario, the
Health-Related Quality of Life after Surgery for Crohn’s Disease
135
measurement of health related quality of life (HRQL) is particularly relevant in these patients. But why is it even more necessary to measure quality of life after surgery in these patients? The first reason is that traditional outcome measures for assessing surgical procedure such as morbidity and mortality provide a partial and, very often, incomplete picture. In the first half of the twentieth century these measures were appropriate because most gastrointestinal surgical procedures were associated with high complication and operative mortality rates. In the first cohort of patients who underwent abdominoperineal resection, for example, the operative mortality was 32% while today the accepted operative mortality is under 5%. In such a scenario, obviously, survival was the main measure to assess a surgical procedure’s success and, consequently, the decision to adopt an operative technique was largely based on operative mortality and on long-term survival (McLeod, 1999). Today, operative mortality rates for intestinal surgery for Crohn’s disease is under 1% [personal data], so, fortunately, operative mortality is of limited use as an outcome measure to discriminate between two surgical techniques or determine the value of a surgical technique as compared with medical therapy (McLeod, 1999). Therefore, HRQL can be used to integrate the traditional outcome measure to assess therapeutic efficacy of a surgical procedure in patients with CD (Irvine, 1994) and to identify treatment strategies that are preferable from the point of view of patients. Furthermore, quality of life represents an indispensable index to evaluate the success of a procedure not only in the short run but also in a longer period of time (Casellas et al., 1999; McLeod and Baxter, 1998). A second reason is that, nowadays, indications for surgery for Crohn’s disease are broader. They are not limited to life saving procedures such as in case of intestinal bleeding, peritoneal sepsis or obstruction but in many cases they include chronic conditions such as failure of medical therapy, or poor quality of life on it self. The adoption of laparoscopic approach has largely minimized the disability associated with surgery, thus, in some cases, surgery may be considered as an alternative to medical therapy. In such situation, where the indication for surgery is poor quality of life, it is thus essential that quality of life should be measured before and after the procedure to determine whether the therapeutic intervention has been worthwhile (McLeod, 1999). Finally, the goal of surgery in Crohn’s disease is not to cure the disease but to improve quality of life, and thus surgical procedures should be evaluated in terms of their impact on HRQL of these patients. The third reason for measuring quality of life in patients with Crohn’s disease after surgery is because it is a more patient orientated measure of outcome that can give the patients’ point of view about the procedure that is proposed (McLeod, 1999). Although physiologic outcomes are easier to measure, they may not necessarily correlate with patients’ perception of their status. Physiologic outcomes provide information to the clinicians but may be of limited interest to patients and often correlate poorly with well being (McLeod, 1999). For example, after reconstructive surgery for ulcerative colitis surgeons have generally tended to assess outcome in terms of stool frequency. However, evidence suggests that this criterion does not correlate well with patients’ perceived quality of life and satisfaction with the outcome of their surgery (McLeod, 1999). Furthermore, the heterogeneity of Crohn’s disease manifestations make it difficult to use a single symptom to measure disease severity, so many indexes, such as the Crohn’s Disease activity Index or the Perianal Crohn’s Disease Activity Index, have been developed to assess function or disease activity. Anyway, measurement of functional status alone is of limited value because only the physical domain is assessed in these indexes and other domains, such as the psycho-social, which have an impact on quality of life, are not included. Several studies investigated the impact of surgery on the HRQL of CD patients submitted to intestinal surgery (Casellas et al., 2000; Delaney et al., 2003; Maartense et al., 2006;
2309
2310
135
Health-Related Quality of Life after Surgery for Crohn’s Disease
Scarpa et al., 2007a; Scott and Hughes, 1994; Thaler et al., 2005; Thirlby et al., 1998, 2001; Tillinger et al., 1999; Yazdanpanah et al., 1997). Some studies were focused on the early post operative outcome or on the comparison between the preoperative and the post operative HRQL (Casellas et al., 2000; Delaney et al., 2003; Maartense et al., 2006; Scott and Hughes, 1994; Tillinger et al., 1999; Yazdanpanah et al., 1997). All these studies demonstrate that, although undergoing surgery seems to be a major preoperative concern (Drossman et al., 2001; Canavan et al., 2006), in the months immediately following surgery patients affected by CD show a significant increase of their HRQL (Casellas et al., 2000; Delaney et al., 2003; Maartense et al., 2006; Scott and Hughes, 1994; Tillinger et al., 1999; Yazdanpanah et al., 1997). Fewer studies assessed the HRQL outcome after a long-term follow up and most of them did it either with a disease specific questionnaire or a generic tool. Therefore, the different studies obtained different results, making this point the most controversial one (Casellas et al., 2000; Scarpa et al., 2007a; Thaler et al., 2005; Thirlby et al., 2001). In fact, the authors of the Seattle University, using a modified version of the generic Short Form 36 (SF36), claimed that the quality of life of CD patients after ileo-colic resection was improved in long term follow up (Thirlby et al., 1998, 2001). On the contrary, the other authors who used respectively Gastro Intestinal Quality of Life Index (GIQLI), Inflammatory Bowel Disease Questionnaire (IBDQ) and Padova Inflammatory Bowel Disease Quality of Life (PIBDQL) score, reported a long term decrease of HRQL in these patients (Casellas et al., 2000; Scarpa et al., 2007a; Thaler et al., 2005). Therefore, the analysis of quality of life after surgery for Crohn’s disease is mandatory if we want to answer correctly to the main question that patients ask when they are proposed the operation: ‘‘How will I feel after the operation; what will my life be like?’’ These sorts of questions are even more crucial if we propose operation to patients just for poor quality or for failure of medical therapy.
3
Preoperative HRQL: Concerns about Surgery
HRQL is a multi-dimensional concept which includes several dimensions based on biological and symptom variables. Disease-related worries and concerns about the disease on itself and its therapy reflect one of these dimensions and it is considered to be a major determinant of HRQL in patients with IBD (Hjortswang et al., 1999). Descriptive research into the concerns of patients with IBD has been advanced by the development of the Rating Form of IBD Patients’ Concerns (RFIPC) by Drossman et al. (1991), which quantifies the degree of concerns with specific issues related to IBD. Different studies in various samples of IBD patients revealed a high concordance in the ranking of worries assessed by the RFIPC (De Rooy et al., 2001; Maunder et al., 1999; Moser et al., 1995). These major concerns are similar across different cultures (Levenstein et al., 2001). Thus, it is important for clinicians to be aware of them so that they take them in account when Crohn’s disease patients have to be counseled for a surgical therapy. In a recent German study which used RFIPC, the top ranked concerns were as follows: effects of medication, having an ostomy bag, uncertain nature of disease, being a burden on others, energy level, loss of bowel control, having surgery, achieving full potential, attractiveness, developing cancer, and feelings about the body (Mussell et al., 2004). In these patients population the concern about having surgery was at the sixth place. The ranking of concerns was consistent with that reported in previous studies (De Rooy et al., 2001; Moser et al., 1995)
Health-Related Quality of Life after Surgery for Crohn’s Disease
135
and in a cross-cultural validation study (Levenstein et al., 2001). These results for the RFIPC total score and the sub-scores were also similar to those reported from another outpatient study that used the same German questionnaire (Moser et al., 1995). Nevertheless, in a French study, the concern of having surgery was at the second place of the in the ranking of the concerns. In fact, this study reported that the most important fears expressed by the patients affected by Crohn’s disease (noted on a scale from 0 – no fear – to 100 – maximal fear –), were: ‘‘lack of energy’’ (65.0%), ‘‘having an intestine operation’’ (64.6%), ‘‘having an ostomy bag’’ (63.8%) and ‘‘the unpredictable nature of the disease’’ (62.8/100) (Etienney et al., 2004). On the contrary, according to a recent British study (Canavan et al., 2006), the most important concerns for patients were the uncertain nature of the disease, adverse effects of medication, having to use an ostomy bag, low-energy levels and the possible need for surgery which was only at the fifth place (Drossman et al., 1989; Levenstein et al., 2001). Diseaserelated concerns and anxieties regarding body image were rated highest in all subgroups, whilst sexual concerns were lowest (Canavan et al., 2006). These concerns did not change with disease duration but changed before and after the therapy: patients who had surgery reported a worse QoL, showed the highest level of concern regarding body image whereas the greatest concerns of patients who had not undergone surgery were about disease complications (Canavan et al., 2006). Among the individual concerns that were greatest across all patient groups, the possible need for an ostomy bag seem to have a great impact on their quality of life (Canavan et al., 2006). In fact, this concern has also been rated amongst the most important in other studies of patients with Crohn’s disease (Blondel-Kucharski et al., 2001; Drossman et al., 1989; Levenstein et al., 2001). It is possible that concerns about ostomy bags could be reduced by better patient education about the likely need for such surgery and what it would entail. In conclusion, concerns about having surgery and having an ostomy bag have a relevant impact on HRQL of Crohn’s disease patients and having surgery increases concerns about body stigma. Nevertheless, a high percentage of patients who actually had operations for CD felt they would be willing to go through surgery again (Delaney et al., 2003) and, also the study by Scott et al. reported that patients with CD with a previous ileocolic resection would have liked to have had their operation earlier (Scott and Hughes, 1994). Therefore, as the possibility of future surgery could be a major concern for these patients, the results of these two studies may help CD patients to understand that surgical intervention can contribute to improvement in HRQL and it may be possible to reduce level of concern regarding body stigma by counseling patients about their surgery preoperatively. Ranking of need of surgery and need of ostomy bag among worries assessed by the RFIPC in different studies is shown in > Table 135-1.
4
Early Postoperative HRQL
The early impact of surgery on HRQL is an important component of the patient’s decision regarding immediate and future surgery and understanding his or her recovery. Obviously, HRQL is expected to improve after operative procedures. However, in CD patients, the role of surgical treatment is merely symptom control rather than cure of disease, and patients are discharged form the surgical ward still on medications that can produce a confounding effect. So, surgery is sometimes perceived as a detrimental step, and this perception may result in
2311
2312
135
Health-Related Quality of Life after Surgery for Crohn’s Disease
. Table 135-1 Ranking of need of surgery and need of ostomy bag among worries assessed by the RFIPC in different studies
Study
Year Patients Country
Most important concern
‘‘Possible need for surgery’’ ranking
‘‘Having an ostomy bag’’ ranking
Blondel2001 Kucharski et al.
231
France
having an ostomy bag
Fourth
First
Etienney et al.
2004
141
France
energy level
Second
Third
Mussell et al.
2004
47
Sixth
Second
Canavan et al.
2006
221
Germany effect of medications UK
uncertain Fifth nature of the disease
Third
procrastination on the part of the patient and the clinician, with prolonged continuation of (sometimes) inappropriate non surgical care (Delaney et al., 2003). Several studies have reported the effect of surgery for CD on HRQL, although many have been retrospective and based on small numbers of patients. A critical review from the group of Toronto evaluated earlier studies on the effect of surgery on HRQL in patients with IBD (Maunder et al., 1995). They evidenced that the earlier studies were uncontrolled and used semi quantitative measures of HRQL so they were scarcely conclusive (Bechi and Tonelli, 1982; Lindhagen et al., 1983; Scott and Hughes, 1994). Other studies were found to be retrospective comparison between current and preoperative function (Cooper et al., 1986; Meyers, 1983). Nevertheless, successive retrospective studies reported an improved HRQL in patients undergoing surgery for CD (Casellas et al., 2000; Meyers et al., 1980; Nissan et al., 1997), and a comparable improvement in HRQL, irrespective of the procedure performed (Broering et al., 2001). Some prospective studies investigated the effect of surgery and other treatment on HRQL as measured by various instruments after 3 months in the postoperative period. Thirlby and colleagues measured HRQL of a group of 36 CD patients with the Health Status Questionnaire and they reported an improvement 3 months after surgery (Thirlby et al., 1998), while Yazdanpanah and associates observed a significant improvement of the HRQL in 26 patients undergoing surgery at 3 and 6 months in the postoperative period (Yazdanpanah et al., 1997). Another prospective study from the University of Vienna measured HRQL by Crohn’s Disease Activity Index, Time Trade-Off technique, Direct Questioning of Objectives and the RFIPC in 16 patients at three 1-month intervals in the postoperative period and found it to be improved significantly at 3 and 6 months after surgery (Tillinger et al., 1999). However, after 24 months, the long term improvement of the quality of life was reported only for the 12 patents with CD in remission. Thus, it is difficult to attribute the good quality of life to surgery instead of disease activity. Three prospective studies specifically investigated the effect of surgery on early postoperative HRQL of CD patients. Blondel-Kucharski and colleagues reported that recent surgery impaired HRQL but it is not clear from their report what the mean duration of follow up after
Health-Related Quality of Life after Surgery for Crohn’s Disease
135
surgery when measurement of HRQL occurred (Blondel-Kucharski et al., 2001). In addition, the sample size of their study population in whom the early postoperative change in HRQL was measured was quite small. In the another study, from the Cleveland Clinic, the effect of surgery were prospectively evaluated on early postoperative HRQL using CGQL, which had initially been validated in patients with ulcerative colitis undergoing restorative proctocolectomy and, alter, it had been validated for CD patients (Delaney et al., 2003; Kiran et al., 2003). In this study, a significant improvement in CGQL early in the postoperative period was observed. Improvement, apparently, occurred irrespective of the disease activity measured with CDAI, the indication for surgery, type of procedure (abdominal or perineal), and history of previous surgery (Delaney et al., 2003). A greater improvement in CGQL in the postoperative period occurred in female patients and those who did not have any complications in the first month of postoperative period. Although the reason for the greater improvement in the latter group is obvious, there was no apparent reason why women benefited more than men, especially as both had similar preoperative CGQL scores (Delaney et al., 2003). As expected, patients who had complications within a month of surgery (irrespective of whether it was major or minor) had a smaller improvement in CGQL than those who had no complication. Patients who were on steroids at the time of surgery had a significantly better improvement than those who were not on steroids. Finally they concluded that HRQL after surgery for CD, as measured by CGQL, seem to improve a month after operation (Delaney et al., 2003). However this result is influenced by the scarce discriminative ability of CGQL which is a generic questionnaire and, therefore, their conclusion might sound a little optimistic (Scarpa et al., 2007b). Finally, the group of Amsterdam compared laparoscopic assisted and open ileocolic resection for primary Crohn’s disease in a randomized controlled trial (Maartense et al., 2006). Sixty patients were randomized for laparoscopic-assisted or open surgery. Primary outcome parameter was postoperative quality of life during 3 months of follow-up, measured by SF-36 and GIQLI questionnaire. And their conclusion was that generic and diseases specific quality of life measured was not different for laparoscopic-assisted compared with the open ileocolic resection (Maartense et al., 2006). Prospective studies investigating short term HRQL after surgery for Crohn’s disease are shown in > Table 135-2.
. Table 135-2 Prospective studies assessing short term HRQL after surgery for CD Country
Follow up
Year Patients
Yazdanpanah et al.
1997
26
France
3–6 months
SF36, RFIPC
improved
Thirlby et al.
1998
36
USA
3 months
HSQ
improved
Tillinger et al.
1999
16
Austria
3–6 months
RFIPC, TTO
improved
BlondelKucharski et al.
2001
231
France
0–3–6–9–12 months
SF36, RFIPC
impaired
Delaney et al.
2003
172
USA
1 months
CGQL
improved
Maartense et al.
2006
62
SF36, GIQLI
improved
Netherlands 1–2–3 months
Questionnaire
HRQL after surgery
Study
2313
2314
135 5
Health-Related Quality of Life after Surgery for Crohn’s Disease
Long-Term Postoperative HRQL
It is generally reckoned that the impairment of HRQL in Crohn’s disease is substantial, it is worse in disease relapses and it is more severe than in ulcerative colitis (Guyatt et al., 1989) but in the literature the analysis of long term HRQL outcome after bowel surgery for Crohn’s disease did not lead to a definite conclusion. On one side there are Thirlby et al., who reported that the quality of life after bowel resection for Crohn’s disease was equal to norms for general population after a median follow up of 16 months (Thirlby et al., 2001). This could possibly be explained by the speculation to those patients with chronic illness experience better HRQL as the duration of their disease increases because their expectations of health decrease (Alison et al., 1997). On the other side Thaler et al. and Casellas et al., who used respectively GIQLI and IBDQ, reported a long term decrease of HRQL in these patients respectively 42 and 34 months after surgery (Casellas et al., 2000; Thaler et al., 2005). This could be explained considering that the model described above probably does not fit Crohn’s disease well because the condition often follows a remitting and relapsing pattern and the response shift model may be more suitable for diseases that remain constant over many years or gradually deteriorate. However, there is also a different and more technical possible explication of this discrepancy. In a recent paper, we demonstrated that the difference in the conclusions of different HRQL studies on the same subjects may be due to the different type of questionnaire used (Scarpa et al., 2007b). In fact, in a further study, we observed a similar discrepancy between the results of generic Cleveland Global Quality of Life which seemed to indicate a HRQL similar to healthy controls and those of disease specific Padova Inflammatory Bowel Disease Quality of Life (PIBDQL) which showed significantly worse HRQL of CD patients compared to healthy subjects (Scarpa et al., 2007a). These results are shown in > Figures 135-2 and > 135-3. The CGQL score consists in three items (current quality of life, current quality of health and current energy level) each on a scale of 0–10 (0, worst; 10, best) (Fazio et al., 1999). The CGQL was created to assess HRQL in patients affected by ulcerative colitis after restorative proctocolectomy and then was successfully used in Crohn’s disease HRQL analysis (Kiran et al., 2003) and its translation in Italian language was validated in one of our previous paper (Scarpa et al., 2007b). The PIBDQL score was developed for patients affected by ulcerative colitis and Crohn’s disease (Martin et al., 1995; Scarpa et al., 2004). This tool, predisposed to be self-administered, consists of 29 multiple choice questions which explore: intestinal symptoms (8 questions), systemic symptoms (7 questions), emotional function (9 questions) and social function (5 questions). The possible answers for each question were graduated on a four point scale and the maximal score was 87 (0, best; 87, worst). The question of whether a generic non specific questionnaire or a disease specific measure for HRQL should be used in patients with Crohn’s disease has not been conclusively solved (Thaler et al., 2005). Generic tools have the advantage that they can be used for comparison between different health conditions and can be easily managed and diseasespecific questionnaires can be used in intra-group comparison, such as seriated evaluation to assess the success of a therapy, and are usually more discriminative (Scarpa et al., 2007b). Probably, the optimal study should include both type of questionnaires to draw any conclusion and the single questionnaire should not be considered separately but should be analyzed in the full context. Furthermore, in our study the postoperative follow up was medially of 46 months (Scarpa et al., 2007a), similarly to those of the studies of Thaler et al. (2005) and Casellas et al. (2000).
Health-Related Quality of Life after Surgery for Crohn’s Disease
135
. Figure 135-2 Quality of life in 96 patients with Crohn’s disease after ileocolonic resection after a 47.1 (40.7–53.5, 95% CI) months follow up compared to that of 69 healthy controls as measured with a generic quality of life questionnaire (CGQL). No significant difference was observed between the two groups with the exception of the current quality of health that resulted lower in Crohn’s disease patients (p < 0.05) (Scarpa et al., 2007a)
. Figure 135-3 Quality of life in 63 patients with Crohn’s disease after ileocolonic resection after a 43.7 (36.6–50.9, 95% CI) months follow up compared to that of 81 healthy controls as measured with a disease specific quality of life questionnaire (PIBDQL). All the scores resulted significantly higher, worse, in Crohn’s disease patients (p < 0.01) (Scarpa et al., 2007a)
2315
2316
135
Health-Related Quality of Life after Surgery for Crohn’s Disease
On the contrary, the patients in the study of Thirlby et al. had a median follow up of 16 months. In our opinion, this short ‘‘long-term’’ follow up could account for another part of the discrepancy above described. Patients in the first months after surgery fully enjoy a restored health compared the poor preoperative HRQL (Casellas et al., 2000; Delaney et al., 2003; Maartense et al., 2006; Scott and Hughes, 1994; Tillinger et al., 1999; Yazdanpanah et al., 1997) and, probably, 16 months after they are still in the ‘‘long wave’’ of the post operative enthusiasm (Scarpa et al., 2007a). After 3 years, or more, from surgery the frank positive effect of the surgical removal of the diseased bowel on HRQL appear to be greatly faded (Casellas et al., 2000; Thaler et al., 2005). After the post-surgical positive peak HRQL of Crohn’s disease patients appears to return to a low plateau where only the current health status really matters. In fact, results form our study showed that the HRQL appear to be significantly related only to current disease activity expressed either by CDAI scores or with number of daily bowel movements (Scarpa et al., 2007a). Also Casellas et al. demonstrated that HRQL is strictly related to clinical disease activity and that the remission is the main predictor of good HRQL irrespective of the means (surgery or medical treatment) used to achieve it (Casellas et al., 2000). Similarly, Andersson et al. concluded that, in patients with colonic Crohn’s disease, health related quality of life seems to be more dependent of present symptoms than type of previous surgery or the need for immunosuppressive medication as aggressive disease as well as previous colonic surgery lacked predictive value (Andersson et al., 2003). McLeod et al. showed that surgical treatment significantly improves the HRQL of patients with severe Ulcerative Colitis but the improvement was independent of the very different surgical procedure employed (conventional ileostomy, Kock pouch or ileal reservoir). HRQL appeared to be a function of therapeutic efficacy rather than of the surgical procedure used (McLeod et al., 1991). This is even truer if we consider the different type of surgery for Crohn’s disease. Our results demonstrated that there is no difference in terms of long term HRQL between patients submitted to open surgery and patients who underwent laparoscopy assisted ileocolic resection. This result confirmed those reported by Maartense et al. in the short term follow up and by Thaler et al. in the long term (Maartense et al., 2006; Thaler et al., 2005). Furthermore not only cosmetic impact of scar sparing surgery (laparoscopy) seemed to play little role in long term HRQL outcome (Dunker et al., 1998) but also bowel sparing surgery (strictureplasty) failed to demonstrate a long term improvement of HRQL (Broering et al., 2001). Disease activity is the most important factor for health related quality of life which underlines the need to use all available measures, medical or surgical, to bring these patients into clinical remission (Andersson et al., 2003). In conclusion the analysis of long term HRQL after surgery for Crohn’s disease with generic questionnaires showed an apparently normal quality of life with a good energy level but with an impaired quality of health. In fact, disease specific questionnaire evidenced a significant impairment of bowel and systemic symptom domains with important consequences on the emotional and social functions. Therefore, HRQL appears to be significantly related only to current disease activity independently form the surgical technique or access. Our studies showed that generic and disease-specific questionnaires give different results so, in our opinion, HRQL should be analyzed with both type of questionnaires at the same time (Scarpa et al., 2007a, b). Generic questionnaire are useful tool to compare the HRQL between different illnesses whereas disease-specific questionnaire are the most useful tools to make intra-group comparison such as seriated evaluations of the success of a therapy. Long term quality of life outcome after surgery for Crohn’s disease in the different studies is resumed in > Table 135-3.
Health-Related Quality of Life after Surgery for Crohn’s Disease
135
. Table 135-3 Long term HRQL after surgery for CD Study
Year
Thirlby et al.
2001
56
USA
16 months SF36
Similar to healthy subjects
Thaler et al.
2005
37
USA
42 months GIQLI
Lower than healthy subjects
SF36
Lower than healthy subjects
Casellas et al.
Patients Country
2000
Scarpa et al. 2007a
Follow up
Questionnaire
HRQL after surgery
29
Spain
34 months IBDQ
Lower than healthy subjects
96
Italy
46 months PIBDQL
Similar to healthy subjects
63
CGQL
Lower than healthy subjects
Summary Points The goal of surgery in Crohn’s disease is not to cure the disease but to improve quality of life.
HRQL can be used to integrate the traditional outcome measure to assess therapeutic efficacy of a surgical procedure in patients with Crohn’s disease.
Need of surgery and need of ostomy bag are among the most important concerns of Crohn’s disease patients.
Generic and disease-specific questionnaires give different results so, HRQL of Crohn’s disease patients should be analyzed with both type of questionnaires at the same time.
Almost all the authors, agree that HRQL of Crohn’s disease patients improves after surgery in the early postoperative period.
The analysis of long term HRQL after surgery for Crohn’s disease with generic questionnaires showed an apparently normal quality of life.
Disease specific questionnaires evidenced that long term HRQL after surgery for Crohn’s disease is significantly related only to current disease activity independently form the surgical technique or access.
References Alison PJ, Locker D, Feine JS. (1997). Soc Sci Med. 45: 221–230. Andersson P, Olaison G, Bendtsen P Myrelid P, Sjodahl R. (2003). Colorectal Dis. 5: 56–62. Bechi P, Tonelli L. (1982). Int Surg. 67: 325–328. Blondel-Kucharski F, Chircop C, Marquis P, Cortot A, Baron F, Gendre JP, Colombel JF. (2001). Groupe d’Etudes The´rapeutique des Affections
Inflammatoires Digestives (GETAID) Am J Gastroenterol. 96: 2915–2920. Broering DC, Eisenberger CF, Koch A, Bloechle C, Knoefel WT, Du¨rig M, Raedler A, Izbicki JR. (2001). Int J Colorectal Dis. 16: 81–87. Canavan C, Abrams KR, Hawthorne B, Drossman D, Mayberry JF. (2006). Aliment Pharmacol Ther. 23: 377–385.
2317
2318
135
Health-Related Quality of Life after Surgery for Crohn’s Disease
Casellas F, Lopez-Vivancos J, Badia X, Vilaseca J, Malagelada JR. (2001). Eur J Gastroenterol Hepatol. 13: 567–572. Casellas F, Lo´pez-Vivancos J, Badia X, Vilaseca J, Malagelada JR. (2000). Am J Gastroenterol. 95: 177–182. Casellas F, Lo´pez-Vivancos J, Vergara M, Malagelada J. (1999). Dig Dis.17: 208–218. Cooper JC, Jones D, Williams NS. (1986). Ann R Coll Surg Engl. 68: 279–282. Delaney CP, Kiran RP, Senagore AJ, O’Brien-Ermlich B, Church J, Hull TL, Remzi FH, Fazio VW. (2003). J Am Coll Surg. 196: 714–721. Delaney CP, Fazio VW. (2001). Surg Clin North Am. 81: 137–158. De Rooy EC, Toner BB, Maunder RG, Greenberg GR, Baron D, Steinhart AH, McLeod RS, Cohen Z. (2001). Am J Gastroenterol. 96: 1816–1821. Drossman DA, Leserman J, Li ZM, Mitchell CM, Zagami EA, Patrick DL. (1991). Psychosom Med. 53: 701–712. Drossman DA, Patrick DL, Mitchell CM, Zagami E, Applebaum MI. (1989). Dig Dis Sci. 34: 1379–1386. Dunker MS, Stiggelbout AM, van Hogezand RA, Ringers J, Griffioen G, Bemelman WA. (1998). Surg Endosc. 12: 1334–1340. Etienney I, Bouhnik Y, Gendre JP, Lemann M, Cosnes J, Matuchansky C, Beaugerie L, Modigliani R, Rambaud JC. (2004). Gastroenterol Clin Biol. 28: 1233–1239. Fazio VW, O’Riordain MG, Lavery IC, Church JM, Lau P, Strong SA, Hull T. (1999). Ann Surg. 1230: 575–586. Guassora AD, Kruuse C, Thomsen OO, Binder V. (2000). Scand J Gastroenterol. 35: 1068–1074. Guyatt G, Mitchell A, Irvine EJ, Singer J, Williams N, Goodacre R, Tompkins C. (1989). Gastroenterology. 96(3): 804–810. Hjortswang H, Almer S, Strom M. (1999). Eur J Gastroenterol Hepatol. 11: 1099–1104. Irvine EJ. (1994). Gastroenterology. 106: 287–296. Kiran RP, Delaney CP, Senagore AJ, O’Brien-Ermlich B, Mascha E, Thornton J, Fazio VW. (2003). Am J Gastroenterol. 98(8): 1783–1789. Kirshner J. (1980). JAMA. 243: 557–563. Levenstein S, Li Z, Almer S, Barbosa A, Marquis P, Moser G, Sperber A, Toner B, Drossman DA. (2001). Am J Gastroenterol. 96: 1822–1830. Lindhagen T, Ekelund G, Leandoer L, Hildell J, Lindstro¨m C, Wenckert A. (1983). Acta Chir Scand. 149: 415–421.
Maartense S, Dunker MS, Slors JF, Cuesta MA, Pierik EG, Gouma DJ, Hommes DW, Sprangers MA, Bemelman WA. (2006). Ann Surg. 243: 143–149. Martin A, Leone L, Fries W, Naccarato R. (1995). Ital J Gastroenterol. 27: 450–454. Maunder RG, Toner B, de Rooy E, Moskovitz D. (1999). Can J Gastroenterol. 13: 728–732. Maunder RG, Cohen Z, McLeod RS, Greenberg GR. (1995). Dis Colon Rectum. 38: 1147–1161. McLeod RS (1999). Adv Surg. 33: 293–309. McLeod RS, Baxter NN. (1998). World J Surg. 28: 375–381. McLeod RS, Churchill DN, Lock AM. (1991). Gastroenterology. 101: 1307–1313. Meyers S. (1983). Mt Sinai J Med. 50: 190–192. Meyers S, Walfish JS, Sachar DB, Greenstein AJ, Hill AG, Janowitz HD. (1980). Gastroenterology. 78: 1–6. Moser G, Tillinger W, Sachs G, Genser D, MaierDobersberger T, Spiess K, Wyatt J, Vogelsang H, Lochs H, Gangl A. (1995). Eur J Gastroenterol Hepatol. 7: 853–858. Mussell M, Bocker U, Nagel N, Singer MV. (2004). Eur J Gastroenterol Hepatol. 16: 1273–1280. Nissan A, Zamir O, Spira RM, Seror D, Alweiss T, Beglaibter N, Eliakim R, Rachmilewitz D, Freund HR. (1997). Am J Surg. 174: 339–341. Scarpa M, Ruffolo C, D’Inca` R, Filosa T, Bertin E, Ferraro S, Polese L, Martin A, Sturniolo GC, Frego M, D’Amico DF, Angriman I. (2007a). Inflamm Bowel Dis. 13(4): 462–469. Scarpa M, Ruffolo C, Polese L, Martin A, D’Inca` R, Sturniolo GC, D’Amico DF, Angriman I. (2007b). Arch Surg. 142(2): 158–165. Scarpa M, Angriman I, Ruffolo C, Ferronato A, Polese L, Barollo M, Martin A, Sturniolo GC, D’Amico DF. (2004). World J Surg. 58(2): 122–126. Scott NA, Hughes LE. (1994). Gut. 35: 656–657. Thaler K, Dinnewitzer A, Oberwalder M, Weiss EG, Nogueras JJ, Wexner SD. (2005). Colorectal Dis. 7: 375–381. Thirlby RC, Sobrino MA, Randall JB. (2001). Arch Surg. 136: 521–527. Thirlby RC, Land JC, Fenster LF, Lonborg R. (1998). Arch Surg. 133: 826–832. Tillinger W, Mittermaier C, Lochs H, Moser G. (1999). Dig Dis Sci. 44: 932–938. Yazdanpanah Y, Klein O, Gambiez L, Baron P, Desreumaux P, Marquis P, Cortot A, Quandalle P, Colombel JF. (1997). Am J Gastroenterol. 92: 1897–1900. Wolff BG. (1998). World J Surg. 22: 364–369.
136 Postoperative Quality of Life Assessment in the Over 80’s After Cardiac Surgery Christoph H. Huber 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2320
2 2.1
Original Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2321 Coronary artery Bypass Grafting in Combination with Aortic Valve Replacement (n = 41) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2322 Outcome of Cardiac Surgery in patients over 80 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2323 Early Outcomes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2323 Mortality (Cumulative Survival in Brackets) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2323 QoF after Cardiac Surgery in the Patients over 80 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2323 The Scoring Classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2323 Nottingham Health Profile . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2325 Short Form-36 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2325 Sickness Impact Profile . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2326 Seattle Angina Questionnaire . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2326 Advantages of the SAQ versus SF-36 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2326 Original Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2326 Methodological Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2328 Gender Difference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2330 Surgical Therapy of Ischemic Heart Disease in the Elderly . . . . . . . . . . . . . . . . . . . . . . 2330 Aortic Valve Surgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2330 Survival and QoL Impacting Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2331 Financial Impact on Healthcare Resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2332 Final Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2332
2.2 2.2.1 2.3 2.4 2.4.1 2.4.2 2.4.3 2.4.4 2.4.5 2.4.6 2.4.7 2.5 2.6 2.7 2.8 2.9 2.10 2.11
Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2332
#
Springer Science+Business Media LLC 2010 (USA)
2320
136
Postoperative Quality of Life Assessment in the Over 80’s After Cardiac Surgery
Abstract: > Heart disease is the leading cause of death in the industrialized nation, and 25% of patients over the age of 80 are functionally limited by their underling cardiovascular condition. Coronary artery diseases have a prevalence of 18–20% in the elderly. There are principally no differences in treatment options and surgical indications in this age group compared to the younger population. Cardiac surgery is a very effective therapy to improve functional status of patients with coronary artery and heart valve disease. In this chapter the impact of heart surgery on Quality of Life (QoL)in this very challenging age segment is summarized. Recent studies confirm the important gain of QoL with a very acceptable mortality in patients over 80 years after heart surgery. These extremely satisfying results should be confronted with the rather reserved referral practice, often withholding old patients for long from cardiac surgery. List of Abbreviations: AVR, > Aortic valve replacement; BMI, Body mass index; CABG, > Coronary artery bypass grafting; CAD, Coronary artery disease; CCS, Canadian Cardiovascular Society grading; COPD, Chronic obstructive airway disease; CPB, Cardiopulmonary bypass; EC, Erythrocyte concentrate; IABP, Intra aortic balloon pump; ICU, Intensive care unit; IDDM, Insuline-dependent diabetes mellitus; LIMA, Left internal mammary artery graft; LVEF, Left ventricular ejection fraction; MI, Myocardial infarction; NHP, > Nottingham Health Profile; NIDDM, Non-insuline-dependent diabetes mellitus; NYHA, New York Heart Association grading; PAD, Peripheral arterial disease; QoL, Quality of life; SAQ, > Seattle Angina Questionnaire; SF36, Short form 36; SIP, > Sickness Impact Profile
1
Introduction
Life expectancy at birth is increasing leading to a growing older population. From 1982, life expectancy at birth has increased in Switzerland, the UK, and the US from 76.3, 75.7, and 75.5 years to 79.9, 77.9, and > 77 years, respectively. By 2022, life expectancy will reach 82.0 years in Switzerland, 80.8 years in the UK, and 80.2 in the US. In the Western European population, 16% is over 65 years old, in the US 13%, and in Switzerland 15% compared to 7% in the world population. It is estimated that the US population will include more than 25 million persons of at least 80 years of age by 2050 (Spencer, 1989). At present, 5% of the Swiss population is aged 80 years and above, and this percentage continues to rise. Life expectancy at the age of 80 in Switzerland reaches an additional 6.6 years for male and another 7.8 years for female habitants. Heart diseases are the number one cause of death in the industrialized nations. More than 25% of the population 80 years and above (Statistical Abstract of the United States/1994 (114th edn.), 1994) are functionally limited by their underling cardiovascular disease, and the prevalence of coronary artery disease is 18–20% in the US and Europe. Many patients in this age group are long withheld from cardiac surgery because of ‘‘too advanced age’’ from a surgical therapy point of view, and remain severely symptomatic despite maximum medical therapy. Nevertheless, as a result of information campaigns based on the surprisingly good results of recent studies suggesting a very acceptable mortality and morbidity, more and more patients over 80 are referred to the cardiac surgeon. This ongoing trend is well reflected in the increased patient population over 80 years undergoing heart surgery at our institution. The numbers have steadily increased from 4.3%
Postoperative Quality of Life Assessment in the Over 80’s After Cardiac Surgery
136
in 2000 to 5.5% in 2002 and to 6.4% in 2004. Interestingly, in the last 3 years the percentage of patients over 80 years has shown accelerated growth with 8.7% in 2006 and 10.8% in 2007. A further decline in mortality and morbidity is mostly due to advances in operative techniques, myocardial protection strategies, and per-operative care. Cardiac surgery can nowadays be performed safely in patients of 80 years and older with good mid-term results. With the advance of elderly surgical candidates and increasing healthcare costs, quality of life (QoL) particularly in this age segment has become a major area of interest. The commonly used objective indicators for risk–benefit analysis such as survival rates, and return to normal activity are of lesser concern in the older patient population as, for example, return to self-care and the more subtle QoL indicators. Outcome assessment requires data collection focusing also on the personal perception of his or her health, physical well-being, and mental state. ‘‘Not just the absence of death but life with the vibrant quality that was associate[d] with the vigour of youth’’ (Elinkton , 1966) described QoL best. Many studies have reported on the QoL after a cardiac surgery (Rothenhausler et al., 2005; Thornton et al., 2005) and an increasing number on the surgical results in patient over 80 (Craver et al., 1999; Kirsch et al., 1998), but only very few recent studies have combined QoL assessment with objective surgical outcomes of octogenarians undergoing heart surgery (Goyal et al., 2005).
2
Original Data
The author has reviewed a single-center experience spanning from 1999 to 2003. A total of 136 consecutive octogenarians underwent either isolated coronary artery bypass grafting (CABG), isolated aortic valve replacement (AVR) or combined CABG/AVR (Huber et al., 2007). Objective data were obtained from the patients’ records. Information on QoL of the 120 surviving patients and causes of deaths were obtained via telephone interviews during a 2-month period by a single investigator. The surviving patient himself was questioned in the first line; relatives, patient’s general practitioners or cardiologists, and hospital autopsy records served to acquire additional information. A modified Seattle Angina Questionnaire was completed in 100% of the survivors, and the preoperative patients characteristics are summarized in > Table 136-1. In the CABG group consisting of 61 patients, unstable angina pectoris was the most common symptom. Forty-two patients (69%) had a CCS of III or more and 27 (44%) presented with a history of myocardial infarction, in 11 patients less than 14 days before and in 16 patients more than 2 weeks before surgery. Only five patients (8.2%) presented with an LVEF of equal to or less than 35%.The LVEF was 59 13%. In 53 patients (87%), threevessel and in seven patients (11%) two-vessel disease was present. Left main stem disease was diagnosed in 25 patients (41%). Multiple revascularization included quintuple CABG in seven cases (11%), quadruple in 25 cases (41%), and triple in 23 cases (38%), whereas two-vessel revascularization was performed six times (10%). In 47 patients (77%), the left anterior descending artery was bypassed with LIMA graft, and eight patients (13%) got a radial artery graft. The operation time was 168 40 min and the bypass time was 77 29 min. Mean aortic cross-clamp time was 46 18 min. Aortic valve replacement was performed 34 times for aortic stenosis. Most patients presented with severe dyspnea; 62% were in NHYA III and LVEF was 60.4 14%. In 16 cases (47%) concomitant aortic insufficiency grade I or II was noted. Stented bioprosthesis were implanted in 25 cases (74%), stentless in 1 case (3%), and a mechanical valve in 8 cases
2321
2322
136
Postoperative Quality of Life Assessment in the Over 80’s After Cardiac Surgery
. Table 136-1 Preoperative patient characteristics
Variables Males Age NYHA or CCS III
All (n = 136)
CABG (n = 61)
AVR (n = 34)
CABG/AVR (n = 41)
No. of patients (%)
No. of patients (%)
No. of patients (%)
No. of patients (%)
80 (59)
39 (64)
15 (44)
26 (63)
82.3 2.1
82.0 1.8
82.6 2.0
82.5 2.6
90 (66)
43 (70)
21 (62)
26 (63)
Hypertension
83 (61)
38 (62)
15 (44)
30 (73)
Dyslipidemia
70 (51)
33 (54)
14 (41)
23 (56)
Pos. family history
34 (25)
14 (23)
6 (18)
14 (34)
NIDDM/IDDM
14 (10)
(10)
(9)
5 (13)
/ 2 (1.5)
/2 (3.3)
0
0
Creatinine120 mmol/l
26 (19)
10 (16)
7 (21)
9 (22)
COPD
21 (15)
7 (11)
7 (21)
7 (17)
Tobacco abuse
33 (24)
23 (38)
4 (12)
6 (15)
PAD
19 (14)
10 (16)
4 (12)
5 (12)
Atrial fibrillation
26 (19)
7 (11)
10 (29)
9 (22)
Previous MI
35 (26)
27 (44)
2 (6)
6 (15)
Anticoagulation
14 (10)
8 (13)
5 (15)
1 (2.4)
BMI Body mass index, CAD Coronary artery disease, CCS Canadian Cardiovascular Society, COPD Chronic obstructive airway disease, MI Myocardial infarction, IDDM Insulin dependent diabetes mellitus, NIDDM Non insulin dependent diabetes mellitus, LVEF Left ventricular ejection fraction, NYHA New York Hear Association, PAD Peripheral arterial disease
(24%). Mean cardiopulmonary bypass time was 76 28 min, and mean aortic cross-clamp time 53 14 min. Overall operation time was 152 37 min.
2.1
Coronary artery Bypass Grafting in Combination with Aortic Valve Replacement (n = 41)
Six (15%) patients had left main disease, 10 (24%) presented with triple vessel CAD, and 35 (87%) had either double- or single-vessel CAD. Twenty (48%) patients had single vessel revascularization, 14 (34%) has double, and 11(27%) had triple or more. Thirteen (31%) LIMAs and 3 (7.3%) radial arteries were harvested. Stented bioprostheses were implanted in 34 cases (83%), stentless bioprostheses in 1 (2.4%), and a mechanical valve in 13 (31%). Mean cardiopulmonary bypass time for combined AVR and CABG was 90 35 min, and mean aortic cross-clamp time was 62 19 min. The overall operative time was 169 41 min.
Postoperative Quality of Life Assessment in the Over 80’s After Cardiac Surgery
2.2
Outcome of Cardiac Surgery in patients over 80
2.2.1
Early Outcomes
136
The most frequent postoperative complication in all groups was arrhythmias with atrial fibrillation in 29 patients (21%). Permanent pacemaker became necessary in two patients only (1.5%). Four patients (3.0%) suffered per-operative myocardial infarction after combined AVR/ CABG, but none required IABP support in this group. Twenty-seven patients (20%) needed inotropic drugs to wean from CPB, and in 21 cases (15%) inotropes had to be continued for more than 24 h. Prolonged mechanical ventilation (>2 days) was necessary in eight cases (5.9%). The average length of stay in the intensive care unit (ICU) was 2.7 1.6 days, and five atients (3.7%) stayed for more than 1 week. Temporary dialysis was required in two cases (1.5%), but in both the renal function recovered completely. Eight patients (5.9%) underwent reintervention: in six cases (4.4%) for persistent bleeding and in two cases (1.5%) because of deep sternal wound infection. Five patients (3.7%) suffered from permanent neurologic impairment and three (2.2%) recovered fully from a transient neurologic impairment. Hospital stay was 14.2 10.1 days, with six patients (4.4%) staying for more than 25 days (range 5–110). In > Table 136-2, the early complications ( Figure 136-1). Causes of the 16 deaths over 5 years were mostly of non-cardiac origin. Fatal pneumonia caused death in five cases (31.3%). In two patients (12.5%) cerebro-vascular accidents, in three (18.8%) septicaemia, in one (6.3%) mesenterial infarction, in one (6.3%) cerebral haemorrhage, and in one female patient (6.3%) euthanasia led to death. In five cases (31.3%) a cardiac cause has been found to be the cause of death. > Table 136-3 shows a comparison of intervention-linked cumulative survival in octogenarians after cardiac surgery.
2.4
QoF after Cardiac Surgery in the Patients over 80
2.4.1
The Scoring Classification
QoF measures are accepted outcome measures well nestled into clinical research, but seldom routinely used in clinical practice (Higginson and Carr, 2001). It might have been Florence Nightingale to first introduce QoF assessment as routine clinical care outcome measure (Rosser, 1985). The term ‘‘quality of life,’’ popularized after a presidential commission to
2323
2324
136
Postoperative Quality of Life Assessment in the Over 80’s After Cardiac Surgery
. Table 136-2 Postoperative complications All (n = 136) Variable Inotropic drug support required IABP
CABG (n = 61)
AVR (n = 34)
No. of No. of No. of patients (%) patients (%) patients (%) 27 (19.9)
9 (14.8)
9 (26.4)
CABG/AVR (n = 41) No. of patients (%) 9 (22.0)
1 (0.7)
1 (1.6)
0
0
2.7 1.6
2.8 1.5
2.6 1.2
2.7 2.1
8 (5.9)
4 (6.6)
3 (8.8)
3 (7.3)
Duration of inotropes (h)
13.5 29.0
8.3 22.3
17.6 33.0
0.75 1.4
No. of intraoperative EC
1.5 2.0
1.4 1.8
1.8 2.2
1.5 2.0
No. of postoperative EC
1.0 5.1
1.4 7.4
0.9 2.0
0.5 1.2
Transient neurologic impairment
3 (2.2)
1 (1.6)
0
2 (4.9)
Permanent neurologic impairment
5 (3.7)
4 (6.6)
1 (2.9)
0
Myocardial infarction
4 (2.9)
0
0
4 (9.8)
Temporary dialysis
3 (2.2)
1 (1.6)
1 (2.9)
1 (2.4)
26 (19.1)
5 (8.2)
11 (32.4)
10 (24.4)
2 (1.5)
0
2 (5.9)
0
ICU (days) Prolonged ventilation (>24 h)
Antibiotic treatment (>48 h) Deep sternal infections needing reoperation Reoperation for bleeding Atrial fibrillation Permanent pacemaker
6 (4.4)
2 (3.3)
2 (5.9)
2 (4.8)
29 (21.3)
8 (13.2)
10 (29.4)
11 (26.8)
2 (1.5)
0
2 (5.9)
0
Hospital stay mean time (days)
14.2 10.1
14.1 13.5
13.8 5.9
14.8 6.2
No. of drugs at discharge
5
5
5
5
Early complications ( Short Form-36 (SF-36) is a multipurpose, short-form health survey with 36 questions and perhaps is the most commonly used healthcare measuring tool. It yields an 8-scale profile of functional health and well-being scores as well as psychometrically based physical and mental health, investigating physical and social functioning, role limitations of physical or emotional origin, vitality, mental health, physical pain, and overall health condition.
2325
2326
136 2.4.4
Postoperative Quality of Life Assessment in the Over 80’s After Cardiac Surgery
Sickness Impact Profile
The SIP is a large 136-item instrument grouped into 12 scales. SIP can be subdivided into two clusters including the physical dimension (three scales) and the psychosocial dimension (four scales) and the remaining five scales.
2.4.5
Seattle Angina Questionnaire
The SAQ regroups 19 items measuring five specific scales: physical limitations, anginal stability, anginal frequency, treatment satisfaction, and disease perception targeting a specific disease and treatment group.
2.4.6
Advantages of the SAQ versus SF-36
In the authors’ study, the use of a modified SAQ instead of the SF-36 questionnaire (Immer et al., 2004, 2005) was motivated by the increased age, the specific disease, and treatment characteristics of the analyzed patient population. The SF-36 is known to be a generic measure, as opposed to one that targets a specific age, disease, or treatment group. The SAQ (Dougherty et al., 1998; Spertus et al., 1995) as opposed to the SF-36 is a shorter 19-item questionnaire. The fewer number and the simple nature of the questions were found to be more suitable to address the very old patient population. All question investigating anginal-related outcomes were supplemented with the symptom of dyspnea in order to address aortic valve disease as well (see > Table 136-4). This modification is by itself not validated, but it does not interfere with the angina assessment and provides a simple tool to measure valve-related QoL perception.
2.4.7
Original Data
Mean follow-up period was 890 days (range from 69 to 1853). Information was collected by telephone interviews. One-hundred and 30 patients left the hospital, and at follow-up 120 patients were alive. A validated SAQ including two additional questions regarding dyspnea was used to assess the QoL (see > Table 136-4). This modified 11-item multiple choice instrument examines mobility and activity, cardiac symptoms perception, disease perception, treatment satisfaction as well as emotional well-being, and enjoyment of life. In order to allow for assessment of valve pathologies, dyspnea as an additional symptom was added to the questions regarding chest pain. QoL improved considerably after cardiac surgery. Overall, 97 patients (81%) were not at all or only slightly disabled in their daily activities (see > Figure 136-2). Physical exercise was not limited or only a little limited in 84 cases (70%). Symptoms decreased post cardiac surgery in 112 patients (93%), only 2 patients (1.7%) felt worse than before operation, and 6 patients (5%) described unchanged symptoms (see > Figure 136-3). Eighty-six patients (72%) were free of angina or dyspnea, while eight (6.6%) remained moderately to severely symptomatic. In the CABG group, 42 patients (77.7%) did not have to take nitroglycerin anymore, and overall 93 patients (77%) were very satisfied and another 21 patients (17.4%) were satisfied
Postoperative Quality of Life Assessment in the Over 80’s After Cardiac Surgery
136
. Table 136-4 Modified Seattle Angina Questionnaire No
Question (Q)/Answer options (A)
1.
Q: How limited are you in your daily activities inside your flat/house? A: severely limited/moderately limited/limited/little limited/not limited
2.
Q: How limited are you moving up stairs or walk up a little hill? A: severely limited/moderately limited/limited/little limited/not limited
3.
Q: Compare your Angina or dyspnea today and before the operation? A: lot more/somewhat more/unchanged/less/much less
4.
Q: Over the past 4 weeks, on average, how many times have you had angina or dysnpea? A: Four times or more a day/ 1–3 times day/three or more times a week/less than once a week/ never in the last 4 weeks.
5.
Q: Over the past 4 weeks, on average, how many times have you taken nitroglycerin? A: Four times or more a day/ 1–3 times day/three or more times a week/less than once a week/ never in the last 4 weeks.
6.
Q: How bothersome is it for you to take your pills as prescribed? A: very/moderately/little/not at all/no drugs prescribed
7.
Q: How satisfied are you that everything possible is being done to treat your heart? A: not satisfied/somewhat satisfied/satisfied/Very satisfied
8.
Q: How satisfied are you with the overall treatment of your heart disease? A: very dissatisfied/somewhat dissatisfied/little dissatisfied/satisfied / very satisfied
9.
Q: Over the last 4 weeks, how much has your angina or dyspnea interfered with your enjoyment of life?
10.
QQ: If you had to spend the rest of your life with your actual discomfort, how would you feel about this?
A: strongly interfered/somewhat interfered/little interfered/not interfered
A: very dissatisfied/somewhat dissatisfied/little dissatisfied/satisfied / very satisfied 11. (a) (b)
Q: (a) How often do you worry that you may have a heart attack or die suddenly? A: can’t stop/often/from time to time/rarely/never Q: Q: (b) Where do you live?
Questionnaire used for quality of life assessment via telephone interview
by the previous treatment. Only six patients (5%) felt not satisfied enough to take their prescribed medication. Furthermore, 112 patients (93.4%) were very reassured to have continuous, full access to medical treatment (see > Figure 136-4a). Interference of cardiac disease with daily enjoyment of life was described by only 9 patients (7.5%), whereas 111 patients (92.5%) had no reduction in their QoL. Sixty-nine patients (58%) were very optimistic to conserve their present activity of life. Thirty-three patients (27.7%) did think about recurrence of their heart disease from time to time, but only 17 patients (14.2%) were anxious more frequently about having a heart attack or die suddenly (> Figure 136-4b). In contrast, 30 patients (45%) undergoing aortic valve replacement with or without concomitant
2327
2328
136
Postoperative Quality of Life Assessment in the Over 80’s After Cardiac Surgery
. Figure 136-2 The percentage and absolute numbers of patients having answered question 1 and 2 with no or little limitations in their daily activities
. Figure 136-3 Percentage and absolute numbers of patients having answered question 3 with much less or less angina or dyspnea and question 4 with angina or dyspnea less than once a week or never in the last 4 weeks
CABG worried at least once a day of dying, versus 6 patients (11.1%) after isolated coronary artery bypass surgery. And finally, 116 patients (97%) at follow-up lived in their own homes and preserved a high degree of self-care.
2.5
Methodological Considerations
Since the mid-1980s, the per- and postoperative survival in octogenarians after cardiac surgery has steadily increased to become highly acceptable nowadays (Alexander et al., 2000). Patients 80 years and older represent a very distinct population from the younger
Postoperative Quality of Life Assessment in the Over 80’s After Cardiac Surgery
136
. Figure 136-4 (a) Percentage and absolute numbers of patients having answered question 7 and 8 with satisfied or very satisfied with their treatment. (b) The percentage and absolute numbers of patients having answered question 9 and 10 with little interference or no interference and satisfied or very satisfied about their emotional well-being, and question 8 with satisfied or very satisfied with their treatment as well as question 11 with rarely or never worrying about a heart attack or sudden death event
cardiac patients (Fruitman et al., 1999). Measurement of morbidity and mortality provide only a small amount of information about the patient’s postoperative physical, functional, emotional, and mental well-being. Little is known about the postoperative symptom perception in patients 80 years and older after a cardiac surgery. The author analyzed the postoperative QoL in 120 consecutive octogenarians post CABG, AVR, or a combination of both procedures. Mean follow-up was about 2½ years, and none of the surviving patients was lost at follow-up. The QoL of 54 patients after CABC surgery, of 31 patients after AVR, and of 35 patients after combined procedures was compared.
2329
2330
136
Postoperative Quality of Life Assessment in the Over 80’s After Cardiac Surgery
The initial assessment was based on the two main cardiac functional symptoms: chest pain and dyspnea. As older patients are known to have advanced symptoms at presentation for surgery, NYHA class III and IV are a more common finding in octogenarians than they are in younger patients (Goyal et al., 2005; Hunt et al., 2000). Alexander et al. (Rothenhausler et al., 2005) described NYHA class III to IV heart failure being present in 16.6% compared to 9.8% in patients younger than 80 years. Fruitman and coworkers (Fruitman et al., 1999) have also shown a significant, higher presence of NYHA IV in octogenarians. Questions 1–5 of the SAQ address either one or both symptoms. In CABG patients, unstable angina pectoris was the presenting symptom, with more than two-thirds of the patients being in a CCS III. Near-equal distribution was found for dyspnea in patients undergoing AVR, with 62% being in NYHA III. In the combined CABG/AVR group, dyspnea was the leading symptom (see > Table 136-1). In contrast, LVEF (59 14%) did not differ from the values in younger collectives possibly because of a selection bias in this older patient segment. Despite a lower than average prevalence of COPD 18, prolonged ventilation (>24 h) for respiratory failure affected 6–9% of the patients (see > Table 136-2). Very similar results have been reported previously (Goyal et al., 2005). This might reflect the diminished physiologic reserves and a more pronounced tendency to fluid retention.
2.6
Gender Difference
Gender difference decreases with age progression. Forty-one percent of patients were women, as opposed to 20–30% described in younger population (Avery et al., 2001; Rothenhausler et al., 2005). The underling study did not identify either a trend for higher female or male inhospital mortality as described by others. Therefore, female gender may be a weaker risk factor in elderly compared to younger women. The higher difference of pre- and postoperative symptoms in this older patient subgroup is a further argument for the benefit of early operative treatment in patients over 80 years.
2.7
Surgical Therapy of Ischemic Heart Disease in the Elderly
Some studies have suggested that severe three-vessel diseases are less prevalent at very advanced age (Goyal et al., 2005), a fact the authors oft the study could not confirm, as 78% of the CABG patients presented with triple-vessel CAD and 42% had left main disease. Independent of age, the primary target of surgical therapy for ischemic heart disease is complete revascularization, reflected in the high number of triple or quintuple CABGs (79%). Arterial grafts such as LIMA are less frequently used as bypass grafts in the elderly (Goyal et al., 2005; Kolh et al., 2001), even though the authors favor the principle of the LIMA as the graft of choice in CABG patients for all ages and therefore should not be withheld from octogenarians as demonstrated by a 77% LIMA and 13% radial artery use for revascularization of the left coronary artery territory.
2.8
Aortic Valve Surgery
The choice of the prosthetic aortic valve is a trade-off between the lower life expectancy for patients 80 years and older and the higher risk of complications from oral anticoagulation
Postoperative Quality of Life Assessment in the Over 80’s After Cardiac Surgery
136
medication. Tissue valves are therefore implanted in 80% of the patients. Twenty percent of patients got a mechanical valve motivated by ongoing oral anticoagulation medication. Most frequent indication for anticoagulation was for atrial fibrillation, present in 25% of patients undergoing aortic valve surgery.
2.9
Survival and QoL Impacting Factors
Survival rates are very acceptable. Three years after surgery 124 patients (90%) were alive, and after 5 years of operation 120 patients (73%) (see > Table 136-3). These survival rates are comparable to or slightly higher than the ones described in other studies (Akins et al., 1997; Asimakopoulos et al., 1997; Craver et al., 1999; Kirsch et al., 1998; Melby et al., 2007; Rosengart et al., 2002; Rothenhausler et al., 2005; Tsai et al., 1994), and show good early and mid-term postoperative results justifying not withholding cardiac surgery from the increasing elderly and very old population. However, longevity is not the primary goal in patients over 80; therefore, good operative outcomes imply not only safety and survival but also the gain of comfort in daily life. The marked improvement of the NYHA functional class as well as improvement of the CCS class that we found (72% free of angina or dyspnea) has also been reported previously (Kumar et al., 1995; Sundt et al., 2000; Thornton et al., 2005). Nevertheless, only marginal attention had been paid in most studies to the improvement of the emotional well-being, treatment satisfaction, and disease perception. The results of the present study in octogenarians demonstrate a remarkable improvement in the QoL and a significant improvement in the patient’s functional status after cardiac surgery. Activity and mobility improved in ischemic and valvular disease, with nearly 80% of the patients feeling no or only very little limitation in their daily activity (see > Figure 136-2). The improvement in exercise tolerance is less homogenously distributed, with 80% of CABG patients reporting virtually free of limitation and only 59% of the patients in the combined procedures group. This difference in exercise tolerance is reflected again in the symptom perception by the patient. Ninety-eight percent of the CABG patients, compared to 85% of CABG and AVR patients, felt considerable improvement of their angina or dyspnea compared to their preoperative clinical condition. The vast majority (93%) of all the octogenarians felt much better after surgery (see > Figure 136-3). In all types of operations, more than 90% of the patients were moderately satisfied or very satisfied with the overall treatment of their heart disease. And, it is note worthy that nearly 100% of the CABG patients as well as 91% of the AVR or CABG and AVR patients felt pleased to have access to full medical treatment, despite their advanced age (see > Figure 136-4a). Over 95% of the patients at follow-up lived in their own homes and enjoyed a high degree of autonomy. Similar results have been found by Fruitman et al. (1999), Kumar et al. (1995), Heijmeriks et al. (1999), Rumsfeld et al. (2001), and Yun et al. (1999). It is noteworthy that 21% of the CABG/AVR patients and 12% of the AVR patients worry at least daily about a heart attack or about sudden death as compared to less than 4% in the isolated CABG group. A possible explanation might be that the more advanced the disease stage, the higher the incidence of so-called near death experiences in patients with aortic stenosis after syncope resulting from a delayed referring practice. This particular emotional dimension might improve if patients with aortic stenosis are operated on earlier independent of their age. The average octogenarians after heart surgery take 5.3 pills, but 75–80% feel only a little or not at all deranged by the daily medication.
2331
2332
136 2.10
Postoperative Quality of Life Assessment in the Over 80’s After Cardiac Surgery
Financial Impact on Healthcare Resources
Happier and healthier octogenarian come at an increased cost. Looking at the economical dimension of patients over 80 undergoing heart surgery does result in 20–27% higher hospital costs as reported (Peterson et al., 1995). Avery et al. attributed the increase of total direct hospital cost in octogenarians to a more severe risk profile and to longer consecutive ICU and hospital stay (6). However, emphasis on early extubation and timely aggressive mobilization after surgery also has successfully decreased the overall intubation time and length of stay of this elderly patient population in ICUs to 2.8 days (see > Table 136-2). This is between the previously reported 6.8 days (4), or 5.1 days (Goyal et al., 2005) and the 1.7–1.1 days of Dalrymple-Hay et al. (1999). The hospital stay of 14.5 days is in the range of previously published values (11). In contrast, with the increased in-hospital cost are the excellent postoperative recovery and gain in QoL, providing back the potential of self-care, which might also be largely compensated by reduced consecutive disease-associated costs compared to medical treatments alone with frequent re-hospitalization for repetitive congestive heart failure and invalidating cardiac symptoms (Sollano et al., 1998).
2.11
Final Considerations
Selected patients of 80 years and older after cardiac surgery show a remarkable QoL and a considerable increase in their emotional well-being (see > Figure 136-4b), as well as an important increase in their functional status with a satisfactory medium-term, 5-year survival (see > Figure 136-1) at a reasonably low risk. The stunning recovery from being a bedridden patient to a self-caring patient is a further very important advantage after cardiac surgery in this challenging age group. Therefore, in selected octogenarians, early operative treatment should not be withheld, and the adoption of an early referral might further increase the patient’s postoperative benefits.
Summary Points Twenty-five percent of patients over the age of 80 are functionally limited by their underling cardiovascular condition.
Heart surgery is a very effective therapy to improve the survival and functional status of patients with coronary artery and heart valve disease in all age groups.
In heart surgery patients of increased age and decreased life expectancy, QoL represents an important end point.
Ninety-three percent of selected patients of 80 years and older show a remarkable QoL and a considerable increase in their emotional well-being after heart surgery.
Earlier and more aggressive surgical therapy substantially increases outcome of heart surgery in octogenarians and should not be withheld until a progressed-disease stage.
Adoption of an earlier referral practice might further increase the postoperative benefits of patients.
Postoperative Quality of Life Assessment in the Over 80’s After Cardiac Surgery
136
References Spencer G. (1989). US bureau of Census: Projections of the Population of the United States by Age, Sex and Race:1988 to 2080. Washington, DC: US Government Printing Office. Current Population Reports, Series P-25, No 1018. Statistical Abstract of the United States/1994 (114th edn.) (1994). Washington DC: Department of Commerce, US Bureau of the Census, 84. Elinkton JR. (1966). Ann Inter Med. 64: 711–714. Rothenhausler HB, Grieser B, Nollert G, Reichart B, Schelling G, Kapfhammer HP. (2005). Gen Hosp Psychiatry. 27(1): 18–28. Thornton EW, Groom C, Fabri BM, Fox MA, Hallas C, Jackson M. (2005). J Thorac Cardiovasc Surg. 130 (4): 1022–1027. Kirsch M, Guesnier L, LeBesnerais P, Hillion ML, Debauchez M, Seguin J, Loisance DY. (1998). Ann Thorac Surg. 66(1): 60–67. Craver JM, Puskas JD, Weintraub WW, Shen Y, Guyton RA, Gott JP, Jones EL. (1999). Ann Thorac Surg. 67(4): 1104–1110. Goyal S, Henry M, Mohajeri M. (2005). ANZ J Surg. 75 (6): 429–435. Huber CH, Goeber V, Berdat P, Carrel T, Eckstein F. (2007). Eur J Cardiothorac Surg. 31(6): 1099–2105. Higginson IJ, Carr AJ. (2001). BMJ. 322(7297): 1297–1300. Rosser RM. (1985). A History of the Development of Health Indices. In: Smith GT (ed.) Measuring the Social Benefits of Medicine. Office of Health Economics, London. William JI. (1991). Theoretic Surg. 6: 152–157. Campbell A, Converse PE, Rodgers WL. (1976). The Quality of American Life. Russel Sage Foundation, New York, pp. 1–583. Dempster M, Donnelly M. (2000). Heart. 83(6): 641–644. Immer FF, Althaus SM, Berdat PA, Saner H, Carrel TP. (2005). Eur J Cardiovasc Prev Rehabil. 12(2): 138–143. Immer FF, Lippeck C, Barmettler H, Berdat PA, Eckstein FS, Kipfer B, Saner H, Schmidli J, Carrel TP. (2004). Circulation. 110(11 Suppl 1): II250–II255. Spertus JA, Winder JA, Dewhurst TA, Deyo RA, Prodzinski J, McDonell M, Fihn SD. (1995). J Am Coll Cardiol. 25(2): 333–341. Dougherty CM, Dewhurst T, Nichol WP, Spertus J. (1998). J Clin Epidemiol. 51(7): 569–575. Alexander KP, Anstrom KJ, Muhlbaier LH, Grosswald RD, Smith PK, Jones RH, Peterson ED. (2000). J Am Coll Cardiol. 35(3): 731–738.
Fruitman DS, MacDougall CE, Ross DB. (1999). Ann Thorac Surg. 68(6): 2129–2135. Hunt JO, Hendrata MV, Myles PS. (2000). Heart Lung. 29(6): 401–411. Avery GJ 2nd, Ley SJ, Hill JD, Hershon JJ, Dick SE. (2001). Ann Thorac Surg. 71(2): 591–596. Kolh P, Kerzmann A, Lahaye L, Gerard P, Limet R. (2001). Eur Heart J. 22(14): 1235–1243. Akins CW, Daggett WM, Vlahakes GJ, Hilgenberg AD, Torchiana DF, Madsen JC, Buckley MJ. (1997). Ann Thorac Surg. 64(3): 606–614; discussion 614–615. Rosengart TK, Finnin EB, Kim DY, Samy SA, Tanhehco Y, Ko W, Lang SJ, Krieger KH, Isom OW. (2002). Am J Med. 112(2): 143–147. Tsai TP, Chaux A, Matloff JM, Kass RM, Gray RJ, DeRobertis MA, Khan SS. (1994). Ann Thorac Surg. 58 (2): 445–50; discussion 450–451. Melby SJ, Zierer A, Kaiser SP, Guthrie TJ, Keune JD, Schuessler RB, Pasque MK, Lawton JS, Moazami N, Moon MR, Damiano RJ Jr. (2007). Ann Thorac Surg. 83(5): 1651–1656; discussion 1656–1657. Asimakopoulos G, Edwards MB, Taylor KM. (1997). Circulation. 96(10): 3403–3408. Kumar P, Zehr KJ, Chang A, Cameron DE, Baumgartner WA. (1995). Chest. 108(4): 919–926. Sundt TM, Bailey MS, Moon MR, Mendeloff EN, Huddleston CB, Pasque MK, Barner HB, Gay WA Jr. (2000). Circulation. 102(19 Suppl 3): III70–III74. Heijmeriks JA, Pourrier S, Dassen P, Prenger K, Wellens HJ. (1999). Am J Cardiol. 83(7): 1129–1132, A9. Rumsfeld JS, Magid DJ, O’Brien M, McCarthy M Jr, MaWhinney S, Scd, Shroyer AL, Moritz TE, Henderson WG, Sethi GK, Grover FL, Hammermeister KE. (2001). Ann Thorac Surg. 72 (6): 2026–2032. Yun KL, Sintek CF, Fletcher AD, Pfeffer TA, Kochamba GS, Mahrer PR, Khonsari S. (1999). Ann Thorac Surg. 68(4): 1314–1320. Peterson ED, Cowper PA, Jollis JG, Bebchuk JD, DeLong ER, Muhlbaier LH, Mark DB, Pryor DB. (1995). Circulation. 92(9 Suppl): II85–II91. Dalrymple-Hay MJ, Alzetani A, Aboel-Nazar S, Haw M, Livesey S, Monro J. (1999). Eur J Cardiothorac Surg. 15(1): 61–66. Sollano JA, Rose EA, Williams DL, Thornton B, Quint E, Apfelbaum M, Wasserman H, Cannavale GA, Smith CR, Reemtsma K, Greene RJ. (1998). Ann Surg. 228 (3): 297–306.
2333
137 Quality of Life and Financial Measures in Surgical and Non-Surgical Treatments in Emphysema J. D. Miller . F. Altaf 1 1.1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2337 Importance of Measuring QOL as an Outcome in COPD Patients . . . . . . . . . . . . . . 2338
2 2.1
QOL Measurements Tools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2339 What Is the Minimal Important Clinical Difference (MICD) of QOL Instrument? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2339
3 3.1
Economic Measures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2342 Calculating Cost per QALY (CUI): An Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2343
4 4.1 4.2 4.2.1 4.2.2 4.2.3 4.2.4 4.2.5
Impact of Medical Intervention on QOL in COPD Patients . . . . . . . . . . . . . . . . . . . 2343 Inhaled Corticosteroids (ICS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2343 Bronchodilators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2344 B2 Agonists . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2344 Anticholinergics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2344 Combination Therapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2344 Pulmonary Rehabilitation (PR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2345 Oxygen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2346
5
Impact of Surgical Intervention on QOL in COPD Patients . . . . . . . . . . . . . . . . . . 2347
6
Financial Impact of Medical and Surgical Intervention in COPD Patients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2349 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2350
#
Springer Science+Business Media LLC 2010 (USA)
2336
137
Quality of Life and Financial Measures in Surgical and Non-Surgical Treatments
Abstract: The cost of health care is continuing to escalate world wide. More people are living longer and often with more challenging medical conditions. The personal and societal cost for survival is growing yearly. Medical and surgical advances, unfortunately, have not been able to stop this growing financial burden. Occasionally new surgical interventions can favorably affect the overall health care cost. If an intervention can create a healthier patient with fewer health care needs, society may experience an overall saving. Such was the early hope for > Lung Volume Reduction Surgery (LVRS). If LVRS can save productive lives and reduce a patients’ ongoing need for medicines, such as oxygen therapy, there may be an overall financial gain to society. If, however, LVRS differs costs into the future as well as add the new surgical costs clearly there will be no overall saving. In this later instance the gain to a patients’ well being must be weighed against the cost of the intervention. Lastly when there is no overall gain for the patient and no saving to society it is clear that we should not invest in the intervention. This chapter reviews the tools used to evaluate a patients health-related quality of life (> HRQOL) and reviews the world literature evaluating health gains and losses following LVRS. The cost of LVRS to society will be reviewed and compared with other medical and surgical interventions with an emphasis on other treatments for advanced > emphysema. We begin with a general outline of the definition of Quality of Life (QOL), and a review of the research tools in use to evaluate QOL. These tools can be designed to assess a patients’ quality of life specifically as it related to a particular disease state such as chronic obstructive lung disease (> COPD) (a disease specific QOL measure), or it can be designed as a broader tool assessing QOL as ones’ general health impacts on their perception of well being. The two major short-comings of a QOL measure is its subjective nature (individual preference based) and its inability to include one of the worst health related outcomes, death. Traditionally a subject who has died is not able to report on their QOL and is omitted from further assessment and is not included in the group assessment. Only patients who are able to complete the questionnaire at the give time are included in that time’s overall group score. Health Utilities, however, is a societal preference based score and ascribes the value zero for death. It therefore can be used as a tool to follow a group of patients over time and include loss of life in the overall scoring of health quality for that group. Healthcare economists can use > health utility (HU) scores of a study group over time as a measure of that groups’ overall health for that time period. It is reported in units called Quality Adjusted Life Years (> QALYs). A comparison between research groups allows the investigator an opportunity to assess the gain or loss of health. This difference is also reported in quality adjusted life years (QALYs). Knowing the gain or loss in QALYs and the cost difference between two groups allows the economist to report the cost per QALY. The cost per QALY is a value of a very general nature and allows for comparison of interventions of various types to one and other. This chapter will outline in more detail each of these measures and tools and discuss their application to the financial assessment of LVRS.
Quality of Life and Financial Measures in Surgical and Non-Surgical Treatments
137
List of Abbreviations: COPD, chronic obstructive pulmonary disease; CRDQ, Chronic Respiratory Disease Questionnaire; CUA, cost utility assessment; CUI, cost utility index; DLCO, Carbon monoxide diffusion capacity; FEV1, forced expiratory volume in one second; HRQOL, health-related quality of life; HUI, > health utility index; ICER, The incremental cost-utility ratio; ISOLDE, inhaled steroid in obstructive lung disease; LTOT, long-term oxygen therapy; LVRS, lung volume reduction surgery; MICD, minimal important clinical difference; MRC, Modified Medical Research Council Dyspnea Index; NETT, National Emphysema Treatment Trial; NHP, Nottingham Health Profile; Pao2, peripheral arterial oxygen content; PFSDQ, Pulmonary Function Status and Dyspnea Questionnaire; PFSS, pulmonary function status scale; PR, > pulmonary rehabilitation; QALYs, quality adjusted life years; QOL, quality of life; QWB, Quality of Well-Being Questionnaire; SF-36, Short Form 36 questionnaire; SGRQ, St. George’s Respiratory Questionnaire; SIP, sickness impact profile; SOLQ, Seattle Obstructive Lung Disease Questionnaire
1
Introduction
Chronic obstructive pulmonary disease (COPD) is chronic progressive and disabling longterm lung diseases that include chronic bronchitis (irreversible narrowing of small bronchi causing airflow limitations) and emphysema (abnormal enlargement of the air spaces distal to the terminal bronchioles accompanied by destruction of their walls causing air trapping and inefficient exchange of gases across to and from the blood). It is a preventable disease that is caused primarily by cigarette smoking. As the disease naturally progress, patients become progressively short of breath and disabled affecting their life in general. Because treatments options has not succeeded in general to reverse the disease progression, interventions that result in improving symptomatology and > quality of life tools (QOL) of theses patients are of prime importance. Several definitions of QOL has been proposed, simply it can be identified as an individual’s satisfaction or happiness with life in domains he or she considers important OR subjective feeling that one’s life overall is going well. Thus, quality of life is, by definition, a subjective concept, dependent on cultural perspectives and values. Several domains and variables are included in QOL assessment, These domains might be psychological and social; physical and mental health; emotional and cognitive dimensions, for example, happiness and satisfaction with life; the ability to function bodily, sexually, socially and occupationally; objective status in terms of finances, working conditions, family conditions. . . . etc. Numerous taxonomies of life domains have been proposed based on studies of general populations of both well and ill people. A typical taxonomy is that of Flanagan (1978), which categorizes 15 dimension of life quality, into five domains, as shown below (> Table 137-1). These factors and domains act and interact in multiple diminutions in determining one’s quality of life (Testa and Simonson, 1996). The concept of HRQOL encompasses the impact of the individual’s health on his or her ability to perform activities of daily living deemed to be important. Generally speaking, assessment of HRQL represents an attempt to determine how variables within the dimension of health (e.g., a disease or its treatment) relate to particular dimensions of life that have been determined to be important.
2337
2338
137
Quality of Life and Financial Measures in Surgical and Non-Surgical Treatments
. Table 137-1 Flanagan’s domains of QOL Physical and material well-being Material well-being and financial security Health and personal safety Relations with other people Relations with spouse Having and rearing children Relations with parents, siblings, or other relatives Relations with friends Social, community, civic activities Helping and encouraging others Participating in local and governmental affairs Personal development, fulfillment Intellectual development Understanding and planning Occupational role career Creativity and personal expression Recreation Socializing with others Passive and observational recreational activities Participating in active recreation As this model includes five domains, other models have included more/different domains and items in an attempt to be reflective of the true person QOL. QOL quality of life
1.1
Importance of Measuring QOL as an Outcome in COPD Patients
More and more studies are including QOL measures as a primary out come in patients with COPD for several reasons
QOL measures are used to quantify the impact of the condition and to compare the effects of lung diseases with the other chronic medical problems.
QOL measures can be used to evaluate changes resulting from therapeutic interventions especially when the results of various therapeutic interventions has failed to show survival benefits in patients with sever COPD. For prognostication. Previous studies showed correlation between measures for QOL and mortality independent of disease severity. In one study (Domingo-Salvany et al., 2002), in addition to the FEV1, both SGRQ and SF-36 scores were an independent predictors of mortality in a cohort of COPD patients. QOL measures are necessary as a central component of cost/effectiveness analysis (Kaplan and Ries, 2005).
Quality of Life and Financial Measures in Surgical and Non-Surgical Treatments
2
137
QOL Measurements Tools
There are several generic and disease-specific instruments that can be used to objectively measure QOL in patients with COPD, each incorporating various aspects of physical, psychological, and social function. 1. Generic QOL instruments: general measures that respond to broad changes in patient’s health state and have the advantage of being a common assessment tool so it can compare QOL across several diseases. 2. Disease-specific QOL instruments: are instruments that are designed to have greater sensitivity to minimal clinical changes in variables that are specific to a particular disease (e.g., COPD) which will thereafter affect the score of that instrument. Each QOL instrument is a questionnaire that has unique domains and items that address specific health related issues; each candidate answers these questionnaires and is subsequently assigned a score for each domain as well as a total score. These scores can be used to assess a study group before and after an intervention, or to compare QOL between study groups. A good QOL instrument must be valid (ability of the instrument to actually measure what it claims to measure), responsive (ability of the instrument to detect clinically significant changes) and reliable (repeated measurements in the same setting yields very similar results), giving these facts, several measures for dyspnea and QOL have been thoroughly verified methodologically in COPD patients.
2.1
What Is the Minimal Important Clinical Difference (MICD) of QOL Instrument?
The Minimally Important Clinical Difference (MICD) is the smallest difference in the score of the QOL instrument that informed patients or physicians perceive as important, either beneficial or harmful. An intervention that led to a change in health reflected by that MCID-score would lead the patient or clinician to consider a change in the management. It is possible for QOL score changes to have statistically significant difference and yet not have meaningful clinical implications. Unfortunately, there is no ‘‘gold standard’’ methodology of estimating the MICD, A possible and widely used, but weak, technique would be to approach a group of experts and ask them whether the particular score looks like a reasonable measure of what is important to patients, as they perceive it (But not as the patient perceive it).Understanding the MICD will help in interpreting the impact of the intervention and in helping the authors in estimating sample size and defining who are the responders (whom they reached MICD). Basic concepts regarding few of these instruments are illustrated in > Table 137-2. The American Thoracic Society maintains a website (www.atsqol.org) that summarizes QOL measures that can be used in outcomes research for lung disease. The site lists measures by disease and offers references on their use. 3. Quality-adjusted life years (QALYs), are the units of quality of life over time (number of years lived). Each year in perfect health is assigned the value of 1.0 QALYs. Death for a full year would have a value of 0. If a patient is alive but not in a perfect health, for example blind or be confined to a wheelchair, then the QALYs would have a value between 0 and 1.
2339
2340
137
Quality of Life and Financial Measures in Surgical and Non-Surgical Treatments
. Table 137-2 Commonly used QOL measures in COPD patients
Instruments
Conduction and estimated time
Domains and/or categories
Item, n
QOL improve as score (increase/ decrease)
MICD
Generic Nottingham Health Profile (NHP) (Hunt et al., 1985)
Self; 5–10 min
Energy, pain, sleep, social isolation, emotional reactions, physical mobility
38
Decrease
Not determined
Medical Outcomes Self or Study Short Form interviewer; 36-Item Health Survey 5 min (SF-36) (Ware and Sherbourne, 1992)
Physical functioning; role limitations due to physical health problems; bodily pain; social functioning; general mental health; role limitations due to emotional problems, vitality, energy or fatigue; general health perceptions
36
Increase
(5–12.5) points change of score
Sickness Impact Profile (SIP) (Bergner et al., 1976; Bergner et al., 1981)
Self or interviewer; 20–30 min
Domains:physical and psychosocial Categories: sleep and rest, eating, work, home management, recreation and pastimes, ambulation, mobility, body care and movement, social interaction, alertness behavior, emotional behavior, communication
136
Decrease
Not determined
Interviewer; 15–25 min
Dyspnea, fatigue, emotional function, mastery
20
Increase
(0.5) points change of score
Disease-specific Chronic Respiratory Disease Questionnaire (CRDQ) (Guyatt et al., 1987)
Quality of Life and Financial Measures in Surgical and Non-Surgical Treatments
. Table 137-2 (continued)
Instruments
Conduction and estimated time
Domains and/or categories
Item, n
137
QOL improve as score (increase/ decrease)
MICD
Pulmonary Functional Self; 15 min Status and Dyspnea Questionnaire (PFSDQ) (Lareau et al., 1994)
Domains functional status, dyspnea Categories: self care, mobility, home management, eating, recreation, social
164
Decrease
Not reported
St. George’s Self or Respiratory interviewer; Questionnaire (SGRQ) 10 min (Jones et al., 1992)
Symptoms, activity, impacts
76
Decrease
4% improvement in separate domains or total score
Seattle Obstructive Self; 5–10 Questionnaire (SOLQ) min (Tu et al., 1997)
Physical function, emotional function, coping skills, treatment and satisfaction
29
Increase
Not determined
Different scaling and scoring system has been developed for each domain instrument, total score can be then evaluated which reflect the QOL status of the candidate. CRDQ Chronic Respiratory Disease Questionnaire; MICD minimal important clinical difference; NHP Nottingham Health Profile; PFSDQ Pulmonary Function Status and Dyspnea Questionnaire; PFSS Pulmonary function status scale; SF-36 Short Form 36 questionnaire; SGRQ St. George’s Respiratory Questionnaire; SIP sickness impact profile; SOLQ Seattle Obstructive Lung Disease Questionnaire
. Table 137-3 Calculating QALYs Years lived (a)
Health state as measured by health utility tool (b)
QALYs gained (¼ a b)
Intervention A
4
0.75
3 QALYs
Intervention B
4
0.5
2 QALYs
Intervention A will generate additional QALYs (1 year of perfect life) in comparison to intervention B. QALYs: quality adjusted life years
The values between 0 and 1 are usually determined by using a QOL measure called a, health utility. Several health utilities are in practice and include: Standard Gamble, Time Tradeoff and the standardized instruments of the Quality of Well-Being Questionnaire (QWB), the Health Utility Index (HUI) and the European Quality of Life tools (Yusen et al., 2002). Calculating QALY is illustrated in > Table 137-3. The Canadian LVRS Study Group used HUI as a measure of QALYs. The HUI is a 15 min self administered generic instrument questionnaire that is used for classification of health status.
2341
2342
137
Quality of Life and Financial Measures in Surgical and Non-Surgical Treatments
Health status is valued according to a multi-active, multi-attribute utility function estimated from preference scores obtained from random sample from the general public. The scoring function assigns a single summary score at an interval between 0.00 (deceased) and 1.0 (perfect health).
3
Economic Measures
Economists have developed different techniques to compare treatments in term of cost and outcome. Several measures have been used to compare the financial impact of different treatments. 1. Cost effectiveness or value-for-money, is a technique used to compares the relative expenditure (costs) and outcomes (effects) of two or more interventions. Although it is the most commonly used, it only measures a single outcome, such as life-years gained or patients cured, as the basis of comparing treatments. 2. Cost utility analysis (CUA), is a technique used to compare the relative expenditures (costs) and outcomes (effects) of two or more interventions where the outcomes are measured in terms of years of full health gained or lost (multiple outcomes) using a measures such as QALYs. The cost utility index (CUI) is expressed as the costs required to generate a year of perfect health (one QALY) using the intervention being studied. A CUA has the generalizability needed to compare various treatment options with one and other. Several registries are now common and are available to compare various interventions including drugs, prostehetics and surgery thus provide health care policy makers with a common unit of cost for health gain for each intervention to determine how to allocate health care funds. In Canada, cost-effectiveness threshold of $20,000/QALY is considered within most healthcare budgets while treatments that cost more than $100,000/QALY are considered to be outside budget limits, these limits differ between countries as their health budgets and health care values differ. An arbitrary figure of $20,000–50,000 dollars per QALY has commonly been applied as the threshold (> Table 137-4). Cost-effectiveness Analysis Registry of various interventions for various diseases can be found at www.tufts-nemc.org/cearegistry/ 3. The incremental cost-utility ratio (ICER) is the ratio between the difference in costs and the difference in QALYs both interventions produced.
. Table 137-4 A sample of cost-effectiveness analysis registry Intervention
Cost/QALY (US dollars)
Mammographic screening/colon cancer screening
10,000–25,000
Implantable cardioverter-defibrillator
30,000–85,000
Dialysis in end stage renal disease
50,000–100,000
LVRS
100,000–300,000
Left ventricular assist device
500,000–1,400,000
The last three interventions would not be considered good value for money (cost effective). LVRS: lung volume reduction surgery
Quality of Life and Financial Measures in Surgical and Non-Surgical Treatments
137
. Table 137-5 Calculating cost per QALY in dollars Intervention B
Difference (or increment)
7,000
4,000
3,000 (a)
0.7
0.5
0.2 (b)
Intervention A Mean costs of treatment per patient (not average) Mean utility (QALYs) per patient (not average) Incremental cost per QALY (ICER) (= a/b)
15,000
Intervention A will coast 15,000 dollars per one full functional year gained by the intervention. QALY quality adjusted life years
3.1
Calculating Cost per QALY (CUI): An Example
Estimates of utility, costs and incremental cost per QALY have been calculated using an updated version of a cost-utility model proposed by Hutton et al. (1996). Cost per QALY estimates should be incremental rather than average since incremental method will present the additional costs and health gains over time for each intervention. Thus it closely reflects the impact of in the real world; especially in COPD patients where long terms follow up is a standard, helping of choosing one intervention over another. A simple example of calculating coast per QALY is illustrated below in > Table 137-5.
4
Impact of Medical Intervention on QOL in COPD Patients
Published guidelines on COPD state that the goals of pharmacologic therapy should be to control symptoms, improve health status, and reduce the frequency of COPD exacerbations (O’Donnell et al., 2003). Bronchodilators are the mainstay of pharmacotherapy for patients with COPD. They are effective in treating symptoms and improving exercise capacity but do not alter disease progression. On the other hand, inhaled corticosteroids (ICS) have beneficial effects in a subset of patients with chronic, stable COPD. They reduce the frequency of exacerbations in those patients with advanced disease (FEV1 50% of predicted) who experience, on average, at least 1 exacerbation per year. Below will discuss the impact of different medical interventions on QOL in patients with COPD. Few studies have looked at the effect of these treatments in comparison to placebo while the others have looked at the effect of treatment combinations in comparison to placebo or single treatment strategy.
4.1
Inhaled Corticosteroids (ICS)
Looking at the effect of ICS on QOL, few studies have looked their effects in comparison to placebo in randomized controlled trials
2343
2344
137
Quality of Life and Financial Measures in Surgical and Non-Surgical Treatments
(ISOLDE) trial (Burge et al., 2000) used the SGRQ as a measure of QOL showed that patients on ICS showed a slower decline in health status compared with those on placebo, although statistically significant these were less than the MICD that is required with the SGRQ. In a recent review (Yang et al., 2007) of randomized trials, rate of change in SGRQ in units/ year was analyzed in five long-term studies (2,507 patients), showed a slowing in the rate of decline of quality of life in ICS group compared to placebo.
4.2
Bronchodilators
4.2.1
B2 Agonists
Few studies compared inhaled B2 agonists (long and short acting) to placebo
QOL improved with formoterol (long acting
> B2 agonist) over 12-week treatment in a randomized controlled trial comparing it with placebo (Dahl et al., 2001), formoterol showed statistically significant improvement in SGRQ of a difference that exceeded the MICD. In a review of studies comparing B2 agonists with placebo in COPD patients (Appleton et al., 2006), Twenty-four studies (6,061 participants) were included, this review demonstrated improvement in SGRQ total score and domains including symptoms, activity and impact over a period of 12 months in treatment arm versus placebo arm.
4.2.2
Anticholinergics
Few studies has looked at the effect of > anticholinergics to placebo
In a randomized-controlled double-blinded studies, tiotropium significantly improved dyspnea scores and the SGRQ total score when compared with placebo (Brusasco et al., 2003; Donohue et al., 2002).
4.2.3
Combination Therapy
Long-acting B2 agonists and ICS have both been recommended in guidelines for the treatment of chronic obstructive pulmonary disease. Their co-administration in a combined inhaler may facilitate adherence to medication regimens, and improve efficacy. Two types of combined inhaler exist currently (B2 agonist and ICS): budesonide/formoterol (Symbicort), and fluticasone/salmeterol (Advair or Seretide). Several studies have looked at the effect of combined therapy against placebo or single therapy in randomized controlled way, recent reviews has looked at these studies looking primarily at exacerbation rate and mortality benefits as well as HRQOL secondarily.
Looking at 11 studies (6,427 participants) were two different combination preparations (fluticasone/salmeterol and budesonide/formoterol) were tested against placebo, both treatments led to significant improvement in QOL measured by CRDQ (Nannini et al., 2007a).
Quality of Life and Financial Measures in Surgical and Non-Surgical Treatments
137
On the other hand, another review looked at the studies that compare above treatments
versus long acting B2 agonists alone, this review included ten studies, a total of 7,598 patients, showed an improvement of QOL measured by the CRDQ and SGRQ (Nannini et al., 2007c). On a third review of randomized controlled studies comparing above combination therapies against ICSs alone, seven studies, a total 5,708 patients, showed an improvement of QOL measured by SGRQ (Nannini et al., 2007b). As above stated, the combined effect of both B2 agonists and ICS on QOL is better than placebo and/or each medication alone. Adding anticholinergics to B2 agonists and ICS (triple combination) has had an added effect; A recent randomized controlled trial (Aaron et al., 2007) looking at combined medical treatments (Tiotropium, salmeterol and inhaled > steroids fluticasone) showed an improvement of QOL measured by SGRQ and CRDQ in patients who received the combination. In summary, ICS and bronchodilators (B2 agonists and anticholinergics) improve QOL in COPD patients but this effect gets more evident as one therapy is added to another but it must be emphasized that this should be interpreted cautiously since these measures of QOL may not reach their minimal clinically significant score. In that context clinical studies should report their results of QOL measures only when they reach their minimal clinically significant score and should attribute their results to the most affected domains.
4.2.4
Pulmonary Rehabilitation (PR)
As defined recently by The American Thoracic Society and the European Respiratory Society: PR is an evidence-based, multidisciplinary, and comprehensive intervention designed for patients with chronic respiratory diseases who are symptomatic and often have decreased daily life activities. Many rehabilitation strategies have been developed for patients with disabling COPD. Programs typically include components such as patient assessment, exercise training, education, nutritional intervention, and psychosocial support. Studies suggest that best results are achieved with programs of 6–10 weeks’ duration that involve 6–8 patients per class. Looking at its effect on QOL
A recent review (Lacasse et al., 2006) involving 31 randomized studies showed statistically and clinically significant improvements in different QOL measures and their domains including dyspnea, fatigue, emotions, and patient control over disease. A recent meta analysis (Cambach et al., 1999) showed significant improvements of QOL measured by CRDQ and all of its domains. Based on theses studies and other studies, American College of Chest physicians guidelines recommendations were as following 1. A program of exercise training of the muscles of ambulation is recommended as a mandatory component of pulmonary rehabilitation for patients with COPD. Grade of recommendation, 1A. 2. Pulmonary rehabilitation improves the symptom of dyspnea in patients with COPD: Grade of recommendation, 1A.
2345
2346
137
Quality of Life and Financial Measures in Surgical and Non-Surgical Treatments
3. Pulmonary rehabilitation improves HRQOL in patients with COPD. Grade of recommendation, 1A. Home-based PR programs, which provide greater flexibility, have recently been compared in several studies with conventional hospital-based programs. Despite few limitations in these studies, improvements in exercise capacity, symptoms, and QOL have been demonstrated with home-based PR (Na et al., 2005). As of now one can conclude that there is strong evidence support the fact of improvement of QOL as a result of pulmonary rehabilitation, more studies are needed to look at the benefit of home based PR programs since it may improve compliance, which may impact the outcome, and decrease the coast.
4.2.5
Oxygen
The basic goal of long term oxygen therapy (LTOT) in patients with chronic stable COPD is to maintain peripheral arterial oxygen content values (Pao2) > / ¼ 60 mm mercury. Long term therapy currently targets patients with sever COPD and clinically significant chronic respiratory failure, which is defined as a Pao2 < / ¼ 55 mm mercury, or as a Pao2 < / ¼ 60 mm mercury and either cor pulmonale (right ventricular failure due to pulmonary hypertension) or hematocrit > / ¼ 55%. Oxygen therapy for patients with chronic, stable COPD is generally administered via nasal prongs at 1–6 L/min, from a cylinder or concentrated oxygen, using one or more of the following approaches: (1) long term oxygen therapy (LTOT) (2) use with exercise (3) use during sleep (4) use during air travel and (5) use as needed for symptoms of dyspnea independent of activity. Of these, LTOT (defined as therapy for >15 h/d) has been shown to have beneficial effects on exercise capacity, mental capacity, cardiopulmonary hemodynamics, and survival in patients with severe COPD with hypoxemia and chronic respiratory failure. Giving that survival benefit, LTOT was targeted in many studies in regard to QOL with variable results. As some studies showed some improvement in QOL, others did not due to financial constrains, decrease in patient mobility and damaging of social relationships as will as a disturbing noise and nasal/ear discomfort associated with the use of nasal tubes.
In one study (Okubadejo et al., 1996) comparing 19 patients that met the criteria for LTOT with 18 less hypoxic patients with better baseline SGRQ scores over a period of 6 months; no change in QOL using SGRQ has been found in patients with severe COPD using an oxygen concentrator. A recent observational study conducted in Brazil (Tanni et al., 2007) looked at the effect of LTOT in hypoxemia patients after transitioned from cylinder to concentrated oxygen, this study resulted in clinically significant improvement in SGRQ total scores mainly manifested on symptoms and impact domains, worth mentioning this result was not observed in non hypoxemic patients (pao2 > 60 mm mercury at rest). Looking at oxygen during exercise training of pulmonary rehabilitation, a review published recently including five randomized trials (Nonoyama et al., 2007), a comparison was done between supplemental oxygen compared to control (compressed air or room air) during the
Quality of Life and Financial Measures in Surgical and Non-Surgical Treatments
137
exercise-training component of a pulmonary rehabilitation in COPD patients whom they did not meet criteria for long-term oxygen therapy (see above) showed no significant differences in QOL. Looking at Ambulatory oxygen, a randomized controlled study conducted on 41 patients, whom they are hypoxic on exercise only, over a 12 weeks period using a cylinder oxygen tanks. Improvements were seen in all domains of the CRDQ for cylinder O2 compared with cylinder air but at study completion, (41%) of responders did not want to continue therapy, with 11 citing poor acceptability or tolerability (Eaton et al., 2002). Finally as of oxygen therapy, LTOT proven survival benefits justify the use of this treatment modality but wither patients should expect an improvement in HRQOL is still not well defined but as technology advances smaller, portable and quieter devices with longer supplement period may give more freedom which may improve the overall QOL though financial restrains will be a limiting factor. Ambulatory and exercise oxygen therapy are not justified based on QOL results alone and one could not conclude that these interventions would necessarily improve QOL.
5
Impact of Surgical Intervention on QOL in COPD Patients
Several surgical procedures has be implemented for the treatment of sever COPD including bullectomies, autonomic denervation, Lung volume reduction surgery (LVRS), endobronchial valve implantation and lung transplantation. Of those the most that has been studies are (LVRS) and lung transplantation which is beyond the scope of discussion in this chapter. Because LVRS has been developed as a surgery to palliate disabling symptoms of emphysema, many studies now have included QOL outcomes along with the commonly measured physiologic and functional outcomes. Many symptom scales and disease-specific and general instruments of QOL have been used for evaluating emphysema patients before and after LVRS. Case-control studies and randomized studies have shown a consistent improvement in symptoms related to emphysema and general QOL tools validating the use of LVRS as a palliative therapy for selected patients with emphysema (> Table 137-6). In a recent review of randomized studies (Tiong et al., 2006), 1,663 patients were included of a total of eight studies, 73% of the study group recruited from NETT study. This study showed Improvement in SGRQ in LVRS group in excess of minimum clinically important difference as will as improvement in all four domains of the CRDQ at 12 and 24 months as well as in SF-36. As there is improvement in QOL measures during the early period after LVRS this usually decline over years and may even get worse than baseline in subset of patients who received LVRS. This has been illustrated in the recent long term update of NETT (Fishman et al., 2003; Naunheim et al., 2006) looking at 1,218 patients with the longest follow up period of 5 years looking at SGRQ as a measure of QOL. This study has showed
Clinically significant improvement in QOL (>8 unit decrease in the SGRQ) occurred in 40%, 32%, 20%, 10%, and 13% of LVRS patients in comparison to the medical group of 9%, 8%, 8%, 4%, and 7% at 1, 2, 3, 4, and 5 years after randomization of all patients (p < 0.001, years 1–3; p < 0.005, year 4). Patients who had Upper-lobe-predominant and low baseline exercise capacity were the most to benefit from this procedure, on the other
2347
2348
137
Quality of Life and Financial Measures in Surgical and Non-Surgical Treatments
. Table 137-6 Major studies reporting health-related quality-of-life assessments after lung volume reduction surgery N
N follow-up (mo) Instruments Outcomes
Author
Study type
Cooper et al. (1995)
Casecontrolled
20
6
MRC
Brenner et al. (1997)
Casecontrolled
145
6
MRC
Improved
Anderson (1999)
Casecontrolled
20
12
QOLS
Improved
Moy et al. (1999)
Casecontrolled
19
6
SF-36
Improved
Hamacher et al. (2002)
Casecontrolled
39
24
MRC
Improved
Appleton et al. (2003)
Casecontrolled
29
51
MRC
Improved
Ciccone et al. (2003)
Casecontrolled
250
52
MRC
Improved
SF-36
Improved
Lofdahl et al. (2000)
Randomized
28
12
SGRQ
Improved
Pompeo et al. (2003)
Randomized
60
6
MRC
Improved
Improved
FF-36
SF-36
Geddes et al. (2000)
Randomized
48
12
SF-36
Improved
Goldstein et al. (2003)
Randomized
55
12
CRDQ
Improved
NETT (Fishman et al., 2003; Naunheim et al., 2006)
Randomized 1218
60
CRDQ
Improved
SGRQ
Improved
QWB
Improved
Hillerdal et al. (2005) Miller et al. (2006)
Randomized Randomized
106 62
12 24
SGRQ
Improved
SF-36
Improved
CRDQ
Improved
SF-36
Improved
HUI
Improved
CRQ Chronic Respiratory Disease Questionnaire; MRC Modified Medical Research Council Dyspnea Index; HUI Health utility index; NETT National Emphysemia Treatment Trial Research Group; SF-36 Medical Outcomes Study Short-Form Health Survey; SGQR St. George’s Respiratory Questionnaire; SF-36 Medical Outcomes Study ShortForm Health Survey; QWB: Quality of Well-Being Questionnaire
hand, this benefit was not obvious in high risk patients (FEV1 < 20% and carbon monoxide diffusion capacity (DLCO) MCID at 2 years, but only one achieved statistical significance. Giving above data, LVRS will offer a transient improvement of QOL but as time progress the benefits start to decrease attributed to the chronic progressive nature of the disease, this decline is noticed earlier and faster (quicker worsening of QOL) in high risk patients (see above) and patients with good baseline exercise capacity and lower lobes predominant disease in comparison to patients with upper lobes predominance and poor baseline exercise capacity which may justify the conduct of this procedure in this set of patients.
6
Financial Impact of Medical and Surgical Intervention in COPD Patients
Patients with COPD have chronic progressive diseases, limited therapeutic options have been proved to only slow the progression of the disease rather than reverse or at most arrest its progression. As this disease progress it affects patient’s life in variety of aspects including clinical deterioration, as this happen patients continue to seek medical attention for long period as the disease progress. Both medical and surgical options for these patients are coasty and need a long term follow up which increase the burden on the health system. In absence of survival benefits of the treatment options and due to the chronic nature of the disease, solutions that would alter patient’s quality of life and decrease the coast will be at most important. Few studies have looked at the cost effectiveness of surgical treatment versus medical treatment in patients with COPD.
Our group has looked at the incremental cost-effectiveness of LVRS compared with best medical therapy alone over a 2-year time horizon (Miller et al., 2006), this was defined as the ratio of the difference between treatment groups in mean costs to the difference in quality-adjusted life years (cost per QALY, ICER, see above) using HUI as a measure.
2349
2350
137
Quality of Life and Financial Measures in Surgical and Non-Surgical Treatments
Dividing the incremental mean cost of LVRS patients ($28,119) by their incremental QALYS (0.21) results in an incremental cost-effectiveness ratio of $133,900 per QALY gained for patients treated with LVRS. (N.B: Lung transplantation costs $137,000–294,000 per QALY gained). A recent follow up of the coast effectiveness of LVRS over medical treatment of NETT study using QWB as a utility measure (Ramsey et al., 2007) confirmed a cost difference of $140,000 per QALY gained in patients treated with LVRS at 5 years. In subgroup analysis, the cost-effectiveness of LVRS in patients with upper-lobe emphysema and low exercise capacity was $77,000 per QALY gained at 5 years. This difference was attributed mainly to the higher QALY rather than a lower coast in comparison to the high risk patients. As a result LVRS is relatively poor cost-effectiveness overall though in a sub group of patients with upper-lobe, low exercise capacity, it appears that rates of cost-effectiveness might be considered a good value for the level of expenditure required. Giving that, Cost/QALYs should not be used in isolation and many other factors should be considered in treatment decision making in COPD patients.
Summary Points Medical and surgical treatments of COPD patients have failed to reverse or arrest the
progression of the disease and improve survival. This made any intervention that will improve the QOL of these patients of at most important. QOL is an individual’s satisfaction or happiness with his/her life that can be objectively measured by several QOL tool (questionnaires) that include different domains of life, these tools can be general or specific to certain disease as their domains change. The effect of any intervention in BOTH improving QOL and number lived years (QALYs) can be measured by health utility tools. The decision to choose any medical intervention by health agencies depend on BOTH its effect (QALYs) and cost of each intervention (cost/QALYs). The best intervention is the one that cost less for better QALYs gained. Registry data can be used to compare the effect (cost/QALYs) of different interventions. These data can help in allocation of financial resources to the neediest interventions in that community. Bronchodilators and ICS inhalation therapy improve QOL especially if they combined together. Community pulmonary rehabilitation improve QOL. LTOT may improve QOL and that will be more observed as technology advances to provided smaller containers at lower cost. LVRS is a palliative procedure that has not clearly showed improvement in survival in COPD patients but it did improve QOL in comparison to medically treated patients at higher coast. Although LVRS improve QOL, this is unfortunately a temporary effect and QOL continue to decline over time up to baseline or even worse giving the progressive nature of this disease. LVRS is poor cost effective intervention, especially in certain set of patients though the decision to proceed with this intervention should not be based solely on this fact since this procedure may prove cost effective as longer follow up results come out.
Quality of Life and Financial Measures in Surgical and Non-Surgical Treatments
137
References Aaron S, Vandemheen M, Fergusson D, Maltais F, Bourbeau J, Goldstein R, Balter M, O’Donnell D, McIvor A, Sharma S, Bishop G, Anthony J, Cowie R, Field S, Hirsch A, Hernandez P, Rivington R, Road J, Hoffstein V, Hodder R, Marciniuk D, McCormack D, Fox G, Cox G, Prins H, Ford G, Bleskie D, Doucette S, Mayers I, Chapman K, Zamel N, FitzGerald M. (2007). Ann Intern Med. 146: 545–555. Anderson KL. (1999). Am J Crit Care. 8: 389–396. Appleton S, Adams R, Porter S, Peacock M, Ruffin R. (2003). Chest. 123: 1838–1846. Appleton S, Poole P, Smith B, Veale A, Lasserson TJ, Chan MM, Cates CJ. (2006). Cochrane Database Syst Rev. Issue 3. Art. No: CD001104. Bergner M, Bobbitt RA, Carter WB. (1981). Med Care. 19: 787–805. Bergner M, Bobbitt RA, Pollard WE. (1976). Med Care. 14: 57–67. Brenner M, McKenna RJ, Gelb AF. (1997). Chest. 112: 916–923. Brusasco V, Hodder R, Miravitlles M. (2003). Thorax. 58: 399–404. Burge PS, Calverley PM, Jones PW. (2000). BMJ. 320: 1297–1303. Cambach W, Wagenaar RC, Koelman TW. (1999). Arch Phys Med Rehabil. 80: 103–111. Ciccone AM, Meyers BF, Guthrie TJ. (2003). J Thorac Cardiovasc Surg 125: 513–525. Cooper JD, Trulock EP, Triantafillou AN. (1995). J Thorac Cardiovasc Surg. 109: 106–116. Dahl R, Grerst LA, Nowak D. (2001). Am J Respir Crit Care Med. 164: 778–784. Domingo-Salvany A, Lamarca R, Ferrer M. (2002). Am J Respir Crit Care. Med 166: 680–685. Donohue JF, van Noord JA, Bateman ED. (2002). Chest. 122: 47–55. Eaton T, Garrett JE, Young P, Fergusson W, Kolbe J, Rudkin S, Whyte K. (2002). Eur Respir J. 20: 306–312. Fishman A, Martinez F, Naunheim K, Piantadosi S, Wise R, Ries A, Weinmann G, Wood DE. (2003). N Engl J Med. 348: 2059–2073. Flanagan JC. (1978). Am Psychol. 33: 138–147. Geddes D, Davies M, Koyama H. (2000). N Engl J Med. 343: 239–245. Goldstein RS, Todd TRJ, Guyett G. (2003). Thorax. 58: 405–410. Guyatt GH, Berman LB, Townsend M. (1987). Thorax. 42: 773–778. Hamacher J, Buchi S, Georgescu CL. (2002). Eur Respir J. 19: 54–60.
Hillerdal G, Lofdahl CG, Strom K, Skoogh BE, Jorfeldt L, Nilsson F. (2005). Chest. 128: 3489–3499. Hutton J, Brown R, Borowitz M. (1996). Pharmacoeconomics. 9(Suppl 2): 8–22. Hunt SM, McEwen J, McKenna SP. (1985). J R Coll Gen Pract. 35: 185–188. Jones PW, Quirk FH, Baveystock CM. (1992). Am Rev Respir Dis. 145: 1321–1327. Kaplan RM, Ries AL. (2005). J Cardiopulm Rehabil. 25: 321–331. Lacasse Y, Goldstein R, Lasserson TJ, Martin S. (2006) Cochrane Database Syst Rev. Issue 4. Art. No.: CD003793. Lareau SC, Carrieri-Kohlman V, Janson-Bjerklie S. (1994). Heart Lung. 23: 242–250. Lofdahl CG, Hillerdal G, Strom K. (2000). Am J Respir Crit Care Med. 161: A585. Miller JD, Malthaner RA, Goldsmith CH, Goeree R, Higgins D, Cox PG, Tan L, Road JD. (2006). Ann Thorac Surg 2006 81: 314–321. Moy ML, Ingenito EP, Mentzer SJ, Evans RB, Reilly JJ Jr. (1999). Chest 115: 383–389. Na JO, Kim DS, Yoon SH. (2005). Monaldi Arch Chest Dis. 63: 30–36 Nannini L, Cates CJ, Lasserson TJ, Poole P. (2007a). Cochrane Database Syst Rev. Issue 4. Art. No.: CD003794. Nannini LJ, Cates CJ, Lasserson TJ, Poole P. (2007b). Cochrane Database Syst Rev. Issue 4. Art. No.: CD006826. Nannini LJ, Cates CJ, Lasserson TJ, Poole P. (2007c). Cochrane Database Syst Rev. Issue 4. Art. No.: CD006829. Naunheim K, Wood D, Fishman A. (2006). Ann Thorac Surg. 82: 431–443. Nonoyama ML, Brooks D, Lacasse Y, Guyatt GH, Goldstein RS. (2007). Cochrane Database Syst Rev. Issue 4. Art. No.: CD005372. O’Donnell DE, Aaron S, Bourbeau J, Hernandez P, Marciniuk D, Balter M, Ford G, Gervais A, Goldstein R, Hodder R, Maltais F, Road J. (2003). Can Respir J. 10(Suppl): 11A–65A. Okubadejo AA, Paul EA, Jones PW, Wedzicha JA. (1996). Eur Respir J. 9: 2335–2339. Pompeo E, Marino M, Notroni I, Matteucci G, Mineo TC. (2003). Ann Thorac Surg. 70: 948–954. Ramsey S, Shroyer L, Sullivan S, Wood D. (2007). Chest. 131: 823–832. Tanni SE, Vale SA, Lopes PS, Guiotoko MM, Godoy IL, Godoy IR. (2007). J Bras Pneumol. 33: 161–167. Testa MA, Simonson DC. (1996). N Engl J Med. 334: 835–840.
2351
2352
137
Quality of Life and Financial Measures in Surgical and Non-Surgical Treatments
Tiong LU, Davies HRHR, Gibson PG, Hensley MJ, Hepworth R, Lasserson TJ, Smith B. (2006). Cochrane Database Syst Rev. Issue 4. Art. No.: CD001001. Tu S, McDonell M, Spertus J. (1997). Chest. 112: 614–622. Ware JE Jr, Sherbourne CD. (1992). Med Care. 30: 473–483.
Yang IA, Fong KM, Sim EHA, Black PN, Lasserson TJ. (2007). Cochrane Database Syst Rev. Issue 4. Art. No.: CD002991. Yusen RD, Morrow LE, Brown KL. (2002). Semin Thorac Cardiovasc Surg. 14: 403–412.
138 Quality of Life After Revascularization and Major Amputation for Lower Extremity Arterial Disease M. Deneuville 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2354 1.1 HRQOL Tools in LEAD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2355 2
Generic Questionnaires . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2355
3 Disease-Specific Instruments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2358 3.1 The Vascular Quality of life Questionnaire (VascuQol) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2359 3.2 Other and Alternative Measures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2360 4
Impact of Revascularization on HRQOL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2360
5
Impact of Revascularization for Intermittent Claudication on HRQOL . . . . . . . . 2362
6
Impact of Revascularization for Critical Limb Ischemia . . . . . . . . . . . . . . . . . . . . . . . . . 2365
7
Impact of Limb Amputation on HRQOL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2370
8
Current limitations/Future Outlook of QOL Assessment of Patients with LEAD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2373
9
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2375 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2375 Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2376
#
Springer Science+Business Media LLC 2010 (USA)
2354
138
Quality of Life After Revascularization and Major Amputation for LEAD
Abstract: Lower extremity arterial disease (> LEAD) is not a curable disease and > revascularization procedures have little or no effect on the overall life expectancy. Hence, treatment should be aimed primarily at alleviating symptoms, controlling risk factors and improving health-related quality of life (HRQOL). LEAD is associated with impaired HRQOL not only in physical domains but also in social function, emotional and mental health. LEAD is commonly associated with many risk factors each being capable to deteriorate HRQOL independently. In contrast to the well-developed body of publications on surgical outcomes, prospective data on patient-oriented outcomes after revascularization are still lacking with a total volume of publications currently below 40. The available data provide some evidence that successful revascularization immediately improves the HRQOL in patients suffering from ischemic claudication with a lasting benefit on physical functioning for at least 12 months while a trend toward return to baseline values in mental health, emotional and vitality domains is commonly observed. Surprisingly, patients with unsuccessful revascularization with minimal increase in lower limb blood flow still experience some improvement in pain, emotional reactions and family relationships in the first year. In the most severe form of LEAD (critical limb ischemia), an immediate and lasting benefit on HRQOL is seen after successful revascularization although less pronounced than in claudicants. However, despite long-term limb salvage and optimal graft functioning, patients successfully revascularized remain functionally disabled when compared to age-matched subjects, nevertheless they report similar well-being. After major limb amputation, some improvement in HRQOL can be expected through pain relief and the maintenance of mobility either with prosthetic rehabilitation or wheel chair ambulation. The measurement of HRQOL is clearly needed at baseline and after vascular operations but its future role in the decision making process is yet to be defined. List of Abbreviations: ABI, > ankle brachial index; > CLI, critical limb ischemia; ET, > endovascular therapy; HRQOL, health-related quality of life; > IC, intermittent claudication; LEAD, lower extremity arterial disease; OVS, open vascular surgery; PTA, percutaneous transluminal angioplasty; TASC, Trans-Atlantic Inter-Society Consensus; VBG, venous bypass graft
1
Introduction
Lower extremity arterial disease (LEAD) is a common illness in adults over 50 years. It is currently established that LEAD is likely to indicate an extensive and severe degree of systemic atherosclerosis with a markedly increased risk of coronary and cerebro-vascular acute events. In population studies, the risk of limb amputation is lower than 1–2% in contrast with a high cardiovascular mortality, ranging from 5 to 10%/year (ACC/AHA, 2006; Novgren et al., 2007). The available data clearly indicates that the general health status and quality of life of patients affected by LEAD are both significantly impaired compared to subjects without LEAD in age-matched general populations in the USA (Gibbons et al., 1995; Holtzman et al., 1999), Germany (Engelhardt et al., 2006), Spain (Hernandez-Osma et al., 2002), UK (Pell, 1995; Basil, 2006) and northern Europe (Klesvgard et al., 2001; Hallin et al., 2002). Self-perceived limitations in their physical activities truly exist in patients with LEAD when compared with other vascular conditions (aortic aneuvrysm, carotid stenosis) which are generally asymptomatic (Hallin et al., 2002). In addition, LEAD is also associated with
Quality of Life After Revascularization and Major Amputation for LEAD
138
substantial impairment in other HRQOL domains including social function, emotional and mental health (Pell, 1995, Khaira et al., 1996; Breek et al., 2002). Many risk factors commonly associated with LEAD are each being capable to deteriorate HRQOL independently (Cherr et al., 2007; Aquarius et al., 2007; Rajagopalan et al., 2006). Revascularization is mainly indicated in patients complaining of lifestyle limitations in their walking abilities (i.e., intermittent claudication, IC) or in limb-threatening conditions such as critical limb ischemia (CLI) or acute ischemia. However, most studies after revascularization reported little or no effect on survival when compared to the overall life expectancy in LEAD. Therefore, the goals of revascularization should be mainly geared toward relief of symptoms and improvement of HRQOL increasing life expectancy should be reached only with an aggressive medical treatment and life style modifications. The outcome of vascular surgery is traditionally assessed by means of technical endpoints that include patency, limb salvage and survival rates, all relevant for the surgeons who perform the procedures (Novgren et al., 2007; ACC/AHA, 2006). However, the patient is more interested in answers to basic concerns like improved mobility, maintained or regained ability to engage in family or social activities, relief of pain and preservation of the body wholeness among many aspects of HRQOL. Surprisingly, patient-oriented measurements were not considered in the assessment of vascular surgery until the early 90’s, more than forty years after the first bypass grafting and 25 years after the development of balloon angioplasty. Clearly, HRQOL measurements have been neglected for too many years and must now be regarded as an important factor in every-day clinical practice (Novgren et al., 2007; ACC/ AHA, 2006). However, despite of a significant increase in the volume of publications in the last 10–15 years, there is still a limitation of data. The emerging role of HRQOL measurements in both the decision making for the surgical interventions in LEAD as well as the post interventional outcome is yet to be defined. This chapter will review and discuss the available data on HRQOL in patients with LEAD eligible for surgical treatment.
1.1
HRQOL Tools in LEAD
Before 1999, few studies were published using many different instruments to measure HRQOL. Interpretation of the available data was limited by the variety of instruments used, the various bias that included small number of patients studied, high rate of nonresponders, survey design and/or inclusion criteria which more likely excluded the most severe patients. Earlier studies used internationally recognized generic questionnaires or instruments adapted from them. Over the last 8 years, such disease-specific tools aimed at HRQOL assessment in LEAD were designed, validated and used in several large prospective studies for evaluating the different degree of severity of LEAD (Chong et al., 2002; Morgan et al., 2001).
2 > The
Generic Questionnaires
Short Form-36 (SF-36) has been used most widely, and since a large component of this questionnaire measures physical function, this instrument has proved to be more suitable for patients with IC (Wann-Hansson et al., 2004) and has been recommended by the TASC as the
2355
2356
138
Quality of Life After Revascularization and Major Amputation for LEAD
preferred generic health outcome measure in lower limb ischemia. However, most studies using the SF-36 have shown that it tends to behave as a functional scale responding best within its physical domain (Currie et al., 1995; Bosch et al., 1999; Deutschmann et al., 2007). The RAND-36 questionnaire is very similar (Bosh et al., 1999). > Figure 138-1–138-3 compares the average scores reported in the largest studies on IC and CLI by using generic (SF-36 and NHP) and disease-specific (VascuQOL) questionnaires. > The Nottingham Health Profile (NHP) has been the second most frequently used generic tool in LEAD. It has been developed first as a measure of perceived stress relating to potentially disabling conditions. The NHP is a two-part instrument. The first part comprises 38 yes-no items investigating the patient’s degree of distress within the domains of physical mobility (8 items), pain (8 items), sleep (5 items), energy (3 items), social isolation (5 items), and emotional reactions (9 items). The answers give a range of possible scores in each domain from zero (no problem at all in the domain) to 100 (all problems present). The second part consists of seven yes-no statements about the frequency of daily living problems with paid employment, housework, family relationships, social life, sex life, hobbies and holidays. The result is presented as a percentage of affirmative responses. The second part is seldom used in LEAD (Klevsga˚rd et al., 2001) and less appropriated to CLI. When compared to SF-36 (Klevsga˚rd et al., 2002; Wann-Hansson et al., 2004), the NHP scores were more skewed and less homogenously distributed. However, it seemed more sensitive in detecting within-patients changes after revascularization for both IC and CLI and to better discriminate among levels of ischemia. Those patients have more problems with mobility and pain which are better screened in the NHP domains of bodily pain and social functioning that contain respectively 8 and 5 items while SF-36 contains only 2.
. Figure 138-1 Short Form-36: comparison of the mean scores (range:0–100) in claudicants (seven pooled studies) versus CLI (five pooled). PF physical functioning; RP role physical; BP bodily pain; GH general health; RE role emotional; VI vitality; MH mental health; SF social functioning
Quality of Life After Revascularization and Major Amputation for LEAD
138
. Figure 138-2 Nottingham Health Profile: comparison of the mean scores (range:0–100) in claudicants (four pooled studies) versus CLI (five pooled). Phys mob: physical mobility; Emo reac: emotional reaction; Soc isol: social isolation
. Figure 138-3 Vascular Quality of life Questionnaire (VascuQOL): comparison of the mean scores (range:0–7) in claudicants (two pooled studies) versus CLI (two pooled)
> The EuroQOL is administered in two parts. The first part consists of a simple questionnaire of 5 questions regarding mobility, self-care, usual activities, pain/discomfort and anxiety/ depression, with each with three possible levels of severity corresponding to ‘‘absence of problem’’ (level 1), some problems’’ (level 2) and ‘‘extreme problems’’ (level 3). The responses
2357
2358
138
Quality of Life After Revascularization and Major Amputation for LEAD
classify the subject into one of 243 different profiles and produce a single numeric index of health status. The EuroQOL also incorporates a visual analogue scale (VAS) on which patients are requested to rate their self-perceived health status on a scale from the worst imaginable (0) to the best (100). The VAS index generated is friendly and easy to read. EuroQOL was used in two large multicenter randomized studies, the Dutch BOA (Bypass, Oral anticoagulants, or Aspirin) study in the Netherlands (Tangelder et al., 1999) and the Bypass versus angioplasty in severe ischemia of the leg trial performed in UK (Basil, 2005). > The WHOQOL-100 is an instrument from the World Health Organization. It consists of 100 questions with 24 facets in six domains. Dutch investigators from the University of Leiden used an abbreviated version with only three domains: physical health, level of independence and social relationships (Breek et al., 2002; Aquarius et al., 2007). The > Ql-index Spitzer was developed in Canada and USA to measure HRQOL in patients with cancer. The index includes five dimensions: involvement in own occupation, activities of the daily living, perception of own health support of friends and family and outlook of life. Each domain is scored 0, 1 or 2 with a best possible score of 10. This generic index has been used only in three of the earlier studies of HRQOL in patients with CLI at Sao Paulo, Brazil (Albers et al., 1992). > The Functional status Questionnaire comprises five domains: general health (1 item), activities of the daily living (6 items), social activities (3 items) and mental well being (5 items). This tool was used only in one study (Gibbons et al., 1995) designed to evaluate the functional outcome and the return to well-being after infrainguinal revascularizations.
3
Disease-Specific Instruments
These tools are supposed to be more accurate and responsive to detect changes in the level of the disease severity within patients before and after therapy. Most of the various attempts made at developing a specific questionnaire for LEAD were in fact limited to HRQOL measurements in patients suffering from IC. The > Walking Impairment Questionnaire (WIQ) was among the first developed. In a small series of fourteen patients after bypass surgery for IC, the functional improvement measured with the WIQ was not predicted from the routine noninvasive testing (Regensteiner et al., 1993). This 15-items questionnaire investigates four domains of walking ability: pain, distance, speed and stair climbing. The distance summary score reflects a patient’s rating of four degrees of difficulty in walking up to seven distances, ranging from indoor to five blocks. An important limitation is that the WIQ does rely on the patient’s estimate of walking distance previously proved to be inaccurate. Similar inaccuracy is likely to affect speed appreciation. In addition, patients may be variously concerned by stair climbing disability. Therefore, the WIQ is a condition-specific scale which is more able to detect impairment and changes in patients with a relatively preserved level of function, rather than being a HRQOL measure. Nevertheless, the WIQ is well validated to use in patient with IC and has been used in several studies as an adjunct to generic HRQOL measures (Feinglass et al., 2000). The > CLAUS-S was initially developed in Germany to investigate pharmacological intervention in claudicants. The original version listed 80 items which required 18 min to score were found unsuitable for clinical practice. A revised version was restricted to
Quality of Life After Revascularization and Major Amputation for LEAD
138
47 items grouped into five domains, including seven specific subgroups, which each generate a score: daily life, pain, social life, disease-specific anxiety and mood. Until 2002, this questionnaire translated in French, Dutch and Flemish had no published validation for its English version. Recently, the questionnaire was assessed for validity and responsiveness together with two generic (SF-36, EuroQOL) and two other disease-specific instruments in a small prospective study at the Hull Royal Infirmary (Metha et al., 2006). In that study, CLAUS-S failed to fully capture mild (pain domain only) and moderate (pain and everyday life) improvements in QOL of patients treated for IC. In term of responsiveness, the CLAUS-S appeared to be equivalent to the SF-36. The authors concluded that the extra time, effort and resources involved in administering the CLAUS could be questioned. However, this tool was used recently in the Oslo Balloon Angioplasty versus Conservative Treatment Study (Nylaende et al., 2007). The > Intermittent Claudication Questionnaire (ICQ) was recently developed at the Imperial College of Medicine and University of Oxford to measure HRQOL of patients with claudication. (Chong et al., 2002). It consists of 16 questions exploring limitations due to leg pain during daytime activities (8 items), at work, hobbies, social life and errands (4 items), emotional impact (3 items) and pain severity (1 item). Each question has a 5 point adjectival scale except the latter which has 6. The instrument produces a single score on a scale from 0 to 100, where 0 is the best possible and 100 is the worst possible health rate. This questionnaire was initially piloted in 20 patients, and then administered to 124 stable claudicants. It was found to be friendly with an average completion time lower than 4 min. Responsiveness of the ICQ was assessed in 60 patients treated conservatively and 40 patients treated with PTA. To our best knowledge, ICQ was not used since the initial report from the promoters (Chong et al., 2002). The > Sickness Impact profile, IC (SIPic) was derived from the generic questionnaire SIP that assesses sickness related behavior. It is a simple instrument that consists of 12 items. The SIPic score is simply the sum of the total number of the dysfunctional items endorsed, ranging from 0 to 12 (Metha et al., 2006).
3.1
The Vascular Quality of life Questionnaire (VascuQol)
This disease-specific questionnaire was designed at the King College in London (Morgan et al., 2001) to cover the whole spectrum (IC to CLI) of patients with LEAD. It contains 25 items subdivided into 5 dimensions: pain (4 items), symptoms (4 items), activities (8 items), social (2 items) and emotional (7 items). Each question has a 7-point response option ranging from 1 (worst score possible) to 7 (best score). Each domain is scored 1–7 and a final total score 1–7 is obtained by summing all the items scores divided by 25. A relevant point regarding self-reported walking distance in this questionnaire was to incorporate a qualitative response to grade the changes in their walking ability instead of absolute measurement of the walking distance which proved to be inaccurate. The validity and responsiveness of VascuQol were tested twice comparatively with different generic (SF-36, EuroQOL) or disease-specific (CLAUS-S, SIPic) QOL instruments in patients with claudication (De Vries et al., 2005; Mehta et al., 2006). In both studies, VascuQol was proved to be the most responsive and the only tool that could detect both mild and moderate clinical improvement.
2359
2360
138
Quality of Life After Revascularization and Major Amputation for LEAD
The authors recommended the VascuQol as the preferred questionnaire in future trials and clinical follow-up, but, to our best knowledge, VascuQol has been only used in one large prospective multicentre study (Nguyen et al., 2006).
3.2
Other and Alternative Measures
Several researchers, mainly in the earlier period of HRQOL investigations in the field of vascular surgery, designed ‘‘home-made tools’’ to selectively investigate specific conditions like night time pain, foot and toe pain at rest, calf muscle cramping, swelling of legs, unhealed sores or ulcers (Gibbons et al., 1995). Johnson et al. (1997) had constructed an ‘‘environment scoring system’’ to assess the suitability of the patients’ home situation. In the same study, the authors also used different generic or condition-specific instruments to measure pain (visual analogue scale/ Burford thermometer), mobility (graded scale), depression and anxiety (Hospital anxiety and Depression scale), independent activities of daily living/self care (Barthe’s score) and lifestyle (Frenchay index). In addition to the generic questionnaires RAND-36 and EuroQol-5D, Bosh et al. (1999) added several valuational measures like time-tradeoff, Standard-gamble and health rating scale. In the latter, patients were asked to rate their current state of health on a scale from 0 to 100. In the > time-tradeoff, patients were asked how many years they would trade in exchange for full health rather than living a full life expectancy in his/her current health status. In the > standard-gamble, patients were asked what risk of death they would willingly take to improve their current state up to full health. Alternatively, several authors used a different approach for assessing QOL either because they felt usual questionnaires could be difficult to administrate, nonvalid or inappropriate in specific communities, like Caribbean patients (Deneuville, 2006). The domains of > maintenance of ambulation and > independence in the residential and living status were pragmatically assessed in a real-world belief that these endpoints represent also major concerns for the patients with LEAD seeking surgical care and, therefore, major determinants in their quality of life (Nicoloff et al., 1998; Taylor et al., 2005; Kalbaugh et al., 2006; Deneuville, 2006).
4
Impact of Revascularization on HRQOL
It has been first argued that revascularization provided the best possible chance for rehabilitation and improvement of the quality of life but the scientific proof of these statements remained unavailable many years after the introduction of OVS and ET. When reporting standards for OVS were originally established, a successful arterial reconstruction was defined as patency of the conduit (permeability rates), increased limb perfusion (ABI) and for CLI, limb salvage. It was assumed that these relevant objectives from a technical/surgeon-oriented point of view would correlate with the unstated ultimate goals which are improvement or maintenance of lower limb function, ambulation and perhaps survival. Over the years, however, a growing body of evidence indicated that optimal technical results were not always translated to improved function or increased survival.
Quality of Life After Revascularization and Major Amputation for LEAD
138
First, despite significant decrease in post operative morbidity, revascularization procedures are commonly associated with complications perceived as minor from the surgeon’s point of view but yet impairing the patient’s quality of life. Decreased sexual ability is commonly an area of concern and dissatisfaction for patients after aortoiliac surgery (Hallin et al., 2002). Infrainguinal VBGs, notably for CLI, are associated with a 11–32% incisional wound complication rate (Novgren et al., 2007; Goshima et al., 2004; Basil, 2005). Healing of ischemic tissue loss occurs in average 15–20 weeks after surgery (Novgren et al., 2007) and requires reoperation in the first 3 months in almost half of the patients (Goshima et al., 2004). Taken together, wound complications and tissue loss may result in a readmission rate up to 50% in the first 3 months, not taking into consideration the pain, discomfort and anxiety experienced for the outpatients during wound management. Another area of concern after infrainguinal VBGs is the surveillance requirement and the subsequent revision of stenosis or failing bypass in 20–40% of cases at 3–5-year follow-up (TASC II). An ideal outcome defined as patent graft without revision, healed wounds and independence in living and ambulation was seen in only 14% of patients (Nicoloff et al., 1998). Redo surgery (secondary patency) is generally associated with a lesser improvement in HRQOL than in patient with primary surgical success (Tangelder et al., 1999; Klevsga˚rd et al., 2001). Indeed, the adverse outcomes as observed after OVS are significantly reduced after endovascular procedures. Since its creation in 1964 by Dotter et al., the use of PTA and its later refinements (ET) have increased exponentially in the management of LEAD. The rapid and widespread acceptance of ET in the clinical practice is supported by intuitive and generally accepted advantages. Those include low procedural morbidity and mortality, early technical success exceeding 90% (Novgren et al., 2007) for most lesions, the speed with which the intervention can be performed and the reduced hospital stay. However, higher immediate failure rate (20%) and redo at 12 months (27%) are commonly reported (Chetter et al., 1999; Feinglass et al., 2000; BASIL, 2005). Undertaken in the UK, the BASIL study was a randomized controlled trial to compare the two treatment strategies in CLI. Among 452 selected patients, 228 were assigned to receive a bypass-first and 224 a PTA-first strategy. The primary endpoint was amputation (of the trial limb) free survival. The trial ran for 5.5 years. At the end of follow up, 55% of patients were alive without amputation, 16% had lost the trial limb with 8% subsequently dead following the amputation and 29% of patients died with their intact limb. The authors concluded that the two strategies were associated with broadly similar outcomes in terms of amputation freesurvival but, in short-term, surgery was associated with a higher morbidity and was more expensive than PTA. However, after 2 years, the data analysis strongly suggested that surgery provides a significantly reduced risk of future amputation, death or both. Both strategies improved similarly the HRQOL at 3 months (Euro5D and SF-36 physical component summary scores), a result that was largely sustained during follow-up. A weak but not significant trend toward better HRQOL was seen in the bypass group. Second, it was found that there is a functional decline in several patients groups with specific morbidities despite successful revascularization. In a large cohort of 841 patients followed in an average of 24 months after revascularization for CLI (endovascular procedures, 35%, open surgery, 62% or both), Taylor et al. reported that the main independent predictors of dismal functional outcome and survival were an impaired ambulatory capacity before surgery and the presence of dementia. Those patients did not only experience a 1.5 to threefold risk of death but also a twofold to threefold risk of ambulatory deterioration and a sixfold risk
2361
2362
138
Quality of Life After Revascularization and Major Amputation for LEAD
of losing their independent living when compared to referent controls. Patients with dementia or walking disabilities at presentation had a decreased survival and poor functional results even when compared with the cohort of patients who lost their limb (Taylor et al., 2005). In the same line, Gibbons found that patient’s perceptions of good function and well being at baseline were predictive of satisfactory function and well being after revascularization (Gibbons et al., 1995). Klevsga˚rd et al. observed that improvement of HRQOL was associated with the patient’s sense of coherence (Klevsga˚rd et al., 2001). Third, several studies reported counterintuitive improvement of HRQOL after failed OVS (Tangelder et al., 1999) or PTA (Klevsga˚rd et al., 2001). Finally, functional impairment in patients with LEAD is far from being confined to hemodynamic abnormalities in the lower limb blood flow. With little doubt, lower limb integrity serve as a major governor of the patient’s physical ability. Likewise, occlusive disease in other vascular beds, target-organ complications of risk factors and the effect of aging have adverse and cumulative effects on the lower extremity function. Those include hip and knee arthritis, overweight, muscle weakness and hemiparesis, breathlessness, angina pectoris, vertigo or decreased visual acuity. Previous population-based studies had shown that 29% of people aged over 65 had mobility-related disability and 11% were living in institutions. In a prospective cohort study of 1122 subjects of 71 years or older in age living in the community and without self-reported disability at baseline, 19% had mobility-related disability at 4-year follow-up and 10% disability in activities of daily living (Guralnik et al., 1995). Back, hip or knee joint disease is present in 23–40% of patients with LEAD (Breek et al., 2002; Rucker-Whitaker et al., 2005), a wide majority of whom being retired (Paaske, 1995). Those figures cannot be ignored in a realistic attempt of improving lower extremity function in patients with LEAD.
5
Impact of Revascularization for Intermittent Claudication on HRQOL
In Western countries, according to large population-based studies, IC affects approximately 5–12% of the population over 55 years. Annual mortality rate in these patients is as high as 5–10%, mainly caused by atherothrombotic events in the coronary and cerebral vascular beds. Data of the natural course of IC in nonwhite populations are scarce but indicate differences in prevalence, mode of presentation (Rucker-Whitaker et al., 2005) and comorbidities (Deneuville et al., 2008). The available literature clearly indicate that most of the deterioration of HRQOL perceived by claudicants are attributable to the limitation in the patient’s daily routine physical activities, interference with social activities, problems at work and limitations of other activities as a result of impaired physical health (Pell et al., 1995, Khaira et al., 1996; Breek et al., 2002). Currently, surgery is indicated for individuals with a vocational or lifestyle disability due to IC (ACC/AHA 2006; Novgren et al., 2007). From the pathophysiological point of view, the restoration of an adequate arterial blood flow in the lower limb during exercise has been regarded for many years as the primary endpoint. Indeed, a considerable amount of surgical reports have repeatedly proven that revascularization for IC does increase arterial blood flow as documented by a significant rise in ABI at rest and after exercise (Currie et al., 1995; De Vries et al., 2005).
Quality of Life After Revascularization and Major Amputation for LEAD
138
Surprisingly, it had been shown that ABI changes correlate poorly with improvement in HRQOL. This finding was consistently present whatever the generic (Currie et al., 1995) or disease specific domains and indices tested (Chong et al., 2002; De Vries et al., 2005; Mehta et al., 2006). In that line, of particular relevance, most studies comparing PTA and OVS for IC indicate similar improvement despite objectively higher increase in ABI -in an average of 0.3– 0.4- after OVS (Regensteiner et al., 1993; Currie et al., 1995; Feinglass et al., 2000) while less significant changes 0.1–0.2 following ET (Whyman et al., 1996; Mehta et al., 2006). By contrast, objective measurement of improved walking ability (ie, treadmill testing or 6-minute walk test) have better correlation with improvement in HRQOL (Regensteiner et al., 1993; Metha et al., 2006). Klevsga˚rd et al. (2001) reported significant improvement in the NHP’s dimensions of pain, physical mobility, energy and emotional reactions after successful revascularization in claudicants in whom treadmill walking distance also improved compared to patients without increased walking distance at 12-months follow-up. This suggests that walking ability is more important for the claudicant that total lower limb blood flow and that reduction in walking ability has a much greater impact than increase in ABI. There is little doubt that successful revascularization immediately improves the HRQOL in claudicants with a lasting benefit for at least 12 months. This consistent finding is also observed either with a patent graft (Regensteiner et al., 1993; Feinglass et al., 2000; Klevsga˚rd et al., 2001) or PTA (Whyman et al., 1996; Chetter et al., 1999; Bosch et al., 1999; Kalbaugh et al., 2006, Deutschmann et al., 2007; Nylaende et al., 2007). > Table 138-1 summarizes the results from four randomized controlled trials (Whyman et al., 1996; Bosch et al., 1999; Nylaende et al., 2007; Sabeti et al., 2007), eleven prospective observational studies, three of which being aimed at validation (Chong et al., 2002) or comparaison of HRQOL questionnaires (De Vries et al., 2005; Metha et al., 2006). A considerable improvement is commonly observed in all dimensions of the NHP (Klevsga˚rd et al., 2001), at least in the physical functioning SF-36 scores (Chetter et al., 1999; Kalbaugh et al., 2006; Deutschmann et al., 2007, Nylaende et al., 2007), all the domains assessed by RAND-36 questionnaire (Bosh et al., 1999), the EuroQOL visual analogue scale (Chetter et al., 1999) and the EuroQoL-5D scores (Bosh et al., 1999; De Vries et al., 2005). Valuational measures (like time trade-off, health utilities index) are also improved (Chetter et al., 1999; Bosh et al., 1999). Regarding disease-specific questionnaire, all four HRQOL instruments showed a significant improvement compared to the baseline values after revascularization, mainly with PTA. This finding was consistently independent of the tool used, the WIQ (Regensteiner et al., 1993), the Intermittent Claudication Questionnaire (Chong et al., 2002), the CLAU-S (Metha et al., 2006; Nylaende et al., 2007), the SIPic (Metha et al., 2006) and the VascuQol. However, the latter was the most responsive instrument to detect small changes after treatment when compared with the other disease-specific and generic questionnaires (De Vries et al., 2005; Metha et al., 2006). An additional proof of the benefits from revascularization is the significant deterioration of HRQOL in significant restenosis not followed by a successful re intervention (Deutschmann et al., 2007) and in symptomatic nontreated occlusions (Tangelder et al., 1999; Deutschmann et al., 2007). The most pronounced effects of PTA can be expected within the immediate postinterventional period (Chetter et al., 1999; Bosch et al., 1999; Kalbaugh et al., 2006; Deutschmann et al., 2007) while a trend toward return to baseline values in several domains is commonly
2363
2364
138
Quality of Life After Revascularization and Major Amputation for LEAD
. Table 138-1 Studies investigating the quality of life (QOL) after open vascular surgery (*) and/or endoluminal therapy (}) for intermittent claudication Study design
Patients
Regensteiner, 1993, Denver, Colorado
PS
14*
0
3
WIQ
Walking distance and speed
Currie, 1995, Bristol, UK
PM
186 (34*, 74})
8
12
SF-36
Pain, physical function * Versus } same results
Whyman, 1996a, Edinburgh, UK
RC
62 (30})
–
6
NHP
Pain only better results after } Versus medical treatment
Chetter, 1999, Leeds, UK
PS
108 }
12
6
SF-36 EuroQOL
All domains except general health, emotional downward trend at 6 m
Bosch, 1999b, The Netherlands
RC
254 }
6
24
RAND-36 EuroQOL
All dimensions physical domains, pain and EuroQol most sensitive
Feinglass,2000, Chicago, Illinois
PM
526 (60*,40})
18
19
SF-36 restricted WIQ
Pain, physical function/higher improvement after * Versus }
Klevsgard, 2001, Lund, Sweden
PM
84 (*, } ND)
12
NHP
All domains
Klevsgard, 2002, Lund, Sweden
PS
40 (25* 15})
–
1
SF36 NHP Pain, vitality, physical function all domains NHP
Chong, 2002, London, UK
PS
100 (40})
–
3
ICQ multiple
ICQ most sensitive better results after } Versus medical treatment
De Vries, 2005, The Netherlands
PS
243 (*, } ND)
13
6
multiple
VascuQol most responsive
Metha, 2006, Hull, UK
PS
70
0
6
multiple
All domains VascuQol most responsive
Kalbaugh, 2006, Greenville, S Carol
PS
54}
ND
12
SF-36
Physical function, role physical, pain
Deutschmann, 2007, Graz, Austria
PS
130}
37
12
SF-36
Same results + social function at 6m
Main author, year
Attrition Follow-up % (months) QOL tool
Improvement in QOL
Quality of Life After Revascularization and Major Amputation for LEAD
. Table 138-1 (continued) Study design
Patients
Nylande, 2007c, Oslo, Norway
RC
56 (28})
ND
24
SF-36 Claus-S
Physical function benefits not sustained at 24 m better results } Versus optimal medical treatment
Sabeti, 2007d Vienna, Austria
RC
104}
ND
12
SF-36
ND
Main author, year
Attrition Follow-up % (months) QOL tool
138 Improvement in QOL
a
Balloon angioplasty vs medical treatment (unsupervised exercise) Primary vs optional stent implantation in the iliac artery c Balloon angioplasty + OMT vs optimal medical treatment (OMT) alone d Primary vs optional stent implantation in the femoral superficial artery ND no data; PS Prospective cohort, single centre, RC randomized controlled, PM Prospective cohort, multicenter b
observed at follow-up. This was true for SF-36 mental health, emotional and vitality domains at 1-year (Kalbaugh et al., 2006). In the OBACT study, the reduction in pain during activity and pain severity (CLAUS-S) which was seen at 6 and 12 months disappeared after 2 years (Nylaende et al., 2007). However, for the largest (Bosh et al., 1999) or more recent studies (Nylaende et al., 2007), the positive effects of PTA on physical functioning lasted at least for 2 years. Surprisingly, patients with unsuccessful revascularization with minimal increase in ABI (Feinglass et al., 2000; Klevsga˚rd et al., 2001) still experienced some improvement in pain, emotional reactions and family relationships in the first year. In the sample of the Dutch BOA study (Tangelder et al., 1999), all eight domains of the SF-36 as well as EuroQOL scores were roughly similar in patients with asymptomatic graft occlusions and in those with patent grafts, with a tendency of lower values in the physical domain in the first group. In selected patients, Klevsga˚rd et al. (2001) reported that 35% of bypass grafts and 28% of PTA performed were unable to increase ABI by more than 0.15. As mentioned by Feinglass et al., poor performances than in published academic series may unfortunately be more representative of results across the broad range of practice. In studies comparing the effect of revascularization on HRQOL in claudicants vs CLI, claudicants consistently outperformed patients undergoing treatment for CLI in nearly all domains (Kalbaugh et al., 2006; Deutschmann et al., 2007). These findings provide some evidence that revascularization, especially with ET, could be of benefit in a significant number of claudicants traditionally treated conservatively.
6
Impact of Revascularization for Critical Limb Ischemia
CLI faces the vascular surgeons as one if not the most difficult challenge. CLI is commonly associated with severe multilevel occlusions of the lower limb arteries and an adaptation
2365
2366
138
Quality of Life After Revascularization and Major Amputation for LEAD
failure in the microcirculation. As CLI generally occurs in elderly patients, all age-related nonvascular conditions (notably hip or knee arthritis, muscle deconditioning or vertigo) are likely to impair the lower extremity function more severely than in claudicants. Likewise, various clustering of complications from both known risk factors and atherosclerosis in other vascular beds may adversely impact far beyond the ability to walk (> Table 138-2). It is universally accepted that CLI has a profound detrimental effect on HRQOL. A highly consistent finding in the literature is the significantly lower HRQOL scores in patients with CLI vs claudication independently of the instrument used (De Vries et al., 2005; Tangelder et al., 1999; Klevsga˚rd et al., 2001, 2002; Kalbaugh et al., 2006; Deutschmann et al., 2007). > Figures 138-1–138-3 summarize the mean scores from the generic instruments SF-36 and NHP and from the disease-specific questionnaire VascuQOL in pooled samples of patients from various surgical studies of IC and CLI. From the technical point of view, revascularization in CLI provides fairly good results with 5-year limb salvage rates exceeding 70–75% (Taylor et al., 2006; ACC/AHA, 2006; Novgren et al., 2007). OVS is required more often in CLI due to the multilevel occlusive lesions. The concept of limb salvage relies on an intuitive expectation of preserving a life-style and a sense of well-being that would be lost with limb amputation. Although mortality after OVS for CLI is currently low (1–2%, TASC, 2007), systemic or neurologic complications after operation are common, up to 10–15% (Gibbons et al., 1995; Tangelder et al., 1999, BASIL, 2005). Furthermore, the expenditure of effort to attain limb salvage is considerable, notably in patients with tissue loss. A critical period of 15–20 weeks (Novgren et al., 2007; Goshima et al., 2004) with reoperation, re admission due to delayed healing of surgical wounds/tissue loss and/or redo surgery has profound detrimental effects on the mental and emotional health (Tangelder et al., 1999, N’Guyen, 2006). The late outcomes in patients operated for CLI are disappointing. The 5-year survival rate after revascularization (Taylor et al., 2006; ACC/AHA, 2006) is low (42–50%) with little or no effect of surgery. Gibbons et al. (1995) reported that less than half (47%) of 156 patients in whom an infrainguinal revascularization had been performed for CLI reported being ‘‘back to normal’’ after 6 months despite optimal technical results of graft patency (93%) and limb salvage (97%). Similarly, Seabrook et al. (1999) have shown that despite long-term limb salvage and optimal graft functioning, patients successfully revascularized for CLI remained functionally disabled when compared to age-matched subjects without LEAD, nevertheless they reported similar well-being. In that case-control study, 70 patients with VBGs performed for limb salvage (mean ABI of 0.93, normal flow velocities and no evidence of flow disturbances or morphological defects), with a mean follow-up of 45 months were compared with age and gender-matched controls with normal ABI and no history of vascular occlusive disease. Less than half of patients (47%) reported that they had no problems with the revascularized limb and only 27% were able to report that the operated leg was better than the opposite unoperated leg. Adverse symptoms when ambulating were reported by 38% of the patients and at rest by 36%. Patients had significant decreased in their functional ability to walk for various distances, perform household tasks including those requiring mild energy level and bath. Patients exhibited also significantly less independence in activities of daily living and reported a
4
3
74
Bedridden
Tissue loss, %
59
ND
3
17
79
96
ND
36
57
ND
13
ND
67 (41)
54
1.31
68
Taylor N = 841
0.84
>70 years
74
ND
ND
ND
ND
ND
ND
33
36
ND
ND
21
80 (36)
72
3
4
9
84
97
2
18
24
12
14
26 (15)
62
75
67% >70 year
42
Deneuville N = 501
BASIL N = 452
Critical limb ischemia
75
ND
ND
ND
ND
ND
ND
28
46
ND
12
20
73 (ND)
64
1.77
69
PREVENT N = 1404
NA
NA
NA
NA
NA
ND
ND
ND
47
ND
ND
28 (ND)
16
4.1
69
Feinglass N = 526
NA
NA
NA
NA
NA
ND
ND
–
36
10
4
ND
93 (ND)
18
3.48
64
NA
NA
NA
NA
NA
ND
ND
ND
ND
ND
ND
ND
90 (ND)
22
1.87
70
Chong N = 224
Claudication DeVries N = 348
NA
NA
NA
NA
NA
ND
ND
ND
32
4
ND
66 (ND)
16
2.07
63
Aquarius N = 200
ESRD end stage renal disease; CAD coronary arterial disease; CHF congestive heart failure; Amputation previous major amputation; Vasc surgery previous vascular surgery; ND no data, NA not applicable
20
62
Vasc surgery, %
Wheelchair
41
CAD, %
With assistance
14
CHF, %
70
18
ESRD, %
Ambulatory
15
Stroke, %
5
70 (21)
Smokers (current) %
ND
81
Diabetes %
Home independent
1.65
Sex ratio male/ female
Amputation, %
66
Age (mean)
Gibbons N = 318
. Table 138-2 Comparison of patient’s demographics in large series of critical limb ischemia versus claudication
Quality of Life After Revascularization and Major Amputation for LEAD
138 2367
2368
138
Quality of Life After Revascularization and Major Amputation for LEAD
significant greater need of using cane indoors or walker and wheelchair outdoors. Only 37% of the patients (vs. 67 in controls) left their home as part of their daily life activities which is likely to explain differences in the domains of social functioning, the patient’s group were being more limited than controls to visit friends or relatives, attend religious services or other social events. Despite the presence of persistent symptoms, 91% of patients stated that they were ‘‘glad they had the bypass operation’’ and that ‘‘the discomfort and time in the hospital’’ was worthwhile. Furthermore, there was no significant difference in the general health perception between patients and controls. All of the studies analyzed except one (Hernandez-Osma et al., 2002) reported significant, immediate and lasting benefit on HRQOL after successful revascularization for CLI (> Table 138-3). PREVENT III is, to our best knowledge, the largest study investigating HRQOL following revascularization in patients with CLI. This multicenter (83 North American sites), doubleblind, randomized trial was primarily aimed to evaluate the efficacy (versus placebo) of intraoperative treatment of venous grafts with edifoligide (thought to inhibit venous stenosis due to smooth muscle cell proliferation) to prevent graft failure in patients undergoing infrainguinal VBGs for CLI. The trial was negative for this primary endpoint. As part of PREVENT III, the effect of infrainguinal VBGs was prospectively assessed by using the VascuQOL in 1404 patients at baseline, 3 and 12 months after surgery. The overall results expand on most findings from earlier studies investigating the changes of HRQOL after OVS for CLI. The mean VascuQOL scores increased in all five domains, resulting in significant changes from baseline to 3-month follow-up. At 1-year, the improvement observed was maintained with an additional modest benefit (mean increase, 15%). As expected, patients free from any graft-related events (stenosis >70%, thrombosis, reoperation) experienced the highest increase in HRQOL scores which is in agreement with the results reported after successful revascularization from previous studies (Chetter et al., 1999; Klevsgard et al., 2001; Thorsen et al., 2002; Tangelder et al., 1999). Despite successful graft revision, patients with graft-related events had lower HRQOL scores at 12 months than patients free from any graft-related events. This finding is in accordance with previous studies (Tangelder et al., 1999; Klevsgard et al., 2001) reporting significant increase in pain and sleep disturbances after redo revascularization while another study showed no impact on HRQOL (Thorsen et al., 2002). Some improvement from baseline was still seen after failed revascularization, notably in the domain of pain, a finding reported by previous authors (Chetter et al., 1998; Klevsgard et al., 2001), especially in case of aymptomatic occlusion (Tangelder et al., 1999). In PREVENT-III, multivariate analysis showed that diabetes was also related to a reduced gain in HRQOL after 1-year, a finding in accordance with other studies (Holtzman et al., 1999; Engelhardt et al., 2006). Unfortunately, the robustness of findings from PREVENT III is threatened by the high number of survey nonresponders (48%) who were more likely to be diabetics, nonwhite patients or patients with graft-related events. Amputation had the greatest effect on 12-month nonresponse (88%) precluding any meaningful measurement of the effect of amputation. Currently, the available literature strongly indicates that revascularization does improve HRQOL in patients suffering CLI, although the magnitude of that improvement is not consistent among studies.
Quality of Life After Revascularization and Major Amputation for LEAD
138
. Table 138-3 Studies investigating the quality of life (QOL) after open vascular surgery (*) and/or endoluminal therapy (}) for critical limb ischemia Study design
Patients
Albers, 1992, Sao Paulo, Brazil
PS
61 (14*)
21
12
Thompson, 1995, Leicester, UK
RS
112 (86*)
35
16–18
Self made Depression, mobility, social functioning no difference between 1st or 2nd patent grafts
Paaske, 1995, Aarhus, Denmark
PS
153*
33
18–36
Self made Physical mobility only
Gibbons, 1995, Boston, USA
PM
250*
37
6
Generic Daily living, vitality, composite mental well-being
Johnson, 1997, Sheffield, UK
PS
150 (44*,26})
27
12
Generic Pain, mobility, selfcomposite care, depression (only after*)
Chetter, 1998, Leeds, UK
PS
55*
22
12
Generic SF36
Physical and social functioning, pain, vitality
Seabrook, 1999, Milwaukee, USA
Casecontrol
70*
–
45
Generic Derived SF-36
Optimal bypass: lower versus controls
Holtzman, 1999, Minneapolis, USA
RS
166 (104*,61})
–
12–84
Tangelder, 1999, The Netherlands
RC
405*
9
21
Generic EuroQOL SF-36
Highest in patent grafts and asymptomatic occlusion
Tretinyak, 2001, Milwaukee, USA
PS
46*
ND
3
Generic SF36
Physical functioning
Klevsgard, 2001, Lund, Sweden
PS
62 (ND*/})
14
12
Generic NHP
Pain, sleep, mobility
Hernandez, 2002, Barcelona, Spain
PS
52 (30*)
43
12
Generic SF36
No changes, trend to degradation
Thorsen, 2002, Copenhagen, DK
PS
60*
20
12
Generic NHP
Pain, sleep
Klevsgard, 2002
PS
40 (34* 6})
–
1
Generic Pain, social isolation, SF36/NHP physical mobility NHP more responsive
Main author
Attrition Follow-up % (months) QOL tools Generic QL-index
Improvement in QOL Improved, similar after * Versus medical therapy
Generic Lower improvement SF12, SF36 in diabetics/older patients
2369
2370
138
Quality of Life After Revascularization and Major Amputation for LEAD
. Table 138-3 (continued) Main author
Study design
Patients
Attrition Follow-up % (months) QOL tools
Improvement in QOL
BASIL trial, 2005, UK
RC
452 (228* 224})
2
60
EuroQol SF-36
All domains EuroQoL physical and mental component SF-36
Deneuville, 2006, Guadeloupe, FWI
RS
175*
17
42
Self made Ambulation, daily living, self care > amputees
Kalbaugh, 2006, Greenville, USA
PS
30}
ND
12
SF-36
Pain only
Engelhardt, 2006, Augsburg, Germany
PS
86*
14
6
SF-36
All domains lower in diabetics
PREVENT III, 2006, USA, Canada
RC
1404*
48
12
VascuQOL All domains lower in diabetics and graft related events
Deutschmann, 2007, Graz, Austria
PS
60}
37
12
SF-36
Physical function, pain (social function at 6 m)
+BOA Dutch Bypass Oral anticoagulants or Aspirin study. Comparison oral anticoagulants versus aspirin in the prevention of infrainguinal bypass grafts occlusion; BASIL Bypass versus Angioplasty in Severe ischemia of the Leg. Comparison of infrainguinal bypass grafts versus balloon angioplasty in the treatment of CLI; PREVENT III Comparison of edifoligide versus placebo to prevent vein graft failure after infrainguinal bypass grafts in the treatment of CLI RS retrospective, single centre; PS prospective cohort, single centre; PM prospective cohort, multicentre; RC randomized controlled
While most studies identified significant improvement in the domains of pain, sleep (Thorsen et al., 2002; Kalbaugh et al., 2006), few authors reported a marked benefit in physical functioning (Chetter et al., 1998; Klevsgard et al., 2001; Engelhardt et al., 2006). The increase in mobility remained generally modest and limited to activities not requiring some initiative. (Jonhson et al., 1997; Seabrook et al., 1999). The effect on social functioning (Chetter et al., 1998), anxiety and depression (Jonhson et al., 1997) was variable among studies. However, there is still no evidence on how HRQOL measurements in everyday practice could modify the current decision making process (see section VI).
7
Impact of Limb Amputation on HRQOL
Several condition-specific instruments have been developed to measure and assess HROL in the prosthetic practice which were recently reviewed (Gallagher and Desmon, 2007). This section rather focuses on the available data of HRQOL after limb amputation related to LEAD which is known to be predominant in Western countries.
Quality of Life After Revascularization and Major Amputation for LEAD
138
. Table 138-4 Studies investigating the quality of life (QOL) after major amputations of the lower limb Main author Albers, 1992, Sao Paulo, Brazil
Study Patients design A1 (A2)
Able to walk %
Follow-up (months)
QOL tools Changes/baseline
PS
16 (+6)
43
12
Generic QL-index
Unchanged trend: worsening in A2
Case control
149
42
–
NHP
Lesser QoL versus controls driven by mobility
Thompson, 1995, Leicester, UK
RS
17 (+7)
ND
16–18
Johnson, 1997, Sheffield, UK
PS
46 (+ND)
ND
12
Generic Improved pain, composite anxiety unchanged self-care deteriorated life style A2
Chetter, 1998, Leeds, UK
PS
(11)
63
12
Generic SF36
Holtzman, 1999, Minneapolis,USA
RS
(28)
36
12–84
Tangelder, 1999, The Netherlands
RC
36 (+38)
ND
21
EuroQOL SF-36
Deneuville, 2006, PAP, FWI
RS
78 (+35)
34
42
Self made Deteriorated daily living, mobility Scores A1 = A2
PREVENT III, 2006
RC
(125)
ND
12
VascuQOL Unconclusive 88% nonresponse
Pell, 1993, Edinburgh, UK
Self made Scores A1 = A2
Improved pain, vitality, social function, mental
Generic ND SF12, SF36 Deterioration all domains except pain worsening in A2
A1 primary amputation, A2 secondary amputation (following failed revascularization) RS retrospective, single centre; PS prospective cohort, single centre; PM prospective cohort multicentre; RC randomized controlled; ND no data
Prospective data on HRQOL after limb amputations for LEAD are scarce in the literature and reflect a generally poor outcome (> Table 138-4). Several studies (Johnson et al., 1997; Chetter et al., 1998; Tangelder et al., 1999) indicate that amputation carries a significant improvement of pain. Chetter et al. reported that amputees had higher SF-36 scores in the psychosocial domains than patients with a patent revascularization. According to the authors, this favorable outcome is mainly explained by pain relief and realization that the problem of life as an amputee can be overcome.
2371
2372
138
Quality of Life After Revascularization and Major Amputation for LEAD
HRQOL studies in dysvascular amputees suggest a strong link between mobility and other physical and nonphysical domains. Among the very first publications in that field, Pell et al. published a large retrospective survey of 149 amputees still alive at a median interval of 38 months after operation. Assessed by using NHP, amputees had significantly the worst scores for all domains compared with age and sex matched controls. After stepwise logistic regression analysis, physical mobility was the only health domain for which the difference between amputees and controls remained independently significant (Pell, 1993). This finding confirms the intuitive statement that impaired mobility account for much of the social isolation and emotional disturbances experienced by amputees. This finding is relevant because the maintenance of mobility, although best reached after prosthetic rehabilitation can be also obtained through wheel chair ambulation. In addition, dysvacular amputees are characterized by low rates of rehabilitation. In a recent large retrospective review of 553 amputees, Taylor et al. (Taylor et al., 2005) reported a maintenance of ambulation rate of 51% at 2-year follow-up, which is in accordance with the findings from other modern (Nehler et al., 2003; Holtzman et al., 2004; Deneuville, 2006) or earlier studies (Pell, 1993). Nehler et al. reported that, despite an aggressive rehabilitation program, less than one fourth of patients were able to walk out of their homes at 10 and 17 months, and that many patients ambulated indoors only in a limited fashion with an assistance device (Nehler et al., 2003). These apparently disappointing results reflect the poor general condition of patients requiring limb amputation performed in the ultimate course of CLI. Thus, as discussed in the previous section, various impairments commonly associated clearly preclude any attempt at rehabilitation. Yet, unsuccessful prosthetic rehabilitation is not synonym of complete loss of ambulation. In older patients or in whom rehabilitation failed after above knee amputation (60–70%), an independent ambulation indoors may be maintained with assistance devices/ wheelchair and even outside with the use of wheelchair/adapted motor vehicle (Nehler et al., 2003; Taylor et al., 2005). Although no specific data on HRQOL in such sample of patients is currently available, pain relief and maintenance of ambulation probably have a favorable impact. Indeed, the degree of handicap may depend upon environmental adaptation at home and outside – car equipment and urban facilities. By contrast, younger healthy amputees with below-knee amputation achieve functional results similar to that might be expected after successful revascularization (Chetter et al., 1998; Nehler et al., 2003; Taylor et al., 2005; Deneuville, 2006) and experience fairly good HRQOL. There is conflicting data about the effect on HRQOL of secondary amputations performed after failed bypass. Several studies reported dramatic impact on HRQOL in all domains except pain (Tangelder et al., 1999) or severe and lasting limitations of physical functioning despite relatively fair rehabilitation rates (Chetter et al., 1998). By contrast, other studies showed little (Albers et al., 1992) or no difference between primary amputees (without revascularization) and those with amputations performed subsequently to a failed bypass (Johnson et al., 1997; Deneuville, 2006). Intuitively, failed attempts at limb salvage are likely to have detrimental effects during the entire process of major surgery often lasting several weeks. In addition of marked physical decline, this period also result in profound psychological deterioration. Several patients may decline an immediate amputation which will be subsequently required for lifethreatening complications. According to Chetter et al. (1998) patients with nonredeemable
Quality of Life After Revascularization and Major Amputation for LEAD
138
occluded graft who declined amputation not only return to the dismal HRQOL experienced before operation but also demonstrated further deterioration in psychosociological domains of the SF-36. This is may be expected as they all the effort and pain had been experienced in vain. This undesirable sequence must certainly be avoided through a more appropriate selection of patients possibly based on pre operative HRQOL measurements.
8
Current limitations/Future Outlook of QOL Assessment of Patients with LEAD
The current data analysis on HRQOL measurements in LEAD identifies several limitations and pitfalls. First, in most research studies, patients with communication difficulties and nonvascular disease in the lower limbs restricting their walking ability are generally excluded in an effort to increase consistency. The percentage of excluded patients ranged from 7 to 21% (Gibbons et al., 1995; Klevsgard et al., 2002; Thorsen et al., 2002; Hernandez-Osma et al., 2002; De Vries et al., 2005). The improvement of HRQOL reported in those favorable samples of selected patients are likely to be the upper limit expected. Second, the number of nonresponders which is at least 14–20% of the sample studied (Klevsgard et al., 2002; Thorsen et al., 2002; De Vries et al,, 2005) exceeds 50% in some occasions (Gibbons, 2005; Deutschmann et al., 2007; N’Guyen, 2006). Furthermore, nonresponders are more likely to have surgical complications and to be current smokers (Gibbons, 2005). In the prospective study PREVENT-III, the rate of nonresponders at 12 months was 52%. Therefore, the lack of data acquisition is likely to lessen, if not negate, the improvement in HRQOL which was identified after VGB. These drawbacks raise the question of the generalization of data of HRQOL studies for all patients considered for surgical treatment of LEAD. Third, there is clearly a link between LEAD and depressed mood, found in up to 30% of patients. It has been suggested that revascularization could improve depression symptoms (Johnson et al., 1997). By contrast, recent studies identified depression (Cherr et al., 2007) as a factor of adverse outcome after revascularization and distressed personality as an independent predictor of less improvement in HRQOL (Aquarius, 2007). Likewise, Thorsen et al. (2002) found that dissatisfied patients who experienced failed surgery with or without subsequent limb amputation, unhealed tissue loss or diabetic neuropathy had greater distress preoperatively in the domains of emotional reactions, social isolation and energy. The poor perceived HRQOL for those patients could not be only explained by the presence of comorbidities, a finding suggesting that preoperative poor perceived health is an independent predictor of revascularization failure (Thorsen et al., 2002). In the same line, it has been previously shown that the disease severity did not have significant effect on mental health and emotional role (Pell et al., 1995). Breek et al. found in claudicants that risk factors and comorbidities adversely affect HRQOL in nonphysical components. With more comorbid diseases, patients had lower scores on the facets of overall health, general health, energy, fatigue and showed more dependence on medications and treatment (Breek et al., 2002). These intuitive findings are in accordance with other studies showing additional negative effects of diabetes (Holtzman et al., 1999;
2373
2374
138
Quality of Life After Revascularization and Major Amputation for LEAD
Engelhardt et al., 2006; N’Guyen, 2006) and end stage renal failure (Rajagopalan et al., 2006) on HRQOL in patients with LEAD. Furthermore, the addition of patient-perceived HRQOL as an outcome measure may influence a trend toward more operations in those patients underscoring their HRQOL and placing unrealistic expectations in surgery. This profile is likely to be found in subjects with behavior counterproductive for long-term outcome (i.e., reluctant to required lifestyle changes like smoking cessation). Feinglass et al. reported that among 277 claudicants patients treated conservatively, 40% were engaged in regular physical exercise while in patients oriented to OVS and PTA, only 15 and 9%, respectively, agreed in such program (Feinglass et al., 2000). The devastating effects of failed revascularization in younger patients with more virulent form of atherosclerosis and forcefully seeking for surgery are well-known outcome in every-day practice. If possible, surgery should be avoided below 50 years (ACC-AHA guidelines 2005). Therefore, the complex interaction between depressed mood and chronic illness associated with LEAD is an important HRQOL parameter that is needed to be addressed in future studies. Current data strongly suggest the benefits of including HRQOL as an additional outcome measure in treatment of LEAD. The first step should clearly be directed toward large data collection on the HRQOL status of patients at baseline and regularly after treatment in prospective studies and in the clinical setting. Therefore, an international standardization should be recommended. There is some evidence that the Vascular Quality of Life Questionnaire (De Vries et al., 2005; Metha et al., 2006; N’Guyen, 2006) is the best candidate to assess HRQOL in the whole spectrum of LEAD. In selective researchs on the impact of treatment in claudicants, the TASC II (Novgren et al., 2007) recently recommended to use a validated, disease-specific health status questionnaire; or the physical functioning domain of a validated generic health status questionnaire. The second step should be a direct application of baseline HRQOL measurements in the decision-making process. For instance, patient-based initial assessment could support a more aggressive approach for claudicants (Feinglass et al., 2000, Taylor et al., 2005). The ‘‘conservative’’ or ‘‘best medical’’ treatment commonly fails because the demand placed on the patient is unaffordable and unrealistic. A new concept in the treatment of vocational or lifestyle limiting IC could be aimed at improving QOL with first-intent endovascular therapy to increase the patient’s performance during exercise therapy and training and encourage him/her to enroll more efficiently in smoking cessation programs and lifestyle changes. Failure to improve HRQOL, patient’s awareness and compliance to all therapies during this more favorable period with high benefit/risk balance and focused lesion highly amenable to endovascular therapy should be seen as a missed opportunity. Regarding the optimal method of assessing interventions in CLI, the role of HRQOL is more questionable. The aim to return healthy individual to their place in the society is unrealistic. Treatment should attempt to maintain the QOL at the premorbid level. Possibly, HRQOL measurements may help to address a difficult issue is the decisionmaking; who will be best served by primary amputation and who will cope well with all the demanding process of revascularization.
Quality of Life After Revascularization and Major Amputation for LEAD
138
Further investigations in the field of decision making analysis (Brothers et al., 2007) are also needed to highlight whether HRQOL measurements are equally efficient alone or integrated in such models.
9
Conclusion
Revascularization has positive effects on various domains of HRQOL in patients with severe forms of LEAD. The highest improvement in physical function is obtained in claudicants, an effect lasting for at least two years after successful OVS or PTA, while more limited improvement are generally observed in the domain of pain in patients treated for CLI. Despite a growing number of publications in the last ten years, prospective data are still lacking in many medical and surgical aspects of HRQoL in patients with LEAD. The measurement of HRQOL is clearly needed at baseline and after OVS and ET but the future role of patient-oriented outcome in the decision making process is yet to be defined. This critical issue is of particular relevance in the projected progression of LEAD related to type II diabetes mellitus all over the world and graving in most industrialised countries.
Summary Points LEAD is associated with impairment not only in physical domains but also social function,
emotional and mental health in most generic and disease-specific questionnaires assessing HRQOL. LEAD is commonly associated with many risk factors (i.e., diabetes mellitus, metabolic syndrome, obesity, end stage renal failure) each being capable to deteriorate HRQOL independently. Successful revascularization immediately improves the HRQOL in claudicants with a lasting benefit on physical functioning for at least 12 months while a trend toward return to baseline values in mental health, emotional and vitality domains is commonly observed. Patients with unsuccessful revascularization with minimal increase in lower limb blood flow still experience some improvement in pain, emotional reactions and family relationships in the first year. Significant, immediate and lasting benefit on HRQOL is seen after successful revascularization for critical limb ischemia although less pronounced than in claudicants. Despite long-term limb salvage and optimal graft functioning, patients successfully revascularized for CLI remain functionally disabled when compared to age-matched subjects, nevertheless they report similar well-being. Younger healthy amputees with successful prosthetic rehabilitation after below-knee amputation experience fairly good HRQOL while acceptable maintenance of HRQOL at the premorbid level is observed in older amputees though pain relief and ambulation with assistance devices. Further investigations are needed on the measurement of HRQOL before and after OVS and ET to best define the future role of patient-oriented outcome in the decision making process of vascular surgery in LEAD.
2375
2376
138
Quality of Life After Revascularization and Major Amputation for LEAD
Appendix Key facts of revascularization for lower extremity arterial disease Level of disease
Type
Aorto- iliac
BPGa
Material Pro
biAxF Pro
Mortality % 3.3 7
Indications Bilateral, Long aortic lesion Rasch analysis has been used in the development of The Quality of Life Impact of Refractive Correction (QIRC) questionnaire which facilitated optimization of question inclusion, unidimensionality and valid linear scoring. The QIRC questionnaire, and others, readily demonstrates the benefits of refractive surgery. However, QIRC is also sensitive to the negative impacts of surgical complications, providing a global assessment of QOL outcome. The Quality of Life Impact of Refractive Correction (QIRC) instrument is demonstrably superior to other instruments in terms of validity and reliability, so is the ideal outcome measure for laser refractive surgery. List of Abbreviations: ADVS, Activities of Daily Vision Scale; ICF, International Classification of Functioning, Disability and Health; LASEK, Laser assisted sub-epithelial keratectomy; > LASIK, Laser in situ keratomileusis; NEI-RQL, The National Eye Institute Refractive Quality of Life questionnaire; PRK, > Photorefractive keratectomy; PRO, Patient reported outcome; QIRC, Quality of Life Impact of Refractive Correction questionnaire; QOL, Quality of Life; RK, Refractive keratectomy; RSVP, The Refractive Status Vision Profile; WHO, World Health Organization
1
Introduction
Refractive errors of the eye are common, and represent a significant burden in terms of disability arising from uncorrected refractive error and economically in the cost of the correction of refractive error. Refractive errors can be treated with spectacles, contact lenses or refractive surgery. Laser refractive surgery has become a highly refined and successful treatment increasingly performed worldwide. The success of laser refractive surgery is typically assessed using objective clinical measures such as postoperative uncorrected > visual acuity and residual refractive error (Waring, 2000). However, these measures do not necessarily correlate well with patients’ postoperative subjective impressions (McGhee et al., 2000). Ultimately, the patient’s perspective is an important outcome of refractive surgery and a number of patientreported outcomes (PROs) have been measured. The most important PRO domain is quality of life (QOL) as this includes all issues that impact a person undergoing laser refractive surgery. Several useful instruments exist including the Quality of Life Impact of Refractive Correction (QIRC) questionnaire (Pesudovs et al., 2004), the Refractive Status Vision Profile (RSVP) (Schein, 2000) and the National Eye Institute Refractive Quality of Life (NEI-RQL) (McDonnell et al., 2003b). These instruments, and other less formal questionnaires, have been used to show the improvement in QOL that occurs with laser refractive surgery, as well as the impact of complications that may arise (Ben-Sira et al., 1997; Garamendi et al., 2005; McDonnell et al., 2003b; McGhee et al., 2000; Schein et al., 2001). In this chapter we outline the important QOL instruments for measuring PRO of laser refractive surgery.
Quality of Life After Laser Surgery for Eye Disorders
2
139
Refractive Error
Refractive error or > ametropia is a very common disorder affecting millions of people and causes poor vision, headaches and eyestrain in addition to other less common symptoms. In a normal, emmetropic eye, the power of the refracting surfaces of the eye, the > cornea and > crystalline lens, are perfectly matched to the axial length of the eye so that light from a distant object is focussed onto the > retina, the light sensitive lining of the eye, performing a function analogous to camera film (> Figure 139-1). Ametropia is caused by a mismatch of the refractive power of the eye and its axial length (> Figure 139-2). If the cornea and lens are too powerful and/or the eye is too long, then light from an object in the distance will be focused in front of the retina and distance vision will be blurred, although vision of close-up objects will be fine. This type of ametropia is called > myopia or short or near sightedness. If the cornea and lens are too weak and/or the eye is too short, then light from an object in the distance will be focussed behind the retina. This type of ametropia is called > hyperopia or long sightedness. Since the eye can increase its refracting power by adjusting the shape of the lens (this is called > accommodation and describes how the eye is able to focus on near objects) the hyperopic eye tries to focus, the power of the lens and eye increases and the distant image can be made clear. The problem then becomes that there is less focusing power to bring close objects into focus as some of the focusing power has been used to see in the distance. Distance vision is typically fine, but near vision can be blurred. An individual with a lot of focusing
. Figure 139-1 The human eye shown in cross-section (courtesy of Dr. Karen Hampson). The human eye as though cut in half from front (cornea) to back, with the important structures named. The only parts of the eye that are visible are the white sclera surrounding the colored part of the eye, the iris. The iris has a central hole that allows light through, the pupil. The cornea and lens are transparent
2381
2382
139
Quality of Life After Laser Surgery for Eye Disorders
. Figure 139-2 Refractive error as illustrated by the focus of parallel rays of light arising at infinity (courtesy of Dr. Karen Hampson). The focus of light from a far away object is shown to focus on the retina at the back of the eye in emmetropia, in front of the retina in myopia (short sightedness), and behind the retina in hyperopia (long sightedness)
power would be able to cope with a slight amount of hyperopia. Other people suffer from headaches due to exerting focusing effort for long periods. Unfortunately as we age, our focusing power decreases until we have relatively little at the age of 40–45 and none left at the age of about 55–60 (Charman, 1989), so that hyperopia becomes increasingly common after 40.
3
Disease Burden of Refractive Error
Estimates of the number of people worldwide with refractive error range from about 800 million to 2.3 billion. The accuracy of estimates is hampered by variation across gender and ethnicity and variation with the type of refractive error. Data from 29,281 people in the US, western Europe and Australia over 40 years of age showed a prevalence for hyperopia (3 > diopters or greater) of 9.9%, 11.6%, and 5.8%, respectively, and for myopia (1 diopter or more) 25.4%, 26.6%, and 16.4% for these population samples (Kempen et al., 2004). Amongst Chinese in Singapore, the prevalence of myopia is twice as high (Wong et al., 2000), although
Quality of Life After Laser Surgery for Eye Disorders
139
in less developed Asian populations the prevalence of myopia is less than in Western populations (Wong et al., 2003). While refractive error is clearly a common problem, corrected refractive error really has only economic burden. However, uncorrected refractive error is a significant public health problem. Worldwide, 259 million people are estimated to have > visual impairment, and 42 million are estimated to be blind (Dandona and Dandona, 2006). Uncorrected refractive error is estimated to be responsible for 98 million cases of visual impairment and five million cases of blindness. Although Vision 2020 (the current WHO global initiative) imposes a mandate to correct refractive errors, little infrastructure and few resources are available to accomplish the task of correcting refractive errors. Access to general medical services is possible for about 25% of populations in developing countries, access to medical eye care, including refraction, could be obtained by only about 10%. It is likely that visual impairment and blindness due to uncorrected refractive error will remain a global public health problem. In the developed world, uncorrected refractive error is also a burden. The economic burden of uncorrected refractive error is of the order of several billon dollars in the USA alone (Rein et al., 2006). However, the economic burden of correcting refractive error in the USA amounts to $5.5 billion per year in direct costs (Rein et al., 2006).
4
Treatments of Refractive Error
There are three main methods for the correction of refractive error; spectacles, contact lenses and refractive surgery. Spectacles dominate the refractive error correction market, in the USA approximately 12% of the adult population wears contact lenses and 6.1 million (2.2%) have had refractive surgery, including 1.2 million (0.4%) in 2002 (Vision Watch, 2003). There are several different types of refractive surgery; incisional procedures such as radial keratotomy (RK) are now obsolete with excimer laser based procedures being the industry standard. Laser in Situ Keratomileusis (LASIK) is the most widely performed refractive surgery procedure in the USA today. Indeed, LASIK dominates the refractive surgery market with all other procedures way behind in terms of volume (Vision Watch, 2003). Photorefractive keratectomy (PRK), which has various synonyms (e.g., LASEK, epi-LASIK, surface ablation), is also a prevalent treatment. PRK and LASIK are fundamentally the same treatment, however, PRK us performed at the surface of the cornea (with only the epithelial cells removed) whereas LASIK is performed within the stroma of the cornea after a thin flap of tissue has been cut and rolled back. In both treatments, the shape of the cornea is altered to change its refractive power such that the ametropia is eliminated. Myopia is more commonly treated with laser refractive surgery than hyperopia, especially with PRK. However, treatment for hyperopia is performed at lower levels of refractive error and with LASIK.
5
Refractive Correction-Related Quality of Life
Quality of life (QOL) is a PRO in which all impacts on a person’s enjoyment of life, ability to function, well being and all other life potentials are assessed. In Medicine, QOL is specifically modified to mean – health-related quality of life. That is, the measurement of healthrelated impacts on all aspects of life. For refractive surgery, the definition can be narrower still, in that the PRO need only tap all aspects of life impacted by the correction of refractive
2383
2384
139
Quality of Life After Laser Surgery for Eye Disorders
error. Many instruments (questionnaires are commonly called instruments in the QOL literature) purport to measure quality of life, but only measure a few dimensions; often vision-related activity limitation only (visual functioning or > visual disability would be more appropriately called vision-related activity limitation to be in line with the World Health Organization (WHO) International Classification of Functioning, Disability and Health (ICF) (World Health Organization, 2001)). However, QOL has many other dimensions e.g., emotional, spiritual, vocational, economical attributes etc. So to purport to measure QOL but to only or principally measure activity limitation means that any inferences one may draw about QOL impacts will be incorrect unless they are confined to activity limitation only. This problem is called construct under representation (Downing and Haladyna, 2004), and is common in vision-related instruments including the popular National Eye Institute Visual Functioning Questionnaire (La Grow, 2007). So the name of an instrument is clearly very important in defining the concept that the instrument purports to measure. The title of the Visual Function Index 14 instrument and the research paper that introduced it quite clearly indicates that it measures activity limitation and does not claim to measure quality of life but it has often been misinterpreted as assessing QOL (Steinberg et al., 1994; Uusitalo et al., 1999; Valderas et al., 2004). For refractive surgery PROs, several QOL instruments do exist, and in the best examples they tap multiple domains of QOL including visual functioning, symptoms, convenience, health concerns, economic concerns and well being (Pesudovs et al., 2004).
6
Questionnaire Technology
Just as LASIK has improved continuously over 10 years due to improved technology, so too has the field of patient-reported measurement improved. Important improvements have occurred in the use of statistical methods for improving the validity and scoring of questionnaires. Considerations in selecting a QOL instrument should include its reliability and validity. Two of the major refractive surgery QOL instruments, the Refractive Status Vision Profile (RSVP) (Schein, 2000, Schein et al., 2001; Vitale et al., 1997, 2000) and National Eye Institute Refractive Quality of Life (NEI-RQL) (Berry et al., 2003; Hays et al., 2003; McDonnell et al., 2003a, 2003b) instruments use traditional Likert scoring (Likert, 1932) where patients’ response scores for a selected set of items are summed to derive the overall score. Likert or summary scoring is based on the hypotheses that all questions (often called items in the QOL literature), have equal importance and equivalent response categories for each question have equal value, so that there are uniform increments from category to category for every question. For example, in a summary scaled visual activity limitation instrument, the Activities of Daily Vision Scale (ADVS), the response category of ‘‘a little difficulty’’ scores four, ‘‘extreme difficulty’’ is twice as bad and scores two, and ‘‘unable to perform the activity due to vision’’ is again twice as bad with a score of one. In cases where the items in an instrument do not have equal importance, the logic of averaging scores across all items becomes questionable. The ADVS ascribes the same response scale to a range of different items, such that ‘‘a little difficulty’’ ‘‘driving at night’’ receives the same numerical score as ‘‘a little difficulty’’ ‘‘driving during the day,’’ despite the former being by far the more difficult and complex task. This rationale of ‘‘one size fits all’’ is flawed in this case, and Rasch analysis has been used to confirm that differently calibrated response categories can help to provide a valid and contextual scale that truly represents QOL (Pesudovs et al., 2003).
Quality of Life After Laser Surgery for Eye Disorders
139
By resolving inequities in a scale arising from differential item difficulty, Rasch analysis provides a self-evident benefit in terms of accuracy of scoring. This process also removes noise from the measurement which in turn improves sensitivity to change and correlations with other variables (Garamendi et al., 2006; Norquist et al., 2004). For example, the standard scoring of the Refractive Status and Vision Profile (RSVP) failed to show any difference in QOL between a group of spectacle and contact lenses wearers in optometric practice and a group of spectacle and contact lenses wearers about to undergo refractive surgery. When Rasch analysis was used to differentially calibrate each item, significant differences between the groups were found, with the pre-refractive surgery group having a lower self-reported QOL than the control group, as might be expected (Garamendi et al., 2006). This occurs through the reduction of noise in the original measurement which chiefly arises from considering all items to be of the same value. Note that conventionally developed instruments can also be reengineered using Rasch analysis (Garamendi et al., 2006; Massof and Fletcher, 2001; Pesudovs et al., 2003) and it is possible to use the Rasch calibrations from these studies to convert summary scaled data from these instruments (Lamoureux et al., 2006; Massof, 2005, 2007). However, this is second best to developing an instrument using Rasch analysis as this technique also gives unparalleled insight into the dimensionality of a questionnaire, so informs the content that should be included in the questionnaire. The importance of Rasch analysis in the development and scoring of questionnaires has been recognized in standards proposed for the assessment of questionnaire quality (de Boer et al., 2004; Pesudovs et al., 2007; Terwee et al., 2007). Clearly, a Rasch-scaled questionnaire should be used wherever possible for the measurement of outcomes. Indeed, calls have been made for this to be the case when assessing refractive surgery outcomes (Weisinger, 2006). There is one Rasch scaled, highly validated instrument developed for measuring quality of life impact of refractive surgery: the Quality of Life Impact of Refractive Correction (QIRC) questionnaire (Pesudovs et al., 2004).
7
The Quality of Life Impact of Refractive Correction (QIRC) Questionnaire
This instrument was developed for the assessing the quality of life impacts of spectacles, contact lenses and refractive surgery (Pesudovs et al., 2004). Visual function, symptoms, convenience, cost, health concerns and well being are included in the content of this instrument which was rigorously developed using literature review, expert opinion, and focus groups. Content was determined using a pilot questionnaire with Rasch analysis for item reduction. A 90-item pilot instrument was implemented on over 300 participants across the United Kingdom. Analysis of these data led to the final 20 item questionnaire (> Table 139-1). QIRC has been ratified as a valid and reliable measure of refractive correction-related QOL by both Rasch analysis and standard psychometric techniques (Pesudovs et al., 2004). QIRC scores are reported on a 0–100 scale which is free of floor and ceiling effects with a higher score representing better QOL and the average score being close to 50 units (> Figure 139-3). The difficulty of the items is well matched to the ability of persons, so QIRC is well targeted to the patients. QIRC has been shown to be responsive to LASIK surgery, demonstrating significant improvements in QOL overall from a mean SD of 40.07 4.30 to 53.09 5.25 (Garamendi et al., 2005). Individual item analysis showed 15 of the 20 items demonstrated statistically significant improvement. Patients reported improved QOL on all five convenience items, both
2385
2386
139
Quality of Life After Laser Surgery for Eye Disorders
. Table 139-1 The 20 items of QIRC Item description 1 How much difficulty do you have driving in glare conditions? 2 During the past month, how often have you experienced your eyes feeling tired or strained? 3 How much trouble is not being able to use off-the-shelf (non prescription) sunglasses? 4 How much trouble is having to think about your spectacles or contact lenses or your eyes after refractive surgery before doing things; e.g., traveling, sport, going swimming? 5 How much trouble is not being able to see when you wake up; e.g., to go to the bathroom, look after a baby, see alarm clock? 6 How much trouble is not being able to see when you are on the beach or swimming in the sea or pool, because you do these activities without spectacles or contact lenses? 7 How much trouble is your spectacles or contact lenses when you wear them when using a gym/doing keep-fit classes/circuit training etc? 8 How concerned are you about the initial and ongoing cost to buy your current spectacles/ contact lenses/refractive surgery? 9 How concerned are you about the cost of unscheduled maintenance of your spectacles/ contact lenses/refractive surgery; e.g., breakage, loss, new eye problems? 10 How concerned are you about having to increasingly rely on your spectacles or contact lenses since you started to wear them? 11 How concerned are you about your vision not being as good as it could be? 12 How concerned are you about medical complications from your choice of optical correction (spectacles, contact lenses and/or refractive surgery)? 13 How concerned are you about eye protection from ultraviolet (UV) radiation? 14 During the past month, how much of the time have you felt that you have looked your best? 15 During the past month, how much of the time have you felt that you think others see you the way you would like them to (e.g., intelligent, sophisticated, successful, cool, etc)? 16 During the past month, how much of the time have you felt complimented/flattered? 17 During the past month, how much of the time have you felt confident? 18 During the past month, how much of the time have you felt happy? 19 During the past month, how much of the time have you felt able to do the things you want to do? 20 During the past month, how much of the time have you felt eager to try new things? The 20 questions identified as the most useful to characterize aspects of quality of life influenced by refractive correction and used in the QIRC questionnaire. QIRC Quality of Life Impact of Refractive Correction Reproduced from Complications in Refractive Surgery, 2008, Influence of refractive surgery complications in quality of life, Pesudovs K, Table 139-1 with kind permission of Springer Science and Business Media
economic items, all four health concern items and on 4 of the 7 items in the well being domain (> Figure 139-4). QIRC is also sensitive to decreased QOL caused by complications of LASIK (Garamendi et al., 2005; Pesudovs et al., 2006). The major complications of LASIK include residual refractive error which can mean that the patient still needs to wear spectacles or
Quality of Life After Laser Surgery for Eye Disorders
139
. Figure 139-3 Person/item map for the QIRC questionnaire. The figure plots the mean score for each item or question from the QIRC questionnaire on the right hand side against subjects on the left represented by # (# ¼ 3 subjects). A higher score indicates a better QOL. The distribution of items is well matched to the distribution of persons (M mean; S1 standard deviation; T two standard deviations) and this question group exhibits excellent targeting of items to subjects. QOL Quality Of Life; QIRC Quality of life Impact of Refractive Correction questionnaire
contact lenses after the surgery and chronic dry eye giving a gritty sensation that often requires regular eye drops to be administered and these lead to reduced quality of life. However, minor complications, like night vision disturbances, may not negatively impact QOL. The QIRC questionnaire effectively differentiates between spectacle wearers, contact lens wearers and post-refractive surgery patients – with the refractive surgery group showing superior QOL (50.23 6.31) than contact lens wearers (46.70 5.49, p < 0.01) and spectacle wearers (44.13 5.86, p < 0.001) (Pesudovs et al., 2006). There were significant differences between scores on 16 of the 20 questions; of the remaining four questions two health concerns and two well being questions did not detect differences between groups (> Figure 139-5). Our study also showed that those spectacle wearers with medium or high refractive error had worse QOL than those with low refractive error. We presume that this is partly due to the nature of spectacles for larger degrees of myopia and hyperopia and also due to the fact that these spectacles would have to be worn full-time as without them the person would have very poor vision. High powered spectacles are relatively heavy with thicker edges (myopic spectacles) or centers (hyperopic spectacles) and either greatly magnify (hyperopic spectacles) or minify (myopic spectacles) the patient’s eyes behind their spectacles. In addition, these spectacles can affect QOL in terms of concerns over cost if patients attempt to improve the appearance and
2387
2388
139
Quality of Life After Laser Surgery for Eye Disorders
. Figure 139-4 QIRC scores before and after LASIK surgery (Reproduced from Complications in Refractive Surgery, 2008, Influence of refractive surgery complications in quality of life, Pesudovs K, > Figure 139-1 with kind permission of Springer Science and Business Media). Mean 1SD scores for each question from the QIRC questionnaire for 66 people with myopia (shortsightedness) before and 3 months after refractive surgery by LASIK. A higher score after surgery indicates an improvement in quality of life related to refractive correction. QIRC Quality of life Impact of Refractive Correction questionnaire, LASIK – Laser in situ keratomileusis
thickness of the lenses by obtaining thin (high refractive index and/or aspheric) lenses with special coatings. It should be noted that the superior QOL for refractive surgery patients comes with a risk in that patients may be one of the small number (approximately 5%) that have significant complications that can lead to reduced QOL (Pesudovs et al., 2006). QIRC was developed to be as short as possible, so contains only 20 items and takes on average slightly under 5 min to complete (Pesudovs et al., 2004). The QIRC questionnaire, scoring spreadsheets and supporting material are available free from http://konrad.pesudovs. com/konrad/questionnaire.html.
8
The Refractive Status Vision Profile (RSVP)
The RSVP is a conventionally developed and Likert scaled instrument developed almost exclusively on a refractive surgery population (92% of subjects) (Schein, 2000), and is really
Quality of Life After Laser Surgery for Eye Disorders
139
. Figure 139-5 Mean 1SD responses on each QIRC question for spectacle, contact lens and refractive surgery groups. Average scores from each question on the QIRC questionnaire for a group of 104 spectacle wearers, 104 contact lens wearers and 104 people who had recently had refractive surgery. A higher score indicates a better quality of life related to refractive correction. QIRC Quality of life Impact of Refractive Correction questionnaire
only valid for refractive surgery and not for spectacle or contact lens wearers. Indeed, the RSVP has been shown to be insensitive to QOL issues relevant to people wearing contact lenses (Nichols et al., 2001). Its 42 items fall into the domains of concern (6 items), expectations (2), physical/social functioning (11), driving (3), symptoms (5), glare (3), optical problems (5) and problems with corrective lenses (7 items) (Schein et al., 2001). The RSVP has been shown to be sensitive to QOL changes related to visual functioning and refractive error, and is responsive to refractive surgery (Schein et al., 2001). Improvements after refractive surgery occurred in the subscales: expectations, physical and social functioning and problems with corrective lenses. The RSVP was developed using traditional techniques, but its psychometric properties were re-evaluated by Garamendi et al. using Rasch analysis (Garamendi et al., 2006). The original 42 item questionnaire showed poor targeting of item impact to patient QOL, items with a ceiling effect, underutilized response categories, a high level of redundancy and no valid subscales. Rasch analysis guided response scale restructuring and item reduction to a 20 item instrument with improved internal consistency and precision for discriminating between groups. It was shown that many of the 14 items relating to functioning and driving were redundant and could be reduced to five items without any loss of accuracy. In addition, many
2389
2390
139
Quality of Life After Laser Surgery for Eye Disorders
of the eight items related to symptoms and glare were shown to be redundant and this could be reduced to three. This is consistent with the content of the QIRC questionnaire, in which the use of Rasch analysis in the development of the original questionnaire identified that patients with corrected refractive error experienced few problems with visual function, and issues of convenience, cost, health concerns and well being were more influential on QOL (Pesudovs et al., 2004). Perhaps the reason why the original RSVP was so heavily weighted with functioning and symptoms questions was because the items were principally determined by clinicians (Schein, 2000), who tend to deal with patients’ presenting complaints of symptoms or functional difficulties, instead of using more objective methodology to discover the important QOL issues for patients when their refractive error has been corrected and their functional vision is normal and they have no symptoms. Most of the time refractive corrections provide adequate functional vision without symptoms and it is only when patients are so motivated by a problem that they visit their clinicians that this is not the case.
9
The National Eye Institute Refractive Quality of Life (NEI-RQL)
The NEI-RQL is a conventionally developed and Likert scaled 42 item questionnaire that included subscales related to clarity of vision, expectations, near and far vision, diurnal fluctuations, activity limitations, glare, symptoms, dependence on correction, worry, suboptimal correction, appearance and satisfaction. The development and validation of the NEI-RQL was spread across three papers and despite rigorous work with focus groups, there is no report on how the final 42 items were selected (Berry et al., 2003; Hays et al., 2003; McDonnell et al., 2003b). The majority of the questions are again related to symptoms or functional difficulties and the content appears to be very clinician-led, similar to the RSVP. However, the NEI-RQL can discriminate between modes of refractive correction (Nichols et al., 2003, 2005) and is sensitive to QOL changes related to visual functioning and refractive error. (Schmidt et al., 2007) Two studies have used the NEI-RQL to demonstrate improved QOL after refractive surgery (McDonnell et al., 2003b; Nichols et al., 2005). The NEI-RQL has not been tested or scaled using Rasch analysis.
10
Other Instruments
The Vision Quality of Life Index is a simple six item questionnaire that does not compute an overall score but compares groups on the basis of agreement with each of six statements. People who had undergone refractive surgery had less concern about four of the six issues included in this questionnaire (Chen et al., 2007). The content of the questionnaire includes concern about injury, coping, friendships, ability to obtain assistance, fulfilling roles and confidence in joining in activities so therefore can be considered a quality of life instrument. This instrument can effectively inform of the rate of concerns within these content areas. The Myopia Specific Quality of Life and the Canadian Refractive Surgery Research Group Questionnaires have been conventionally validated and shown to be responsive to refractive surgery (Brunette et al., 2000; Lee et al., 2005). Other studies that report QOL issues before and after refractive surgery have used informal, non-validated questionnaires (Bailey et al., 2003; Ben-Sira et al., 1997; McGhee et al., 2000; Rose et al., 2000).
Quality of Life After Laser Surgery for Eye Disorders
11
139
Comparing Quality of Life Questionnaire for Refractive Surgery Outcomes
We recently compared several refractive surgery QOL PROs – QIRC, the NEI-RQL and the RSVP, across a comprehensive framework for assessing instrument validity and reliability. This study concluded that QIRC was superior in terms of methods used for item reduction, demonstration of unidimensionality, response scale construction, scoring system and aspects of reliability (Pesudovs et al., 2007). These findings are summarized in > Table 139-2. Disadvantages of the RSVP and NEI-RQL
Likert or summary scoring (Likert, 1932). This scoring method assumes that categories of responses have specific values and that these values represents appropriate positions along a linear scale; this is invariably incorrect (Pesudovs et al., 2003). This method also assumes that all items have the same value; this is also invariably incorrect (Pesudovs et al., 2003). These two assumptions cause the scoring to be noisy and non-linear. This decreases both the accuracy and precision of measurement (Garamendi et al., 2006; Norquist et al., 2004) and leads to an increase in the required sample size for studies. Since neither were developed using Rasch analysis, neither have had the benefits of stringent assessment of internal consistency afforded by Rasch analysis. Rasch analysis gives greater insight into the dimensionality of a questionnaire than available with conventional methods, so items which do not fit the construct under measurement can be excluded. Importantly, redundant items can be identified by Rasch analysis and removed. If a Likert scored questionnaire contains redundant items this leads to overemphasis of that issue within the ‘‘score’’ and this is a serious problem of the RSVP (Garamendi et al., 2006). Conventionally developed questionnaires consistently require modification when assessed by Rasch analysis in that misfitting and redundant items need removal and often the response scale needs revision (Garamendi et al., 2006; Lamoureux et al., 2006; Massof Fletcher, 2001; Pesudovs et al., 2003). Despite four papers outlining the development and validation for the NEI-RQL, nowhere is the method for arriving at the final 42 items reported (Berry et al., 2003; Hays et al., 2003; McDonnell et al., 2003a, 2003b). This is a serious failing in content validity. The NEI-RQL does not allow for an overall score of quality of life, only subscale scores. The validity of these subscales has not been demonstrated and many contain too few items to validly form a subscale. The RSVP has the same problem with invalid subscales. Advantages of QIRC over the RSVP or the NEI-RQL
True linear scoring on an interval scale. More precise measurement of quality of life. More accurate measurement of quality of life. Smaller sample size required for outcome studies. All questions fit the construct of quality of life. Content of the questionnaire driven by patients, not by clinicians, which means that it is more relevant.
2391
üü
üü
QIRC
üü
üü
üü
üü
üü
üü
üü
û
ü
üü
û
û
üü
û
û
Actual Item Item content identifi- reduc- Unidimen- Response area cation tion sionality scale
üü
û
û
Scoring scale
0
0
0
0
üü
ü
üü
üü
ü
Discrim- Converinant gent ‘‘Other’’ validity validity validity
üü
üü
ü
0
0
0
üü
0
0
üü
üü
üb
üü
üü
üb
Rasch separation reliInterpre- Responsability tation iveness
A Rasch-analyzed version of the RSVP (Garamendi et al., 2006) with a modified response scale and a reduced number of items has been shown to have greater responsiveness and test-retest reliability than the standard instrument. It also provides a unidimensional score, statistically justified response and scoring scales and good Rasch separation reliability b Conflicting reports of normative data levels and responsiveness of the RSVP are provided by (Schein et al., 2001) and (Nichols et al., 2001) An assessment of three questionnaires that attempt to assess aspects of quality of life influenced by refractive correction. The assessment rates aspects of their development, reliability and accuracy and scoring is either üü which is the top score and a positive rating; ü is a minimally acceptable or ‘‘just OK’’ score, û which is a negative rating and 0 indicates that this aspect of quality has not been reported to date. More details of the quality assessment are provided in the paper listed in the title RSVP Refractive Status Vision Profile; NEI-RQL National Eye Institute Refractive Quality of Life; QIRC Quality of Life Impact of Refractive Correction. Reproduced from Pesudovs et al. (2007)
a
üü
üü
ü
Intended population
NEIRQL
RSVPa üü
Hypothesis
Testretest reliability
Interobserver or intermode agreement
139
. Table 139-2 Quality assessment of three refractive error-related quality of life instruments: the RSVP, NEI-RQL and QIRC
2392 Quality of Life After Laser Surgery for Eye Disorders
Quality of Life After Laser Surgery for Eye Disorders
139
QIRC is less than half the length of the RSVP or the NEI-RQL, plus a consistent question
and answer format ensures time of completion is less than half. Brevity encourages completion making QIRC more likely to supply a higher participation rate in a study and more complete data. The need for Rasch scaled questionnaires is broadly recognized across ophthalmology (Massof, 2007; Pesudovs, 2006; Spaeth et al., 2006; Weisinger, 2006) There is no justification for using a non-Rasch analyzed questionnaire (both scored and developed using Rasch analysis) when one is available. To do so leaves the study open to criticism of its methodology and therefore its outcome. For refractive surgery outcomes research, QIRC is the preferred instrument.
Summary Points Refractive error is common. Uncorrected refractive error is a significant disease burden worldwide with 98 million people visually impaired and five million people blind from refractive error alone.
Refractive error can be corrected by spectacles, contact lenses or laser refractive surgery. The outcome of laser refractive surgery is assessed by clinical measures of visual
acuity and residual refractive error plus patient reported outcomes, especially quality of life. A number of questionnaires exist for the measurement of quality of life after refractive surgery but all questionnaires are not equal in validity. Rasch analysis is important in the development of questionnaires to optimize question inclusion, unidimensionality and to provide valid linear scoring. A refractive error related quality of life instrument should include a breadth of content areas e.g., well being, convenience and concerns, not just functioning or satisfaction. Quality of life instruments readily demonstrate the benefits of refractive surgery. A sound QOL instrument is also sensitive to the negative impacts of surgical complications, providing an insight into the real impact of the intervention on the person. The ideal QOL outcome measure for refractive surgery would contain broad content, be developed and validated with Rasch analysis and have valid linear scoring. The Quality of Life Impact of Refractive Correction (QIRC) instrument is superior to other instruments in terms of validity and reliability.
References Bailey MD, Mitchell GL, Dhaliwal DK, Boxer Wachler BS, Zadnik K. (2003). Ophthalmology. 110: 1371–1378. Ben-Sira A, Loewenstein A, Lipshitz I, Levanon D, Lazar M. (1997). J Refract Surg. 13: 129–134. Berry S, Mangione CM, Lindblad AS, McDonnell PJ. (2003). Ophthalmology. 110: 2285–2291. Brunette I, Gresset J, Boivin JF, Pop M, Thompson P, Lafond GP, Makni H. (2000). Ophthalmology. 107: 1790–1796.
Charman WN. (1989). Ophthalmic Physiol Opt. 9: 424–430. Chen CY, Keeffe JE, Garoufalis P, Islam FM, Dirani M, Couper TA, Taylor HR, Baird PN. (2007). J Refract Surg. 23: 752–759. Dandona L, Dandona R. (2006). BMC Med. 4: 6. de Boer MR, Moll AC, de Vet HC, Terwee CB, Volker-Dieben HJ, van Rens GH. (2004). Ophthalmic Physiol Opt. 24: 257–273.
2393
2394
139
Quality of Life After Laser Surgery for Eye Disorders
Downing SM, Haladyna TM. (2004). Med Educ. 38: 327–333. Garamendi E, Pesudovs K, Elliott DB. (2005). J Cataract Refract Surg. 31: 1537–1543. Garamendi E, Pesudovs K, Stevens MJ, Elliott DB. (2006). Vision Res. 46: 1375–1383. Hays RD, Mangione CM, Ellwein L, Lindblad AS, Spritzer KL, McDonnell PJ. (2003). Ophthalmology. 110: 2292–2301. Kempen JH, Mitchell P, Lee KE, Tielsch JM, Broman AT, Taylor HR, Ikram MK, Congdon NG, O’Colmain BJ. (2004). Arch Ophthalmol. 122: 495–505. La Grow S. (2007). Optom Vis Sci. 84: 785–788. Lamoureux EL, Pallant JF, Pesudovs K, Hassell JB, Keeffe JE. (2006). Invest Ophthalmol Vis Sci. 47: 4732–4741. Lee J, Park K, Cho W, Kim JY, Kang HY. (2005). J Refract Surg. 21: 59–71. Likert RA. (1932). Arch Psychol. 140: 1–55. Massof RW. (2005). Ophthalmic Epidemiol. 12: 103–124. Massof RW. (2007). Optom Vis Sci. 84: 689–704. Massof RW, Fletcher DC. (2001). Vision Res. 41: 397–413. McDonnell PJ, Lee P, Spritzer K, Lindblad AS, Hays RD. (2003a). Arch Ophthalmol. 121: 1577–1581. McDonnell PJ, Mangione C, Lee P, Lindblad AS, Spritzer KL, Berry S, Hays RD. (2003b). Ophthalmology. 110: 2302–2309. McGhee CN, Craig JP, Sachdev N, Weed KH, Brown AD. (2000). J Cataract Refract Surg. 26: 497–509. Nichols JJ, Mitchell GL, Saracino M, Zadnik K. (2003). Arch Ophthalmol. 121: 1289–1296. Nichols JJ, Mitchell GL, Zadnik K. (2001). Ophthalmology. 108: 1160–1166. Nichols JJ, Twa MD, Mitchell GL. (2005). J Cataract Refract Surg. 31: 2313–2318. Norquist JM, Fitzpatrick R, Dawson J, Jenkinson C. (2004). Med Care. 42: I25–136. Pesudovs K. (2006). BMC Ophthalmol. 6: 25. Pesudovs K, Burr JM, Harley C, Elliott DB. (2007). Optom Vis Sci. 84: 663–674. Pesudovs K, Garamendi E, Elliott DB. (2004). Optom Vis Sci. 81: 769–777. Pesudovs K, Garamendi E, Elliott DB. (2006). J Refract Surg. 22: 19–27.
Pesudovs K, Garamendi E, Keeves JP, Elliott DB. (2003). Invest Ophthalmol Vis Sci. 44: 2892–2899. Rein DB, Zhang P, Wirth KE, Lee PP, Hoerger TJ, McCall N, Klein R, Tielsch JM, Vijan S, Saaddine J. (2006). Arch Ophthalmol. 124: 1754–1760. Rose K, Harper R, Tromans C, Waterman C, Goldberg D, Haggerty C, Tullo A. (2000). Br J Ophthalmol. 84: 1031–1034. Schein OD. (2000). Trans Am Ophthalmol Soc. 98: 439–469. Schein OD, Vitale S, Cassard SD, Steinberg EP. (2001). J Cataract Refract Surg. 27: 665–673. Schmidt GW, Yoon M, McGwin G, Lee PP, McLeod SD. (2007). Arch Ophthalmol. 125: 1037–1042. Spaeth G, Walt J, Keener J. (2006). Am J Ophthalmol. 141: S3–S14. Steinberg EP, Tielsch JM, Schein OD, Javitt JC, Sharkey P, Cassard SD, Legro MW, Diener-West M, Bass EB, Damiano AM, Steinwachs DM, Sommer A. (1994). Arch Ophthalmol. 112: 630–638. Terwee CB, Bot SD, de Boer MR, van der Windt DA, Knol DL, Dekker J, Bouter LM, de Vet HC. (2007). J Clin Epidemiol. 60: 34–42. Uusitalo RJ, Brans T, Pessi T, Tarkkanen A. (1999). J Cataract Refract Surg. 25: 989–994. Valderas JM, Alonso J, Prieto L, Espallargues M, Castells X. (2004). Qual Life Res. 13: 35–44. Vision Watch. (2003). Vision Correction Market Review. Jobson Publishing, New York. Vitale S, Schein OD, Meinert CL, Steinberg EP. (2000). Ophthalmology. 107: 1529–1539. Vitale S, Schein OD, Steinberg EP, Ware JE, Jr. (1997). Invest Ophthalmol Vis Sci. 38: S841. Waring GO, 3rd. (2000). J Refract Surg. 16: 459–466. Weisinger HS. (2006). J Refract Surg. 22: 14–15. Wong TY, Foster PJ, Hee J, Ng TP, Tielsch JM, Chew SJ, Johnson GJ, Seah SK. (2000). Invest Ophthalmol Vis Sci. 41: 2486–2494. Wong TY, Foster PJ, Johnson GJ, Seah SK. (2003). Invest Ophthalmol Vis Sci. 44: 1479–1485. World Health Organization. (2001). The International Classification of Functioning, Disability and Health (ICF). World Health Organization, Geneva.
3
Quality of Life Measures and Indices 3.3 Early Life Stages and Aging
140 Intrauterine Growth Restriction and Later Quality of Life D. Spence 1 1.1 1.2 1.3
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2398 Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2398 Small for Gestational Age Versus Intrauterine Growth Restriction . . . . . . . . . . . . . . . . 2399 Overview of Intrauterine Growth Restriction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2399
2 2.1 2.2 2.3
Etiological Determinants of IUGR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2399 Fetal Abnormality and Infections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2400 Maternal Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2400 Lifestyle Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2401
3
Pediatric Implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2401
4
Long-term Implications for Adult Health . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2402
5 5.1 5.2 5.3 5.4 5.5
Original Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2402 Quality of Life . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2403 Short Form 36 Health Survey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2403 Definition and Measurement of Quality of Life . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2404 1950’s Cohort . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2405 Health-Related Quality of Life . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2405
6
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2407 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2408
#
Springer Science+Business Media LLC 2010 (USA)
2398
140
Intrauterine Growth Restriction and Later Quality of Life
Abstract: > Intrauterine growth restriction (IUGR) remains a major clinical problem in modern obstetrics. While technology has yet to aid significant prevention of IUGR, the past few decades has seen the introduction of some of the most innovative therapies in the history of > neonatal intensive care. This has resulted in an increased survival rate of a heterogeneous group of babies, including those with intrauterine growth restriction. It is therefore important to assess if the associated problems with these babies, impact on health related > quality of life long-term. Subtle psychological and social differences have been reported in childhood and early adult life and higher incidence of coronary heart disease, high blood pressure, and diabetes have also been reported in later life. However, it is unclear how this impacts on overall quality of life. Quality of life is considered an important outcome measure for healthcare interventions in adults and health status measures are key determinants of health service use. Despite the extensive cohort studies undertaken to date, there is a dearth of literature on the relationship between IUGR and quality of life, particularly in later life. A recent study addressed this gap in the literature, comparing subjects born with IUGR and a control group with normal birth weight for > gestation. Quality of life in adulthood was assessed using the Short Form-36 health survey (SF-36). The two groups reported similar quality of life on each of the eight dimensions of the SF-36 and there were no significant differences between them. Adjusting for potential confounding variables did not alter this conclusion. This chapter summarizes the impact of IUGR on later quality of life. List of Abbreviations: AGA, Appropriate for gestational age; CI, Confidence intervals; CMV, Cytomegalovirus; g, Grams; GP, General Practitioner; IUGR, Intrauterine growth restriction; LMP, Last menstrual period; SD, Standard deviation; SF-36, Short Form 36 health survey; SGA, Small for gestational age; SPSS, Statistical package for social sciences; UK, United Kingdom
1
Introduction
Intrauterine growth restriction (IUGR) continues to be a significant clinical problem in modern obstetrics. Whilst there has been substantial progress regarding IUGR, there is still some way to go as many difficulties and challenges still exist in improving clinical outcome and in understanding the scientific mechanism.
1.1
Definition
Growth is a basic fundamental of life and of particular importance is the intrauterine growth of the > fetus. Normal fetal growth has been defined as ‘‘that which is neither significantly restricted nor promoted by extrinsic factors’’ (Robinson et al., 1997, p. 29). It may be assumed that the area of research surrounding growth disturbance is well established, yet definitions which are concise and rigorously laid out are not readily available. IUGR has been described as ‘‘a concept signifying that the fetus has not achieved its optimal growth’’ by Kingdom and Baker (2000, p. 1). Many factors affect fetal growth and inhibit it, yet although there is diversity in the causes of IUGR, the fact that the baby with IUGR would have been bigger were it not for suboptimal genetic and/or environmental factors, is common to all.
Intrauterine Growth Restriction and Later Quality of Life
1.2
140
Small for Gestational Age Versus Intrauterine Growth Restriction
The International Committee at Geneva recommended a definition of prematurity in 1937 (Crosse, 1949). As a result, healthy full-term babies which often weigh less than 3,200 g, perhaps even as little as 2,300 g, were termed premature by definition if the birth weight was below 2,500 g. This meant babies born less than 2,500 g were regarded as premature, regardless of the period of gestation. This birth weight standard inevitably included some full-term but small for gestational age infants. Definitions have changed and since the abandonment of the concept that weight determines age, a variety of terms exist to describe babies who demonstrate altered growth. One such term is ‘‘small for gestational age’’ (SGA), with a synonym ‘‘small for dates.’’ Henriksen (1999), indicates that SGA and IUGR are different concepts and views SGA as a size measurement which may or may not reflect restricted fetal growth. However, as they share core clinical problems, all infants with birth weight at or below the 10th percentile for gestational age are usually considered SGA (Fanaroff et al., 1989). By convention the 10th centile is the most commonly used cutoff (Bakketeig, 1996), however the 5th centile (Strauss, 2000) or even the 3rd centile (Paz et al., 1995) may be used. There are many reasons why a baby may be small and these are related to both maternal and fetal factors. Growth restriction, congenital infection, or congenital malformations are some examples which may lead to, or are associated with small for gestational age. The terms IUGR and SGA are frequently used in an interchangeable manner, however throughout this chapter the term IUGR will be used.
1.3
Overview of Intrauterine Growth Restriction
IUGR is a major cause of > perinatal > morbidity and > mortality (Kilby and Hodgett, 2000). It is defined as less than 10% of predicted fetal weight for gestational age and can result in significant fetal morbidity and mortality if not properly diagnosed (Vandenbosche and Kirchner, 1998). IUGR is associated with increased risk of stillbirth, neonatal death, and other adverse outcomes (Clausson et al., 2001). Most cases of IUGR present during the third trimester, with most fetal deaths involving IUGR occurring after 36 weeks’ gestation and before onset of labor. In accordance with the common definition of IUGR, the expected incidence of IUGR should be about 10%, however, the actual incidence is currently about 4–7%, increasing to around 15–25% in twins. On occasions, some infants thought to have IUGR, in retrospect can be found to be constitutionally small (Vandenbosche and Kirchner, 1998).
2
Etiological Determinants of IUGR
As outlined, IUGR refers to a biological process in which the fetus fails to achieve its genetically programed growth potential and contributing factors which interact, include the external environment, coexistent maternal disease, and the placental and fetal adaptive responses (Kingdom and Baker, 2000). Many different factors may lead to, or are associated with IUGR, including genetic, nutritional, uterine, and placental factors, fetal infection and maternal oxygen-carrying capacity may also be associated (Neerhof, 1995). > Table 140-1 outlines attributable causes of IUGR for which etiological roles are well established.
2399
2400
140
Intrauterine Growth Restriction and Later Quality of Life
. Table 140-1 Etiological determinants of intrauterine growth restriction in a developed country Attributable causes of IUGR Cigarette smoking
Pregnancy induced hypertension
Low weight gain
Congenital anomalies
Low body mass index
Other genetics
Primiparity
Alcohol/drugs
Short stature
Unknown
These are attributable causes of intrauterine growth restriction (IUGR) for which etiological roles are well established. They refer to a developed country in which 25% of women smoke during pregnancy and a substantial minority are nonwhite
2.1
Fetal Abnormality and Infections
Fetal abnormality may be associated with IUGR (Vandenbosche and Kirchner, 1998) and trisomies likely to be associated are trisomy 18 (Edward syndrome) and 13 (Patau syndrome). Infection particularly viral can severely impair the growth potential of a fetus. The classic example is rubella, now uncommon in the United Kingdom (UK) due to widespread immunization, but which remains a problem worldwide. Rubella is an important embryopathic virus and if infection occurs in the first trimester, the fetus is at considerable risk of developing profound abnormalities such as blindness, cardiac malformations, and deafness. This virus appears to have the potential to interfere with the genomic drive to growth, which leads to a growth-restricted baby. There are other viral infections which are associated with growth restriction, including cytomegalovirus (CMV) and varicella (chickenpox), if they infect the fetus in the early weeks of pregnancy. Infections with agents such as Treponema pallidum, are uncommon in the UK, but globally are a significant cause of fetal growth failure (Gross, 1989).
2.2
Maternal Disease
A plethora of different coexistent maternal diseases has been linked to the development of IUGR. IUGR may ensue if any maternal disease is of sufficient severity. Maternal diseases such as hypertension, renal disease, autoimmune disorders, and microvascular disease associated with long-standing insulin dependent diabetes, can increase the risk of a fetus developing IUGR. Such conditions affect fetal growth, as they lead primarily to reduced blood supply via the uterine arteries. This arises from inadequate changes in the maternal spiral arteries at the time of placentation and in the subsequent weeks. Secondary changes known as acute atherosis also occur, a process which narrows the blood vessels still further. These changes combine to restrict blood supply and as a result, substrate delivery to the placental bed and thence to the fetus (Whittle, 1999). This situation is exacerbated if preexisting maternal vascular disease is present. The mother is at particular risk of developing pregnancy-induced hypertension or even preeclampsia and consequently the fetus fails to grow normally (Neerhof, 1995). Maternal illness can also indirectly create growth problems in the fetus, not necessarily from the illness itself, but from the drugs required to control it. A classic example is anticonvulsant
Intrauterine Growth Restriction and Later Quality of Life
140
therapy which is known to be associated with IUGR. It is assumed that these drugs affect cellular mechanisms, rather than the substrate supply previously described.
2.3
Lifestyle Factors
Toxins from smoking and excessive alcohol are also well recognized as being associated with IUGR, however maternal alcohol intake is not as consistent in affecting fetal growth as maternal smoking (Gross, 1989). The evidence linking smoking with IUGR is unequivocal and there is no doubt that fetotoxic effects result from high maternal alcohol intake. At its extreme form, this presents as fetal alcohol syndrome. It is thought that excessive alcohol taken in early pregnancy may act as a cellular poison, thus reducing the fetal potential for growth. The effect of smoking may be due to the vasoconstrictive effect of nicotine in the maternal circulation and also because of the displacement of oxygen from hemoglobin by carbon monoxide. Drugs such as cocaine and ‘‘crack’’ are potent vasoconstrictors which may have adequate influence on the uterine vasculature to be a potential cause of severe IUGR. Demographic studies have long linked inequalities in health to economic and social variables and many lifestyle factors may contribute to fetal growth restriction. Socioeconomic group is a factor which may be related to fetal growth, maternal height is another factor, and small women from certain ethnic groups are more likely to have small babies, but not necessarily babies with IUGR (Whittle, 1999). It is apparent that the impact of other factors which affect fetal size such as smoking and hypertension, is likely to be greatest amongst those women who have constitutional factors which also reduces fetal size. This presents the challenge in maternal-fetal medicine, to develop techniques which will successfully separate out these different factors and their relative contribution.
3
Pediatric Implications
IUGR babies can have a variety of problems, having an increased risk in various perinatal morbidities (Ergaz et al., 2005). Condition at birth is poorer compared with ‘‘appropriate for gestational age’’ (AGA) babies (Ariyuki et al., 1995). In the short-term they have difficulty with temperature control related to their small body mass to surface area, lack of glycogen stores, and low brown fat which reduces their nonshivering thermogenesis. Perinatal hypoxia is also associated with poor glycogen reserves. These babies are at increased risk of necrotizing enterocolitis due to the adaptation to prolonged intrauterine hypoxia associated with fetal gut ischemia. Babies born with IUGR can develop respiratory complications. However, lung maturity is not solely related to IUGR but is dependant on various factors including for example, gestational age. It has also long been recognized that these babies are at risk of neonatal hypoglycemia (Drossou et al., 1995). > Table 140-2 outlines conditions specifically associated with IUGR. The care of these babies can be complex requiring advanced neonatal support. Subsequent development depends both on the extent and etiology of the IUGR and the gestational age at delivery. In infancy and early childhood IUGR infants are at risk for growth failure and neurodevelopmental sequelae, including cerebral palsy, cognitive deficit, and behavioral problems (Yanney and Marlow, 2004).
2401
2402
140
Intrauterine Growth Restriction and Later Quality of Life
. Table 140-2 Conditions specifically associated with intrauterine growth restriction Neonatal condition Perinatal hypoxia Necrotizing enterocolitis Respiratory distress syndrome and respiratory complications Hypoglycemia Polycythemia Anemia, leucopenia, and low platelets Postnatal growth Intrauterine growth restriction increases mortality and morbidity in the neonatal period
4
Long-term Implications for Adult Health
Epidemiological studies have influenced our thinking regarding IUGR. The work of Barker and his colleagues often referred to as the ‘‘fetal origins of adult disease’’ highlights the possibility that those babies born small for dates are more likely to develop hypertension and noninsulin dependent diabetes in adult life (Barker, 1992). It must be noted however, that it was the effect of low-birth weight rather than IUGR which formed the basis of Barker’s studies (Whittle, 1999). These original UK findings have been replicated in different countries, leading to a wide acceptance that low rates of fetal growth is associated with > cardiovascular disease in adult life (> Table 140-3).
5
Original Research
There are major gaps in the knowledge of long-term effects of IUGR. In our study the aim was to determine whether health problems reported in adult life, in particular health-related quality of life at age 50 years, are associated with IUGR at birth. We took account of gestational age and birth weight.
. Table 140-3 Adult diseases associated with fetal growth restriction Adult disease Coronary heart disease
Type 2 diabetes
Stroke
Hypertension
Polycystic ovary syndrome
Osteoporosis
Hormone dependent cancers There is an inverse relationship between birth weight and the prevalence of chronic diseases in adulthood
Intrauterine Growth Restriction and Later Quality of Life
5.1
140
Quality of Life
‘‘Quality of life’’ is one of the fastest growing areas of research and policy. It has become an accepted endpoint in clinical research trials in recent years, particularly as interest in patients’ experiences and preferences has grown. Health related quality of life is the primary outcome of this study and was measured at 48–50 years in two groups; those born with IUGR and those with normal birth weight. Despite the extensive cohort studies undertaken to date, there is a dearth of literature on the relationship between IUGR and quality of life. The past few decades has seen the introduction of some of the most innovative therapies in the history of neonatal intensive care (Saigal and Rosenbaum, 1996). This has resulted in a further decline in neonatal mortality of high-risk babies; however, improvements in morbidity have not been as significant. These major improvements in perinatal care have seen an increased survival rate of a heterogeneous group of babies, including those born preterm or low-birth weight and also those with intrauterine growth restriction (> Table 140-4). For some of these babies who survive there . Table 140-4 Notes on intrauterine growth restriction Birth weight is the most important determinant of perinatal outcome IUGR is associated with many medical problems for the baby, before and after delivery The mechanisms involved in fetal growth are not well understood Various factors (maternal and fetal) may inhibit fetal growth Diagnosis of IUGR is thus complex Both stillbirths and neonatal deaths are strongly associated with IUGR Effective monitoring of fetal growth is of major importance in antenatal care Identification and management of the growth restricted baby are a major challenge for clinicians Key points related to intrauterine growth restriction
may be adverse outcomes and such impairments may profoundly influence health and quality of life during childhood and adult life. These issues have received little attention to date, as focus has been on for example, physical handicap, academic difficulties, and behavioral problems experienced by these babies particularly in infancy, childhood, and adolescence (Hack et al., 1994; McCormick et al., 1992; Saigal et al., 1991), few extend to early adulthood. Concerns regarding the care of babies of borderline viability have resurfaced and the appropriateness of offering intensive care to all babies is being questioned. This creates ethical dilemmas and perhaps ‘‘bad publicity’’. It is therefore important to assess if the associated problems with these babies, impact on quality of life long-term. A quality of life approach is much broader than traditional measures of health outcome, including morbidity and mortality, which insufficiently capture the full spectrum of potential ‘‘health’’ problems. It is essential to promote a wider perspective on health and social wellbeing.
5.2
Short Form 36 Health Survey
The health status measure used to assess quality of life in this study was the Short Form 36 Health Survey (SF-36). This short 36-item questionnaire is used worldwide and covers a wide range of
2403
2404
140
Intrauterine Growth Restriction and Later Quality of Life
areas that may be adversely affected by illness. The SF-36 is a valuable tool in research which has undergone validity testing in the UK including assessment of the content, criterion, and construct validity of the instrument which is widely used particularly as it is appropriate for use with adults of working age (Jenkinson et al., 1993) and within the National Health Service (Garratt et al., 1993). It was chosen as the quality of life measurement in this study, as it is easy to administer and quick to complete and is a comprehensive tool applicable across social and demographic groups (Hayes et al., 1995). It has been used in a cohort of very low birth weight infants to determine if their birth weight had impacted on their lifestyle and quality of life at age 19–22 years. In this study, Cooke (2004) concluded that problems experienced by those born very preterm, were not perceived to influence quality of life in early adulthood. It covers a wide range of areas that may be adversely affected by illness. It measures eight multi-item dimensions (> Table 140-5). For each dimension item scores are coded, summed, and transformed onto a scale from 0 (worst possible health state measured by the questionnaire) to 100 (best possible health state).
. Table 140-5 Short Form-36 dimensions Dimension Physical functioning
Label
No. of items
PF
10
Role limitation due to physical problems
RP
4
Pain
BP
2
General health perception
GH
5
Energy/vitality
EV
4
Social functioning
SF
2
Role limitation due to emotional problems
RE
3
Mental health
MH
5
For each dimension item scores are coded, summed, and transformed onto a scale from 0 (worst possible health state measured by the questionnaire) to 100 (best possible health state)
5.3
Definition and Measurement of Quality of Life
The definition and measurement of quality of life has been an issue of considerable debate (Eiser and Morse, 2001). It is a term often used vaguely and without clear definition. This is due to the broad nature of a concept which incorporates physical functioning (can undertake activities of daily living including self-care and mobilizing), psychological functioning (emotional and mental wellbeing), social functioning (relationships with others and ability to take part in social activities), and perception of health status, pain, and general satisfaction with life (Sanders et al., 1998). Eiser and Morse (2001) outlined three key ideas which define the concept of quality of life. The first is the idea that individuals have their own distinct perspective on quality of life, depending on current lifestyle, past experience, hopes for the future, dreams, and goals. The second relates to a medical setting, where quality of life is usually viewed as a multidimensional construct incorporating numerous domains. The third idea is that quality of life can include both objective and subjective perspectives in each domain. The former
Intrauterine Growth Restriction and Later Quality of Life
140
perspective concentrates on what the individual can do and the latter includes what things mean to the individual. More recently in all areas of medicine there is increasing recognition of the need to measure outcomes from the perspective of patients. To do so, it is essential to understand the personal experience and effect of disease and disabilities on individuals and their families (Saigal and Rosenbaum, 1996). The measurement of health-related quality of life provides useful descriptive and discriminative information about the various health problems of individuals. Attention to quality of life has emphasized the need to consider outcomes in terms of the individual as a whole, rather than focusing on a narrow range of clinical indicators. Mortality is no longer viewed as the only endpoint when considering the efficacy of medical intervention. Health related quality of life is now considered an important outcome measure for healthcare interventions in adults and health status measures are key determinants of health service use (Hofman et al., 2004).
5.4
1950’s Cohort
The cohort in our study consisted of babies who were born in Royal Maternity Hospital, Belfast between 1954 and 1956 and who were traced and assessed in adulthood at the age of about 50 years. Information on each birth between 1954 and 1956 (n = 6366) was manually abstracted and entered on a database. Gestational age was calculated based on the first day of the last menstrual period (LMP) (Hypponen et al., 2003; Kiserud and Marsal, 2000; Strauss, 2000) and birth weight recorded in pounds and ounces was converted to grams. The study group consisted of growth restricted babies (birth weight < 10th centile), born at term. The control group was term, normal birth weight babies (10th centile). Exclusion criteria were multiple pregnancies, babies born with major congenital abnormalities, and surviving adults deemed inappropriate for the study by their General Practitioner (GP). The study and control groups were selected from 4,667 births that met the inclusion criteria. Software from the Child Growth Foundation was used to adjust birth weight for gestation and gender and to convert these measures to standard deviation scores (SDS). SPSS (version 11) was used to identify the study group (n = 491). A random selection of the nonstudy group based on a one-to-one ratio formed the control group. Losses to follow up, dropouts, and nonparticipation were recorded at the various stages to enable the researcher to examine potential sample bias. The sample was traced with the assistance of the Central Services Agency in Northern Ireland and GPs. For each dimension item scores were coded, summed, and transformed onto a scale from 0 to 100. Mean scores between groups were initially compared using t-tests. Multiple linear regression analysis was then undertaken to adjust for potential confounding variables.
5.5
Health-Related Quality of Life
Results from this study showed both groups reported similar health-related quality of life on each dimension of the SF-36 and there were no significant differences between them (Spence et al., 2007). Adjusting for potential confounding variables did not alter this conclusion (> Tables 140-6–140-9). The IUGR group did have a tendency to use Health Services more for example outpatient departments, accident and emergency, and reported having more GP consultations, illnesses, use of medication and admissions to hospital. This clearly has implications for Health Service resources.
2405
2406
140
Intrauterine Growth Restriction and Later Quality of Life
. Table 140-6 Short Form 36 health survey sub-dimensions related to physical and social functioning before and after adjustment for confounding variables
SF-36
IUGR Group n = 111 Mean (Standard Deviation)
Mean Control Difference Group n = 124 Mean Difference Adjusteda Mean Unadjusted (95% (Standard (95% Confidence Confidence Deviation) Intervals) P Value Intervals) P Value
Physical function
86.3 (21.5)
82.9 (23.7)
Social functioning
85.8 (22.9)
86.0 (23.8)
3.4 ( 2.5 to 9.2) 0.2 ( 6.3 to 5.8)
0.26
4.4 ( 1.3 to 10.1)
0.13
0.94
1.1 ( 4.7 to 6.8)
0.71
a
Adjusted for gender, social class at birth, marital status, education, Townsend deprivation index, and age at time of study The group with intrauterine growth restriction (IUGR) reported similar health-related quality of life on each of these dimensions of the Short Form 36 Health Survey (SF-36). Data are mean and standard deviation, with the scale from 0 (worst possible health state measured by the questionnaire) to 100 (best possible health state)
. Table 140-7 Short Form 36 health survey (SF-36) sub-dimensions related to role limitations before and after adjustment for confounding variables
SF-36
a
IUGR Group n = 111 Control Group Mean Difference Mean n = 124 Mean Unadjusted (Standard (Standard (95% Confidence Deviation) Deviation) Intervals) P Value
Role limitation due to emotional problems
79.9 (34.9)
81.7 (35.1)
Role limitation due to physical problems
83.3 (32.7)
80.6 (35.9)
1.8 ( 10.9 to 7.2)
2.7 ( 6.2 to 11.6)
Mean Difference Adjusteda (95% Confidence Intervals)
P Value
0.69
0.2 ( 8.3 to 8.7)
0.96
0.55
4.8 ( 3.8 to 13.3)
0.27
Adjusted for gender, social class at birth, marital status, education, Townsend deprivation index, and age at time of study The IUGR group reported similar health-related quality of life on each of these dimensions of the SF-36. Data are mean and standard deviation, with the scale from 0 (worst possible health state measured by the questionnaire) to 100 (best possible health state)
Intrauterine Growth Restriction and Later Quality of Life
140
. Table 140-8 Short Form 36 health survey (SF-36) sub-dimensions related to mental health and energy level before and after adjustment for confounding variables
SF-36
IUGR Group Control Group n = 111 Mean n = 124 Mean (Standard (Standard Deviation) Deviation)
Mental Health
72.3 (19.3)
73.4 (19.7)
Energy/ Vitality
61.1 (21.7)
61.0 (21.1)
Mean Difference Unadjusted (95% Confidence Intervals) 1.1 ( 6.1 to 3.9) 0.1 ( 5.4 to 5.6)
Mean Difference Adjusteda (95% Confidence P Value P Value Intervals) 0.67
0.4 ( 4.4 to 5.1)
0.88
0.98
1.5 ( 3.8 to 6.8)
0.58
a
Adjusted for gender, social class at birth, marital status, education, Townsend deprivation index, and age at time of study The IUGR group reported similar health-related quality of life on each of these dimensions of the SF-36. Data are mean and standard deviation, with the scale from 0 (worst possible health state measured by the questionnaire) to 100 (best possible health state)
. Table 140-9 Short Form 36 health survey (SF-36) sub-dimensions related to bodily pain and general health before and after adjustment for confounding variables IUGR Group Control n = 111 Group Mean n = 124 Mean (Standard (Standard Deviation) Deviation)
SF-36 Pain
76.9 (25.6)
76.8 (24.1)
General health perception
70.1 (21.6)
71.0 (20.1)
Mean Difference Unadjusted (95% Confidence Intervals) 0.1 ( 6.3 to 6.5) 0.9 ( 6.3 to 4.4)
P Value
Mean Difference Adjusteda (95% Confidence Intervals)
P Value
0.98
1.3 ( 4.9 to 7.6)
0.68
0.74
0.3 ( 5.0 to 5.6)
0.90
a Adjusted for gender, social class at birth, marital status, education, Townsend deprivation index, and age at time of study The IUGR group reported similar health-related quality of life on each of these dimensions of the SF-36. Data are mean and standard deviation, with the scale from 0 (worst possible health state measured by the questionnaire) to 100 (best possible health state)
6
Conclusion
Most follow-up studies of those born preterm or very low-birth weight examine quality of life in childhood or adolescence, with only a few extending to early adulthood. Less is known about health-related quality of life in later adult life and there is little information on those born with IUGR. Health-related quality of life is regarded as an increasingly important outcome measure for healthcare interventions and health status measures are significant
2407
2408
140
Intrauterine Growth Restriction and Later Quality of Life
determinants of health care utilization (Hofman et al., 2004). The SF-36 is used worldwide to assess quality of life and general health and covers a wide-range of areas that may be adversely affected by illness. In our study the scores on each domain of the SF-36 were similar to published UK age-related norms for males and females aged 45–54 years (Ware et al., 1993). Similar findings have been reported in a study using the SF-36 to assess quality of life in 19–22 year olds who had been born very preterm (Cooke, 2004). Our study indicates that adults who were born growth restricted do not perceive themselves to have worse health-related quality of life than their normally grown peers. It could be argued that participants in this study have survived the short-term effects of IUGR and although they see themselves as being healthy, this needs to be assessed by formal physical examination. The impact of being born with IUGR should not be underestimated in terms of later health and wellbeing and potential implications for Health Service resources.
Summary Points IUGR is a concept indicating a fetus has not reached its optimal growth. IUGR increases mortality and morbidity in the neonatal period but there is limited information on its effect on health-related quality of life in adulthood.
The importance of measuring health-related quality of life is recognized by clinicians and policy makers to inform patient management and policy decisions.
Our study indicates that adults who were born growth restricted do not perceive themselves to have worse health-related quality of life than their normally grown peers.
Being born with IUGR has potential implications for Health Service resources.
References Ariyuki Y, Hata T, Kitao M. (1995). Pediatrics. 96: 36–42. Bakketeig LS. (1998). Eur J Clin Nutr. 52: Suppl 1. Barker DJP (ed.). (1992). Fetal and Infant Origins of Adult Disease. BMJ Publishing Group, London. Clausson B, Gardosi J, Francis A, Cnattingius S. (2001). Br J Obstet Gynaecol. 108: 830–834. Cooke RWI. (2004). Arch Dis Child. 89: 201–206. Crosse VM. (1949). The premature baby, 2nd ed. J & A Churchill Ltd, London. Drossou V, Diamanti E, Noutsia H, Konstantinidis T, Katsougiannopoulos V. (1995). Acta Paediatr. 84: 1–5. Eiser C, Morse R. (2001). Arch Dis Child. 84: 205–211. Ergaz M, Avgil M, Ornoy A. (2005). Reprod Toxic. 20: 301–322. Fanaroff AA, Hack M, Kliegman RM. (1989). In: Gross TL, Sokol RJ (eds.) Intrauterine Growth Restriction a Practical Approach. Year Book Medical Publishers Inc, London.
Garratt AM, Ruta DA, Abdalla MI, Buckingham JK, Russell IT. (1993). Br Med J. 306: 1440–1444. Gross TL. (1989). In: Gross TL, Sokol RJ (eds.) Intrauterine Growth Restriction a Practical Approach. Year Book Medical Publishers Inc, London. Hack M, Taylor G, Klein N, Eiben R, Schatschneider C, Mercuri-Minich N. (1994). N Engl J Med. 331: 753–759. Hayes V, Morris J, Wolfe C, Morgan M. (1995). Age Ageing. 24: 120–125. Henriksen T. (1999). Acta Paediatr Suppl. 88: 4–8. Hofman A, Jaddoe VWV, MacKenbach JP, Moll HA, Snijders RFM, Steegers EAP, Verhulst FC, Witteman JCM, Buller HA. (2004). Paediatr Perinat Epidemiol. 18: 61–72. Hyppo¨nen E, Power C, Davey Smith G. (2003). Diabetes Care. 26: 2515–2517. Jenkinson C, Coulter A, Wright L. (1993). Br Med J. 306: 1437–1440.
Intrauterine Growth Restriction and Later Quality of Life Kilby M, Hodgett S. (2000). In: Kingdom J, Baker P (eds.) Intrauterine Growth Restriction a Practical Approach. Year Book Medical Publishers Inc, London. Kingdom J, Baker P (eds.). Intrauterine Growth Restriction Aetiology and Management. Springer-Verlag, London. Kiserud T, Marsˇa´l K. (2000). In: Kingdom J, Baker P (eds.) Intrauterine Growth Restriction a Practical Approach. Year Book Medical Publishers Inc, London. McCormick MC, Brooks-Gunn J, Workman-Daniels K, Turner J, Peckham GJ. (1992). JAMA. 267: 2204–2208. Neerhof MG. (1995). Clin Perinatol. 22: 375–385. Paz I, Gale R, Laor A, Danon YL, Stevenson DK, Seidman DS. (1995). Obstet Gynecol. 85: 452–456. Robinson JS, Owens JA, McMillen IC, Erwich JJ, Owens PC. (1997). In: Cockburn F (ed.) Advances in Perinatal Medical. The Parthenon Publishing Group, London.
140
Saigal S, Rosenbaum P. (1996). Semin Neonatol. 1: 305–312. Saigal S, Szatmari P, Rosenbaum P, King S, Campbell D. (1991). J Pediatr. 118: 751–760. Sanders C, Egger M, Donovan J, Tallon D, Frankel S. (1998). Br Med J. 317: 1191–1194. Spence D, Alderdice FA, Stewart MC, Halliday HL, Bell AH. (2007). Arch Dis Child. 92: 700–703. Strauss RS. (2000). JAMA. 283: 625–632. Vandenbosche RC, Kirchner DO. (1998). Am Fam Physician. 15: 1384–1393. Ware JE, Snow KK, Kosinski M, Gandek B. (1993). The Health Institute, New England Medical Center, Boston, MA. Whittle MJ. (1999). In: Rodeck CH, Whittle MJ (ed.) Fetal Medicine Basic Science and Clinical Practice. Churchill Livingstone, London. Yanney M, Marlow N. (2004). Semin Fetal Neonatal Med. 9: 411–418.
2409
141 Assessment of Quality of Life During Pregnancy and in the Postnatal Period C. R. Martin . J. Jomeen 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2412 2 Quality of Life in Pregnancy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2412 3 Postnatal Quality of Life . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2414 4 Measuring Perinatal Quality of Life . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2416 5 SF-36 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2416 6 The Mother Generated Index (MGI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2416 7 Maternal Postpartum Quality of Life Tool . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2418 8 Inventory of Functional Status after Childbirth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2420 9 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2420 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2420
#
Springer Science+Business Media LLC 2010 (USA)
2412
141
Assessment of Quality of Life During Pregnancy and in the Postnatal Period
Abstract: Quality of life has become an area of increasing importance to the area of pregnancy and childbirth. Much has been written on specific clinical presentations that occur during the > perinatal period that may deleteriously impact on perceived quality of life, for example > postnatal depression, complications during labor or abnormalities in the baby. This chapter will explore the salient issues regarding assessment of quality of life during pregnancy and in the postnatal period. List of Abbreviations: EPDS, Edinburgh Postnatal Depression Scale; IFSAC, The Inventory of Functional status after Childbirth; MAPP-QOL, The Maternal Postpartum Quality of Life Tool; MGI, Mother Generated Index; SF-36, Short-Form 36
1
Introduction
Pregnancy induces a unique physiological response, which stresses the body more than any other physiological event in a health women’s life resulting in metabolic, hormonal, cardiovascular, respiratory and musculo-skeletal adaptations (Sternfield, 1997). Many of theses changes manifest in many of the so called minor disorders of pregnancy such as nausea and vomiting, dizziness and fatigue (Gross and Pattison, 2007), some of which are transient some are more enduring. Following birth, the postnatal period is a time when a woman has to adjust to physical changes and new emotional demands and the impact of aspects such as perineal trauma and associated pain on women’s general health (Albers and Borders, 2007), ability to fulfill daily activities (Lydon-Rochelle et al., 2001) and ability to adapt to motherhood (Mason et al., 1999) has also been investigated and documented. The increasing literature, documenting the impact of pregnancy and childbirth on women’s normal daily activities, identifies the concept of Quality of Life as an important psychological domain worthy of further debate, both as a concept but also in terms of its effective assessment in women across the perinatal period.
2
Quality of Life in Pregnancy
There is a general paucity of evidence exploring the relationship between perinatality and women’s quality of life. Decreased physical functioning, as a dimension of quality of life, however has been demonstrated during normal pregnancy (Haas et al., 2005; Heuston and Kasik-Miller, 1998; Ochet et al., 1999). In pregnancy, nausea and vomiting, is experienced by more than 70% of pregnant women with 28% reporting the negative consequences of pregnancy on family, social and occupational functioning (Attard et al., 2002; O’Brien and Naber, 1992). However, broader quality of life limitations have also been identified by women themselves, including physical symptoms/aggravating factors; fatigue and emotions (Magee et al., 2002). In a study conducted at the end of pregnancy McKee et al. (2001) emphasized the impairment of pregnancy on the dimensions of life quality that require physical activity and in particular the detriment to ‘‘vitality’’ dimension of quality of life. More recently Forger et al. (2005) noted increased bodily pain and impaired physical functioning in a cohort of Austrian women during late pregnancy. There is very little work considering pregnant groups and physical activity in relation to clinical outcomes, indeed the association between quality of life decrements and pregnancy outcomes remains equivocal. Whilst poor physical functioning
Assessment of Quality of Life During Pregnancy and in the Postnatal Period
141
has been significantly associated with an increased risk of > preterm labor (Haas et al., 1999), other studies conversely, have linked increased or strenuous domestic activity in > multigravida women with higher rates of preterm labor (Launer et al., 1990) and small for dates babies (Woo, 1997). Some albeit rather scant evidence suggests that the restriction of physical activity is associated more with psychological rather than clinical outcomes, with implications for women’s self-esteem and a sense of control associated with being active and the freedom to decide what activities to pursue (Gross and Pattison, 2007). Clearly pregnancy related factors which impact on daily life are important to pregnant women. It is also feasible that a woman’s perception of their pregnancy is important in assessments of life quality. Pesavento et al. (2005), investigating quality of life in pregnant Italian women, found that those with a normal pregnancy have a good perception of their quality of life, whilst women with a high risk pregnancy do not. This may link to the findings, which identify the unwelcome nature of the restrictions associated with a high risk pregnancy and suggest that any potential reduction in psychological wellbeing can be attributed more to the restrictions placed on activity than to the pregnancy risk itself (Mackey and Coster-Schultz, 1992; Monaham and De Joseph, 1991). A further dimension to the debate around quality of life in pregnancy is explored by Kelly and colleagues (2001). They demonstrate the relationship between amplified physical symptoms in pregnancy and the existence of depression and anxiety. This concurs with findings from primary and secondary care settings, which have shown that unexplained medical symptoms are associated with psychopathology (Russo et al., 1994; Simon et al., 1999). Further, anxiety and depression amongst patients with a known medical disease are associated with an amplification of the disease specific and non-specific symptoms (Dwight et al., 2000). Physical symptoms are common in pregnancy, predominantly associated with the normal physiological changes that occur and expected by both clinicians and women themselves. However, an elevated incidence of somatic complaints or negative reports/perceptions of quality of life may rather be associated with psychological disturbance. This may also be true in reverse, and the amplification of somatic symptoms may contribute to the existence of anxiety and depression. The direction of this relationship is yet to be established but it seems legitimate, to suggest that quality of life detriment has often been perceived as a normative part of pregnancy and as such largely ignored. Abundant evidence demonstrates a strong association between depression and decrements in self-reported functional status or quality of life (Simon, 2003) and also that effective treatment helps to restore function. Other studies have reported that outpatients with depressive disorders experienced functional impairment and decreased well-being comparable to or greater than that of people with chronic medical conditions (Hays et al., 1995; Wells et al., 1989). Other trials have demonstrated that improving depression leads to significant improvements in quality of life (Coulehan et al., 1997). Only a few studies to date have considered these relationships in pregnant women. An inverse relationship has been demonstrated between elevated levels of depressive symptomatology and lowered health-related functioning and perceived well-being in Black and Hispanic pregnant women (McKee et al., 2001) and in a more diverse group of pregnant American women (Nicholson et al., 2006). More recent study findings also identified that British women in early pregnancy classified as possibly anxious using the Anxiety subscale of the Hospital Anxiety and Depression Scale (Zigmond and Snaith, 1983) or suffering minor/major depression as measured by the Edinburgh Postnatal Depression Scale (Cox et al., 1987), experienced poorer quality of life compared with their non-anxious and non-depressed counterparts (Jomeen and Martin, 2005).
2413
2414
141
Assessment of Quality of Life During Pregnancy and in the Postnatal Period
. Figure 141-1 Factors which may impact on antenatal quality of life. It shows the constellation of factors which may significantly impact on antenatal quality of life. This list is not exhaustive and the factors described may interact. Baby screen results in the figure explicitly means tests for the detection of fetal abnormality
Once again, however, the direction of causality remains ambiguous but it seems clear that psychological detriment and quality of life are inherently linked. Factors which may impact on perceived antenatal quality of life are shown in > Figure 141-1.
3
Postnatal Quality of Life
Women experience a broad spectrum of physical as well as emotional challenges following childbirth; some of these are persistent and clearly have the capacity to impact on a woman’s quality of life. Recognition of the extent of maternal morbidity and its effect on quality of life has been on the increase in recent years (Symon et al., 2002) and as a result has become the focus of several studies. The impact of perineal trauma and associated perineal pain, a common complication of labor, is well documented. Perineal pain has been linked to both short term and long term perinatal morbidity including perineal pain, perineal healing, urinary incontinence, flatus incontinence, fecal incontinence, sexual morbidity and > dyspareunia (Williams et al., 2007), as well as fatigue and depression (Brown and Lumley, 2000). Both the level of perineal trauma and the degree of suturing were significant in women’s selfreports of poor general health (Eason et al., 2000). Women with an > episiotomy, for example, demonstrated longer periods of disruption to their daily life including sleeplessness, difficulty bathing and resuming normal daily activities (Okubo et al., 2000). Quality of life in the postnatal period is a complex and personal area affected by many different aspects of health and well-being (Symon et al., 2002) and can determined by the physical experience of pregnancy and childbirth itself but also the emotional, social, sexual and spiritual dimensions of the transition to motherhood. From a physical perspective, women with significant obstetric morbidity in pregnancy and/or labor report lower general health (Waterstone et al., 2003). Important differences in fatigue have been observed with emergency and elective
Assessment of Quality of Life During Pregnancy and in the Postnatal Period
141
caesarean section women reporting more prolonged fatigue and lower physical quality of life. A study more specifically examining sleep quality, found positive relationships between good sleep quality and self-report health (Hyyppa et al., 1991) seeming to concur with Tulman and Fawcett (1990) that physical energy levels were overall the strongest correlate of functional status and highlights the dynamic of sleep and fatigue in terms of postnatal quality of life. Indeed in a study which assessed women’s quality of life and sleep quantity/quality at 2 weeks and 6 months postnatal revealed profiles of improving quality of life concurrent with improved sleep patterns (Jomeen and Martin (in press)). Such findings seem intuitive and support the suggestion that functional status increases progressively to reflect the physical recovery and psychosocial adjustment to motherhood. Tulman and Fawcett (1990) reveal the significant relationships between mother’s functional status and variables such as confidence in motherhood, family, relationship and demographic variables. Higher levels of physical energy, parity, confidence in mothering, and infants with predictable temperament were all associated with increased functional status in household, social, community and self-care activities. A study which considered the role of quality of life and its relation to postnatal depressive symptomology in Mexican women, found antenatal family quality of life, and postnatal family quality of life to be significant predictors along with other risk factors to postnatal depression (Martinez-Schallmoser, 1992). As in the antenatal period, studies demonstrate a co-morbid relationship between postnatal depression and negative quality of life assessments (Small et al., 2000) yet once again the causality of the relationship is not firmly established. However a recent study exploring relationships between postnatal depression and quality of life implies that it is postnatal depression that negatively influences all dimensions of life quality explored using the SF-36 (De Tychey et al., 2007). An interesting dimension, which supports this assumption, appears to be the influence of baby gender on postnatal depression and postnatal quality of life, with the birth of a boy having negative consequences. Factors which may impact on perceived postnatal quality of life are shown in > Figure 141-2. . Figure 141-2 Factors which may impact on postnatal quality of life. It shows the myriad of factors which may significantly impact on postnatal quality of life. This list is not exhaustive and the factors described may interact
2415
2416
141 4
Assessment of Quality of Life During Pregnancy and in the Postnatal Period
Measuring Perinatal Quality of Life
An increasing interest in quality of life as a psychological domain of interest and clinical significance requires a quality of life instrument that reliably measure both physical and psychological functioning. The concern around screening measures in pregnancy appears to focus around the unique and dynamic physiological nature of pregnancy. Several assessment and screening tools that have been demonstrated as reliable and valid in generic populations have been observed to be unstable in terms of factor structure, unreliable and inaccurate (Jomeen and Martin, 2004; Karimova and Martin, 2003; Martin and Jomeen, 2003).
5
SF-36
Quality of Life has been widely investigated in a number of clinical conditions utilizing the Medical Outcomes Survey Short Form 36 (SF-36: Ware and Sherbourne, 1992; Ware et al., 2000). This measure is a widely used generic, multipurpose self-report quality of life questionnaire. The SF-36 consists of 36 questions from which functional health and well-being subscale scores are calculated for eight sub-scale domains comprehensively describing quality of life attributes; these being physical functioning (1), role-physical (2), bodily pain (3), general health (4), vitality (5), social functioning (6), role-emotional (7) and mental health (8). Summary measures of physical health (scales 1–4) and mental health (scales 5–8) can also be calculated. An improved version of the SF-36 (version 2) has been introduced which incorporates instruction and item changes and a better layout for questions and answers. The psychometric properties of the SF-36 have been extensively evaluated with good support found for the taxonomy of eight scales and two higher order factors (physical health and mental health domains). The SF-36 has been used in pregnancy (Heuston and Kasik-Miller, 1998) and in postnatal women (De Tychey et al., 2007; Small et al., 2000; Small et al., 2003). Its psychometric properties in early pregnancy have been recently explored by Jomeen and Martin (2005), utilizing Exploratory and Confirmatory Factor Analysis. The recommendations from this study are for the use of the SF-36 as an eight subscale measure. The merits of using the instrument as a two subscale measure of physical and mental health require further evaluation (Jomeen and Martin, 2005). The SF-36 has also demonstrated as a feasible and reliable tool (a > 0.7) in postnatal women, able to discriminate between groups by mode of delivery and to detect moderate recovery in physical and small recovery in mental status over time, when combined with other health related quality of life measures (Jansen et al., 2007). There are a number of suggested measurement models of the SF-36 sub-scales (Jomeen and Martin, 2005) and these are described in > Figures 141-3–141-6.
6
The Mother Generated Index (MGI)
The Mother Generated Index (MGI) is subjective tool designed specifically for use in postnatal women, in order to assess postnatal quality of life in new mothers. The MGI gives a primary index quality of life score and a secondary index that identifies those areas considered most important by the mother. The validity of the MGI was tested in a cohort of 103 postnatal women and was sought from concurrent use of the EPDS and the SF-12 which has both
Assessment of Quality of Life During Pregnancy and in the Postnatal Period
141
. Figure 141-3 Diagrammatic representation of a uni-dimensional model of the SF-36. The figure shows the sub-scales of the SF-36 and their relationship to a single domain of quality of life. This presents a uni-dimensional measurement model of the SF-36
. Figure 141-4 Diagrammatic representation of a bi-dimensional model of the SF-36 specifying separable but correlated physical and mental quality of life domains. The figure shows the sub-scales of the SF-36 and their relationship to two higher order physical health and mental health domains. These higher order domains are often found to be significantly correlated in clinical groups. The Figure reveals one common measurement model of the SF-36, however a number of others have been proposed. Calculation of separate mental health and physical health component scores has been suggested and are often reported in the clinical research literature utilizing this commonly used quality of life assessment tool
physical (PCS-12) and mental component (MCS-12) scores (Symon et al., 2002). The primary index correlated significantly with the EPDS and the MCS-12 yet not with the PCS-12 scores at both, the differences were more marked at the 8 month than 6–8 weeks observation point. The authors postulate that women do not consider physical problems as important
2417
2418
141
Assessment of Quality of Life During Pregnancy and in the Postnatal Period
. Figure 141-5 Diagrammatic representation of a bi-dimensional model of the SF-36 specifying uncorrelated physical and mental components within a higher order health quality of life domain. The figure shows the sub-scales of the SF-36 and their relationship to uncorrelated higher order physical health and mental health domains. These higher order domains may correlate significantly with the overall quality of life score but may not be correlated with each other. Though less established in the literature, this model may be relevant to certain clinical groups with specific symptom presentations, for example a group with very poor physical functioning but relatively normal levels of mental health functioning
in terms of assessing quality of life, a conclusion not necessarily borne out by the other study findings, which have been previously explored in this chapter. Marked differences between high and low scores in terms of quality of life lead the authors to suggest that the measure might serve as a valuable screening tool and a cut-point score of six is also explored. The unique benefit of the MGI, which could also be considered its drawback for use in large scale research studies, is that it allows women to comment on any aspect of their own lives after childbirth and is individualized in its nature. Symon et al. (2002) study was not without limitations, including the small sample utilized and the MGI needs further testing in terms of its reliability and validity. Its use since its introduction appears to have been limited.
7
Maternal Postpartum Quality of Life Tool
A recent study developed and tested a 41 item self-report measure to assess postpartum experience and quality of life of new mothers in the early postnatal period (Hill et al., 2006). The Maternal Postpartum Quality of Life Tool (MAPP-QOL), adopted the definition, domains and conceptual model from the work of Ferrens (1990) and Ferrens and Powers (1992).
Assessment of Quality of Life During Pregnancy and in the Postnatal Period
141
. Figure 141-6 Diagrammatic representation of a tri-dimensional model of the SF-36 specifying separable physical, mental and general well-being components within a higher order health quality of life domain. The figure shows the sub-scales of the SF-36 and their relationship to physical, mental and general well-being quality of life higher order domains. This represents a sophisticated conceptualization of a unique measurement model of the SF-36, informed by clinical research observations. These more sophisticated measurement models are of importance as they provide useful insights into the responding characteristics of distinct clinical groups on the instrument which may have unique responding ‘‘signatures’’ or profiles. This model also highlights the developing area of clinical interest in the measurement characteristics of commonly used quality of life assessment tools
The Quality of Life Index (Ferrens, 1990) has core items and condition specific items and has demonstrated good internal consistency across various clinical groups. The bulk of the items in the MAPP-QOL replicate those of the QLI, with items modified or developed to fit specific aspects of quality of life domains for postpartum mothers. Component analysis revealed to scale to consist of five domains linked to postnatal quality of life including psychological/baby; socio-economic; relational/spouse-partner; relational/family-friends; health and functioning. The measure performed relatively well from a psychometric perspective, with alpha co-efficients across the domains ranging from 0.82–0.96, convergent validity was supported by a strong correlation (r ¼ 0.69), whilst discriminate validity was supported by negative correlations with the four negative mood states of the Multiple Affect Adjective Check List Revised (anxiety; depression; hostility and dysphoria). Hill et al. (2006) study was not without limitations in terms of sample size and type and the utility of this scale clearly requires further evaluation. Despite it relatively good performance, this measure remains in its infancy.
2419
2420
141 8
Assessment of Quality of Life During Pregnancy and in the Postnatal Period
Inventory of Functional Status after Childbirth
The Inventory of Functional status after Childbirth (IFSAC: Fawcett et al., 1988) was designed for assessment of the functional status in the specific situation of recovery from childbirth. With functional status in this context defined as the mother’s readiness to assume infant care responsibilities and resume self-care, household, social, community and occupational activities. The model was conceptually derived from and an attempt to operationalized Roy’s Adaptation model. The IFSAC is a 36 item measure arranged into five subscales, reflecting the definition above infant care responsibilities (six items); self care activities (eight items); household activities (12 items); social and community activities (six items) and occupational activities (four items). Items are scored on a four point scale ranging from ‘‘never’’ to ‘‘all the time,’’ high mean scores for individual subscales and total scale indicated greater functional status. Initial psychometric testing revealed modest but acceptable levels of internal consistency, poor construct validity and correlations between subscales which suggested the need for further testing and refinement.
9
Conclusion
Pregnancy and childbirth are associated with intense physical changes and often a great deal of emotional upheaval, with the ability to perform usual roles affected (Attard et al., 2002). Even in an uneventful pregnancy women have subtle changes that may detract from their quality of life (Heuston and Kasik-Miller, 1998). It seems apparent that quality of life may have a significant role to play in the psychological well-being of pregnant and postnatal women, with a possible suggestion that recognition and validation by caregivers of the need for pregnant women to make changes in lifestyle will contribute to improved quality of life and less risk of psychological sequalae. However evidence regarding outcomes related to quality of life and maternity experience remains scant and merits further investigation before any clear associations can be made between quality of life issues and pregnant and postnatal women’s psychological well-being. Further research must also concentrate on the utility of the measures available to measure quality of life and contribute to the debate on reliable instruments to identify and/or predict physical detriment and psychological sequelae.
Summary Points Pregnancy is associated with physiological and psychological changes which may impact in quality of life.
The postnatal period is also associated with specific presentations such as postnatal depression which may impact negatively on quality of life.
There is a relative vacuum in the research literature on the relationship of pregnancy and the postnatal period to quality of life.
Depression can occur during pregnancy and have a negative impact on quality of life. The physical consequences of childbirth including birth trauma may impact negatively on perceived quality of life.
The presence of a relatively low perceived quality of life during the antenatal period may predict later postnatal depression.
Assessment of Quality of Life During Pregnancy and in the Postnatal Period
141
Gender of the baby may be an important contributor to the mothers perceived quality of life, with a male baby having a more negative effect in this domain.
Assessment of key psychological domains germain to quality of life is difficult during pregnancy and the postnatal period with evidence of a number of significant psychometric concerns related to measurement. The evidence base in relation to quality of life during pregnancy and in the postnatal period is developing rapidly but much further research is needed in this area.
References Albers LL, Borders N. (2007). J Midwifery Womens Health. 52: 246–253. Attard CL, Kohli MA, Coleman S, Bradley C, Hux M, Atanackovic G, Torrance GW. (2002). Am J Obstet Gynecol. 186: 222–227. Brown S, Lumley J. (2000). Br J Obstet Gynaecol. 107: 1194–2001. Coulehan J, Schulberg H, Block M, Madonia M, Rodriguez E. (1997). Arch Intern Med. 157: 1113–1120. Cox JL, Holden J, Sagovsky R. (1987). Br J Psychiatry. 150: 782–786. de Tychey C, Briancon S, Lighezzolo J, Spitz E, Kabuth B, de Luigi V, Messmebourg C, Girvan F, Rosati A, Thockler A, Vincent S. (2007). J Clin Nurs (Online Early Articles). Available from http://www. blackwell-synergy.com/doi/abs/10.1111/j.1365–2702. 2006.01911.x. Eason E, Labrecque M, Feldman P. (2000). Obstet Gynecol. 95: 464–471. Dwight M, Kowdley K, Rus J, Ciechanowski P. (2000). J Psychosom Res 49: 311–317. Fawcett J, Tulman L, Myers ST. (1988). J Nurse Midwifery. 33(6): 252–260. Ferrens C. (1990). Oncol Nurs Forum. 17(3): 15–21. Ferrens C, Powers M. (1992). Res Nurs Health. 15: 29–38. Fo¨rger F, Østensen M, Schumacher A, Villiger PM. (2005). Ann Rheum Dis. 64: 1494–1499. Gerard Jansen AJ, Essink-Bot M-L, Duvekot JJ, van Rhenen DJ. (2007). J Psychosom Res. 63: 275–281. Gross H, Pattison H. (2007). Sanctioning Pregnancy. Routledge, Hove. Haas JS, Jackson RA, Fuentes-Afflick E, Stewart AL, Dean ML, Brawarsky P, Escobar GJ. (2005). J Gen Intern Med. 20: 45–51. Haas J, Meneses V, McCormick M. (1999). J Womens Health Gend Based Med. 8: 547–553. Hays R, Wells K, Sherbourne C, Rogers W, Spritzer K. (1995). Arch Gen Psychiatry. 53: 899–904. Heuston W, Kasik-Miller S. (1998). J Fam Pract. 47: 209–212.
Hill PD, Aldag JC. (2007). J Gynecol Neonatal Nurs. 36(4): 328–334. Hill PD, Aldag JC, Hekel B, Riner G, Bloomfield P. (2006). J Nurs Meas. 14(3): 205–220. Hyyppa MT, Kronholm E, Mattler CE. (1991). Br J Med Psychol. 64: 24–34. Jansen G, Essink-Bot M-L, Duvekot JJ, van Rhenen DJ. (2007). J Psychosom Res. 63(3): 275–281. Jomeen J, Martin CR. (2004). Psychol Health. 19: 787–800. Jomeen J, Martin CR. (2005). J Psychosom Res. 59: 131–138. Jomeen J, Martin CR. (in press). J Eval Clin Pract. Karimova GK, Martin CR. (2003). Psychol Health Med. 8(1): 89–103. Kelly RH, Russo J, Katon W. (2001). Gen Hosp Psychiatry. 23(3): 107–113. Launer LJ, Villar J, Kestler E, de Onis M. (1990). Br J Obstet Gynaecol. 97: 62–70. Lydon-Rochelle MT, Holt VL, Martin DP. (2001). Paediatr Perinat Epidemiol. 15(3): 232–240. Magee LA, Chandra K, Mazzotta P, Stewart D, Koren G, Guyatt, G. (2002). Am J Obstet Gynecol. 186(5): 232–238. Mackey MC, Coster-Schultz MA. (1992). Clin Nurs Res. 1: 366–284. Martin CR, Jomeen J. (2003). J Reprod Infant Psychol. 21: 267–278. Martinez-Schallmoser LD. (1992). Perinatal depressive symptoms, quality of life, social support and risk factors in Mexican-American women. PH.D., University of Illinois. Mason L, Glenn S, Walton I, Appleton C. (1999). Birth. 26(3): 164–171. McKee MD, Cunningham M, Janklowski MA, Zayas L. (2001). Obstet Gynecol. 97: 998–993. Monaham PA, De Joseph JF. (1991). J Perinat Neonat Nurs. 4: 12–21. Nicholson WK, Setse R, Hill-Briggs F, Cooper LA, Strobino D, Powe NR. (2006). Obstet Gynecol. 107: 798–806.
2421
2422
141
Assessment of Quality of Life During Pregnancy and in the Postnatal Period
O’Brien B, Naber S. (1992). Birth. 19: 138–143. Ochet F, Carey MA, Adam L. (1999). Obstet Gynecol. 94: 935–941. Okubo N, Mitsuhashi Y, Saito K. (2000). 14: 35–44. Pesavento F, Marconcini E, Drago D. (2005). Minerva Ginecol. 57: 451–460. Russo J, Katon M, Sullivan M, Clark M, Buchwald D. (1994). Psychosomatics. 35: 546–556. Simon GE. (2003). Biol Psychiatry. 54: 208–215. Simon G, Von Korff M, Piccinelli C, Fullerton C, Ormel J. (1999). N Engl J Med. 341: 1329–1335. Small R, Lumley J, Donohue L, Potter A, Waldenstrom U. (2000). Br Med J. 321: 1043–1047. Small R, Lumley J, Yelland J. (2003). J Psychosom Obstet Gynaecol. 24: 45–52. Sternfield B. (1997). Sports Med. 23: 33–47. Symon A, MacDonald A, Ruta D. (2002). Birth. 29: 40–46.
Tulman L, Fawcett J. (1990). 22(3): 191–194. Ware Jr JE, Sherbourne CD. (1992). Med Care. 30: 473–483. Ware JE, Kosinski MA, Dewey JE. (2000). How to Score Version 2 of the SF-36 Health Survey. Quality Metric: Lincoln (RI) Waterstone M, Wolfe C, Hooper R, Bewley S. (2003). Br J Obstet Gynaecol. 110: 128–133. Wells K, Stewart R, Hays M, Burnam W, Rogers W, Daniels M, Berry S, Greenfield S, Ware J. (1989). JAMA. 283: 212–230. Williams A, Herron-Marx S, Knibb R. (2007). J Clin Nurs. 16, 549–561. Woo GM. (1997). Ann Behav Med. 19: 385–398. Zigmond AR, Snaith RP. (1983). Acta Psychiatr Scand. 67: 361–370.
142 Generic Quality of Life Measures for Children and Adolescents K. J. Zullig . M. R. Matthews . R. Gilman . R. F. Valois . E. S. Huebner 1 1.1 1.2
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2424 Relevance of Quality of Life Measurement to Health Care . . . . . . . . . . . . . . . . . . . . . . . 2424 The Challenge of Quality of Life Measurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2426
2 2.1
Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2426 Generic Measures Including both Objective and Subjective Aspects of QOL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2427 Comprehensive Quality of Life Scale-School Version . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2427 Quality of Life Profile: Adolescent Version . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2428 Generic Subjective QOL Measures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2429 Youth Quality of Life-Research Version . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2429 Multidimensional Students’ Life Satisfaction Scale . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2431
2.1.1 2.1.2 2.2 2.2.1 2.2.2 3 3.1
Future Directions for Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2433 Quality of Life Measurement: How many Domains can and Should be Measured? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2434
4
National Adolescent Indexes to Match Objective and Subjective QOL Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2435 Limitations of the CWI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2436 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2436
4.1 4.2
Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2436 Appendix A . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2438
#
Springer Science+Business Media LLC 2010 (USA)
2424
142
Generic Quality of Life Measures for Children and Adolescents
Abstract: The purpose of this chapter was to review generic measures of child and adolescent (ages 18 or less) quality of life (QOL) that have been applied in at least two populations (e.g., age groups, nationalities, etc.) and have yielded strong psychometric properties. The chapter begins with a discussion regarding the relevance of quality of life measurement to health care. Next, we briefly review the challenges associated with QOL measurement in general followed by multidimensional measures that have been applied to various general (i.e., non-clinical) samples. Multidimensional measures were chosen for this review because they gather important information across a number of life domains. For example, understanding the contribution of environmental factors to an adolescent’s behavioral, cognitive, and overall health functioning is critical to understanding human development In this regard, multidimensional QOL measures can help assess the potential influence of various proximal (e.g., quality of school experiences) and distal factors (e.g., quality of neighborhood environment) on life quality among youth. The measures in this review are classified into two categories: (1) those that contain both objective and subjective QOL indicators, and (2) those comprised of subjective QOL measures only. Specifically, we review the Comprehensive Quality of Life Scale-School Version (ComQOL-S5: Cummins, 1997), the Quality of Life Profile-Adolescent Version (QOLPAV: Raphael et al., 1996), the Youth Quality of Life (YQOL: Edwards et al., 2002; Patrick et al., 2002), and the Multidimensional Students’ Life Satisfaction Scale (MSLSS: Huebner, 1994). Lastly, we recommend some suggestions for future research in youth QOL measurement, including promising efforts to use QOL as one overall index of well-being. List of Abbreviations: BASC 2, behavior assessment system for children-second edition; CWI, child and youth well-being index; CHIP-AE, child health and illness profile; CHQCF80, child health questionnaire-child form; CHQ-PF50, child health questionnaire-parent form; CDI, children’s depression inventory; ComQOL-S5, comprehensive quality of life scaleschool version; CADS-A, conners’ auxiliary adhd/dsm iv instrument; FDI, functional disability inventory; GCQ, generic children’s quality of life measure; KINDL, German Munich quality of life questionnaire for children; MSLSS, multidimensional students’ life satisfaction scale; PedsQL, pediatric quality of life inventory; PWI-SC, personal well-being index-school children; QOL, quality of life; QOLPAV, quality of life profile-adolescent version; SF-36, short form; SF-12, short form-12 health survey; SIP, sickness impact profile; TedQL, teddy quality of life; C-QOL, the quality of life measure for children; YQOL, youth quality of life
1
Introduction
1.1
Relevance of Quality of Life Measurement to Health Care
A number of recent studies consistently note how increasing levels of life quality appear to protect against factors that may potentially contribute to psychological and interpersonal distress. For example, Gilman and Huebner (2006) found that adolescents reporting exceedingly high overall QOL (i.e., scores in the top 20% of the distribution) were significantly less likely to report elevated levels of psychopathology (e.g., depression, anxiety), and to report significantly more positive interpersonal relationships and school attitudes than adolescents reporting exceedingly low QOL (i.e., scores in the bottom 20% of the distribution). Equally if not more important, adolescents in the high QOL group also reported greater adaptation than peers who reported ‘‘average’’ QOL (scores in the middle of the distribution), suggesting that optimal levels of QOL
Generic Quality of Life Measures for Children and Adolescents
142
confer benefits that are not found among even among youth reporting typical levels (see Greenspoon and Saklofske, 2001 for further support). Similar findings have been reported in other studies-both with respect to self-reported overall QOL (Suldo and Huebner, 2006) and to specific QOL domains (e.g., school experiences; see Huebner and Gilman, 2006). Given the findings that suggest that high QOL and various indices of positive behavioral, interpersonal, and educational adjustment are interrelated, the application of QOL measurement to healthcare professions is important for a number of reasons. First, the use of QOL ratings may provide important information above and beyond that what is traditionally measured in health care settings-typically symptom presence and severity (see Frisch, 1998). Nevertheless, anecdotal (Frisch et al., 1992) and empirical evidence (Keyes, 2007) finds that successful elimination of maladaptive behaviors does not necessarily lead to optimal life quality. That is, one cannot assume that individuals will behave or feel ‘‘good’’ simply because they no longer feel ‘‘bad’’. Although these findings and comments were specific to adults, similar statements have been made among adolescent researchers as well (Huebner et al., 2006). Second, although most adolescents are healthy and report being in good health (Irwin et al., 2002), the assumption that physical health difficulties may go unrecognized among adolescents until an actual serious physical illness manifests itself is problematic. For example, Zullig et al. (2005) found that adolescents who reported just 1–2 poor physical health days during the past 30 were significantly more likely to report dissatisfaction with life. In this same study with dissatisfaction with life increasing linearly as number of reported poor days increased. Studies with adults have revealed associations between life satisfaction reports and a variety of health conditions (see Frisch et al., 2003 for a review). For instance, low life satisfaction reports are associated with physical health problems, including myocardial infarctions, chronic pain syndrome, respiratory tract infections, and colds. Low life satisfaction has been shown to predict mental health problems, such as depression, anxiety, and somatoform disorders as well as suicide. Individual differences in life satisfaction also significantly predict longevity (Frisch et al., 2003). Thus, youth reporting a greater number of poor physical health days during the past 30 days may be at greater risk for any of these conditions as they age. Third, QOL ratings may help monitor how adolescents’ in health care settings perceive the quality of their treatment, which is in turn a factor contributing to compliance. There is an axiom that what gets measured gets done (Moore et al., 2003). Yet often what is monitored is outcome data based solely on measures of psychopathology or objective indicators (e.g., frequency of disruptive behaviors observed). However, there is an association between intervention compliance and positive perceptions of the intervention itself. For example, among adults Carlson and Gabriel (2001) reported that patients in a drug treatment program who rated the quality of services as high were more likely to abstain from drugs even 1 year after receiving treatment. Among adolescents, QOL reports appear to show treatment sensitivity and compliance. For instance, longitudinal studies of adolescents first entering a residential treatment setting reported different (and consistently higher) QOL levels through the course of time, with the level of QOL related to treatment compliance (Gilman and Handwerk, 2001). These findings suggest that perceived life quality is an important and meaningful component of compliance and should be monitored throughout the course of treatment. Indeed, considering the aforementioned psychological, psychosocial, and psychoeducational benefits that are conferred to youth reporting optimal QOL, such monitoring may provide healthcare professionals with repeated opportunities to assess domain-specific life quality across treatment phases, in hopes that the reports change from ‘‘low’’ to ‘‘neutral’’ to ‘‘high.’’
2425
2426
142
Generic Quality of Life Measures for Children and Adolescents
Similarly, decrements in QOL reports across particular domains would suggest detrimental effects for youth receiving treatment.
1.2
The Challenge of Quality of Life Measurement
We must acknowledge at the outset that the definition of QOL and its subsequent measurement have been varied and diverse. Part of the challenge of QOL measurement stems from the fact that there has been little consensus as to what QOL ‘‘really is’’ (Schalock and Parameter, 2000), leading some QOL researchers to question the utility of the Term altogether (Rapley, 2003), mainly because it is not a directly measurable construct, nor is it a consistent entity across populations. Historically, QOL measurement as a societal-level concern has been approached from two broad perspectives: objective and subjective. A traditional objective approach focused on external, quantifiable conditions of a particular culture or geographic area such as income levels, quality and available housing, crime rates, divorce rates, access to medical services, school attendance, life expectancy, and suicide rates. On the other hand, a traditional subjective approach disputed measuring QOL in an exclusively objective sense, and instead, suggested subjective indicators (e.g., sense of community, satisfaction with life, sense of safety, relationships with family, etc.) should take precedent and complement more traditional, objective QOL indicators. Pragmatically speaking, the relationships between the two have been surprisingly modest, indicating that both approaches reveal unique information that is crucial to the understanding of overall life quality. As are result, some QOL instruments combine of self-reported subjective perceptions of life circumstances and objective circumstances (e.g., the number of physician visitations within a 3-month period, number of drugs or alcohol use events within the last 30 days, etc). Our take on QOL measurement in general, but specifically for > children and adolescents in this chapter, is the appropriateness of a measure for the research topic. For example, if a researcher is interested in measuring the QOL of adolescents’ family life, then it is best to have a family domain measure of QOL in a chosen instrument. More importantly, that family measure should reliably and validly assess what it purports to assess. Thus, the purpose of this chapter is to review generic measures of child and adolescent (ages 18 or less) QOL that have been applied in at least two populations (e.g., age groups, nationalities, etc.) and have yielded strong psychometric properties. Given the relevance of quality of life measurement to health care, we have reviewed measures that fit our criteria for inclusion by offering credible > reliability and > validity evidence.
2
Methodology
In order for measures to be included in this chapter, instruments must have met the following criteria: they were developed for children and adolescents under the age of 18, applied in at least two populations (e.g., age groups, nationalities, etc.) and have yielded strong psychometric properties. When a particular measure was not reported directly from a child or adolescent respondent, or if a parent or guardian provided the response, these QOL measures were also included. Measures were considered to be generic if the measure demonstrated
Generic Quality of Life Measures for Children and Adolescents
142
validity and applicability to non-clinical child or adolescent populations. Given these criteria, 15 generic quality of life measures for children and adolescents were identified through an extensive review of the literature. In examining the 15 generic quality of life measures, four met our search criteria. These four are: (1) the Comprehensive Quality of Life Scale-School Version (ComQOL-S5; Cummins, 1997); (2) the Quality of Life Profile-Adolescent Version (QOLPAV; Raphael et al., 1996); (3) the Youth Quality of Life (YQOL; Edwards et al., 2002; Patrick et al., 2002); and (4) the Multidimensional Students’ Life Satisfaction Scale (MSLSS; Huebner, 1994). Study sample demographics for each reviewed objective and subjective measure are provided in > Table 142‐1. Unless otherwise noted, additional tables and detailed scoring information of each reviewed scale are provided within the text. The 11 others [The Quality of Life Measure for Children (C-QOL); Personal Well-Being Index-School Children (PWI-SC); Child Health and Illness Profile (CHIP-AE); Child Health Questionnaire-Child Form (CHQ-CF80); Child Health Questionnaire-Parent Form (CHQPF50); Pediatric Quality of Life Inventory (PedsQL); Generic Children’s Quality of Life Measure (GCQ); Teddy Quality of Life (TedQL); Short Form-12 Health Survey (SF-12); Short Form (SF-36); Sickness Impact Profile (SIP)] were excluded because additional research is necessary or are they are not appropriate measures of QOL. All measures are completed by the child or adolescent with the exception of the CHQ-PF50, which is completed by the child’s parent or guardian.
2.1
Generic Measures Including both Objective and Subjective Aspects of QOL
2.1.1
Comprehensive Quality of Life Scale-School Version
The ComQOL-S5 (Cummins, 1997) (see > Table 142‐1 for full study references) is selfadministered and is comprised of seven domains (Material Well-Being, Health, Productivity, Intimacy, Safety, Place in Community, Emotional Well-Being). Scoring for each domain is computed by obtaining a satisfaction score for each domain and weighting that score by the perceived importance to the individual. Either component can be administered independently if so desired, and it has demonstrated validity and reliability with adolescents aged 12–18 years old (see Appendix for full scale complete with scoring instructions). Being the fifth edition of the ComQOL-S, the ComQOL-S5 has similar validity and reliability measures as the ComQOL-S fourth edition (ComQOL-S4). The ComQOL-S4 was originally examined by Bearsley (1997) and included 524 adolescents aged 14–17 years old who were: (1) homeless or ‘‘at risk’’ for homelessness; (2) community school students (with a high frequency of behavioral, emotional, family, or learning problems); and (3) non-homeless secondary school students. For the subjective portion of the ComQOL-S4, Importance (a = 0.76) and Satisfaction (a = 0.80) yielded acceptable levels of internal consistency. For the objective portion of the scale, correlations for each of the individual items versus the domain score were found to be significant (p < 0.01). Criterion-related validity was determined through comparisons with the Life Attitude Profile-Revised Scale (Recker, 1992) and the ComQOL-S4’s subjective portion. Results revealed the ComQOL-S4 was positively correlated with choice and responsibleness (r = 0.45), goal seeking (r = 0.16), and personal meaning (r = 0.61), and negatively with death acceptance (r = 0.01) and existential vacuum (r = 0.48).
2427
2428
142
Generic Quality of Life Measures for Children and Adolescents
. Table 142-1 Sampling characteristics of generic QOL measures for children and adolescents Measure
Location
ComQOL-S5
QOLPAV
YQOL-R
MSLSS
N
Mean Age Gender
Racea
Melbourne, Australia (Gullone and Cummins, 1998)
264
14.6
53% female
NA
Australia (Bearsley, 1997)
524
15.8
57% female
NA
Perth, Western Australia
363
13.8
67% female
NA
(Meuleners et al., 2003, 2005)
167
17.4
62% female
NA
Great Britain (Bradford et al., 2002)
899
14.0
46% female
94% C, 6% O
Seattle, WA, USA (Patrick et al., 2002)
236
12–18
30% female
80% C, 5% AA, 1% H
Seattle, WA, USA (Topolski et al., 2004) 214
14.8
Columbia, SC USA (Huebner and Dew, 222 1993a, b)
15.5
52% female
50% C, 51% AA
Columbia, SC USA (Huebner et al., 1998)
291
12.9
57% female
55% C, 41% AA, 4% O
Columbia, SC USA (Gilman et al., 2000) 321
16.1
65% female
42% C, 56% AA, 2% H
Midwest and Northeast, USA (Gilman et al., 2004)
13.0
44% female
84.3% C, 5% AA, 10.7% O
South Korea and South Carolina, USA 1,657 12.8 (Park et al., 2004)
49% female
50.4% K; 49.6% A
Croatia and Southeast, USA (Gilman et al., 2005)
632
14.7
60% female
46% CRxO; 54% A
Tokyo, Japan and Honolulu, USA (Ito and Smith, 2006)
995
NA
51% female
62% J; 38% A
159
80.5% C, 4.7% AA, 13.9% O
NA not available a C Caucasian; AA African American; H Hispanic; O other; K Korean; CRO Croatian; J Japanese; A American; NA not available
2.1.2
Quality of Life Profile: Adolescent Version
The Quality of Life Profile: Adolescent Version (QOLPAV) (Raphael et al., 1996) has also been applied to numerous populations (see > Table 142‐1 for full study references). The QOLPAV began its development with six high school student focus groups (grades 9–13) to gather information regarding what the term ‘‘quality of life’’ means and to gain insight on specific concerns that adolescents have in their lives. Questions developed from these focus groups were then examined by different adolescents to verify content relevance, identify missing topics, and assess the wording of the QOLPAV. This information led to a 54-item
Generic Quality of Life Measures for Children and Adolescents
142
questionnaire containing three broad domains (Being, Belonging, and Becoming), which were formed from nine subdomains (physical being, physical belonging, practical becoming, psychological being, social belonging, leisure becoming, spiritual being, community belonging, growth becoming). All items were responded to with a five-point Likert response scale ranging from 1 (‘‘Not at all Important’’/’’No Satisfaction at All’’) to 5 (‘‘Extremely Important’’/‘‘Extremely Satisfied’’). Reliability for the QOLPAV was calculated using a Cronbach’s alpha for each of the subdomains, the broad domains and for the overall score. For all of the broad domains, a 0.80 (Being items a = 0.85, Belonging items a = 0.83, Becoming items a = 0.87). For the subdomains all coefficients approached or exceeded 0.70 (physical being a = 0.68; physical belonging a = 0.67; practical becoming a = 0.74; psychological being a = 0.72; social belonging a = 0.69; leisure becoming a = 0.74; spiritual being a = 0.71; community belonging a = 0.68; growth becoming a = 0.79). The overall score demonstrated high internal consistency (a = 0.94). To assess the validity of the QOLPAV, correlational analyses were performed on measures of self-esteem, life satisfaction, social support, and life chances taken from the Youth Transition Study (Bachman et al., 1967). The QOLPAV demonstrated significant (p < 0.01) correlations with all four measures (ranging from 0.45 to 0.56). The QOLPAV is a private domain scale and therefore the actual scale items are not provided. However, complete information regarding the QOLPAV, including ordering information can be located at http://www.utoronto.ca/qol/qolPublications.htm.
2.2
Generic Subjective QOL Measures
2.2.1
Youth Quality of Life-Research Version
The Youth Quality of Life-Research Version (YQOL-R: Edwards et al., 2002) has been administered to separate groups of youth aged 12–18 years with and without disabilities (see > Table 142‐1 for full study references). YQOL-R validity analyses included 210 adolescents (with and without disabilities) with a test battery of 49 items. Convergent validity was established by comparing the measure with the German Munich Quality of Life Questionnaire for Children (KINDL; Ravens-Sieberer and Bullinger, 1998) and the Children’s Depression Inventory (CDI; Kovacs, 1992). Group/discriminant validity was assessed through the use of the Conners’ Auxiliary ADHD/DSM IV Instrument (CADS-A; Conners, 1997), the Functional Disability Inventory (FDI; Walker and Greene, 1991), and the Youth Disability Screener (Patrick et al., 1998). These analyses resulted in a 41-item measure consisting of five domains: Self, Relationship, Environment, and General Quality of Life. All of the domains in the YQOL-R correlated highly with all domains of the KINDL (average r = 0.73, p < 0.05). Additionally, findings indicate that the correlations between the YQOL-R and KINDL were significant and in the expected direction. These findings suggest that the YQOL-R is sensitive to adolescents’ current symptom status (e.g., the YQOL-R perceptual scores were significantly lower for those adolescents who scored above the depression cut-point on the CDI, and above the ADHD cut-point on the CADS-A; Patrick et al., 2002). > Table 142‐2 provides the YQOL-R items by domains. The QOL scores are summed and those are then transformed to a scale from 0 to 100 for ease of interpretation, where higher scores indicate higher QOL.
2429
2430
142
Generic Quality of Life Measures for Children and Adolescents
. Table 142-2 YQOL-R items by domain Self 1. I keep trying even if at first I don’t succeed 2. I can handle most difficulties that come my way 3. I am able to do most things as well as I want 4. I feel good about myself 5. I feel I am important to others 6. I feel comfortable with my sexual feelings and behaviors 7. I have enough energy to do the things I want to do 8. I am pleased with how I look 9. I feel comfortable with the amount of stress in my life 10. I feel it is okay if I make mistakes 11. I feel my life has meaning 12. My personal beliefs give me strength 13. I feel alone in my life 14. I feel left out because of who I am Relationships 15. I feel most adults treat me fairly 16. I feel I am getting the right amount of attention from my family 17. I feel understood by my parents or guardians 18. I feel useful and important to my family 19. I feel my family cares about me 20. My family encourages me to do my best 21. I feel I am getting along with my parents or guardians 22. I feel my parents or guardians allow me to participate in important decisions which affect me 23. I try to be a role model for others 24. I can tell my friends how I really feel 25. I am happy with the friends I have 26. I am satisfied with my social life 27. I feel I can take part in the same activities as others my age 28. People my age treat me with respect Environment 29. I feel my life is full of interesting things to do 30. I like trying new things 31. I like my neighborhood 32. I look forward to the future 33. My family has enough money to live a decent life 34. I feel safe when I am at home 35. I feel I am getting a good education 36. I know how to get the information that I need
Generic Quality of Life Measures for Children and Adolescents
. Table 142-2 (continued)
142
37. I enjoy learning new things 38. I feel safe when I am at school General QOL 39. I enjoy life 40. I am satisfied with the way my life is now 41. I feel life is worthwhile Items 2–4, 6–8, 9–11, 13–15, 19, 20, 23, 25, 26, 27, 33–36, 40, 41 use a 11-point rating scale with adjectival anchors ‘‘Not at All’’ to ‘‘Completely’’. Items 1, 5, 12, 16–18, 21, 22, 24, 28–32, 37, 39 use a 11-point rating scale with adjectival anchors ‘‘Not at All’’ to ‘‘A Great Deal’’
2.2.2
Multidimensional Students’ Life Satisfaction Scale
The Multidimensional Students’ Life Satisfaction Scale (MSLSS) is a 40-item, subjective QOL instrument that assesses life satisfaction across five domains shown to important to the lives of youth (Family, Friends, School, Living Environment, and Self) in addition to an overall life satisfaction assessment (Huebner, 1994) (see > Table 142‐1). The MSLSS is appropriate for use on children between the ages of 8 and 18 and has been administered to over 4,000 children and adolescents world-wide (see > Table 142‐1), making it one of the most frequently administered scales of its kind. The psychometric properties of the scale consistently yield robust values, with internal consistency estimates of each MSLSS domain ranging from 0.80 and higher. In addition, the Total score (which is an aggregate of all items and is akin to measure of general satisfaction) is 0.95 or higher in several studies (see Gilman and Huebner, 2000). Evidence for the construct validity has been found by comparing the domain scores with well-known behavior and self-concept scales, including the Behavior Assessment System for Children-Second Edition (BASC 2: Reynolds and Kamphaus, 2004) and the Self-Concept Scale (Marsh et al., 1984). Evidence for the construct validity has been consistently found using both exploratory (Huebner, 1994) and confirmatory factor analysis (Gilman et al., 1999). Finally, the factor structure of the scale is invariant with respect to nationality (Gilman et al., 2008). > Table 142‐3 provides the MSLSS items by domains. Scoring the MSLSS is straightforward. The four response options are assigned points as follows: (never = 1); (sometimes = 2); (often = 3); and (almost always = 4). Negatively-keyed items must be reverse scored (see > Table 142-3 for the list of negatively-keyed items). Hence, negatively-keyed items are scored so that almost always = 1, and so forth. Higher scores thus indicate higher levels of life satisfaction throughout the scale. It should be noted that a 6-point agreement format has been used with middle and high school students. In this case, response options are assigned points as follows: (1 = strongly disagree, 2 = moderately disagree, etc.). Because the domains consist of unequal number of items, the domain and total scores are made comparable by summing the item responses and dividing by the number of domain (or total) items. An abbreviated version of the MSLSS is also available for research purposes or when resources are limited: the Brief Multidimensional Students’ Life Satisfaction Scale (BMSLSS) (Seligson et al., 2003) (see > Table 142‐4). The BMSLSS contains one item for each domain of the MSLSS in addition to an overall item. Although space limits a detailed discussion of the BMSLSS, it has demonstrated acceptable psychometrics in several samples of youth ranging
2431
2432
142
Generic Quality of Life Measures for Children and Adolescents
. Table 142-3 Multidimensional students’ life satisfaction (MSLSS) scale structure Family I enjoy being at home with my family My family gets along well together I like spending time with my parents My parents and I doing fun things together My family is better than most Members of my family talk nicely to one another My parents treat me fairly Friends My friends treat me well My friends are nice to me I wish I had different friendsa My friends are mean to mea My friends are great I have a bad time with my friendsa I have a lot of fun with my friends I have enough friends My friends will help me if I need it School I look forward to going to school I like being in school School is interesting I wish I didn’t have to go to schoola There are many things about school I don’t likea I enjoy school activities I learn a lot at school I feel bad at schoola Living environment I like where I live I wish there were different people in my neighborhooda I wish I lived in a different housea I wish I lived somewhere elsea I like my neighborhood I like my neighbors This town is filled with mean peoplea My family’s house is nice There are lots of fun things to do where I live Self I think I am good looking
Generic Quality of Life Measures for Children and Adolescents
. Table 142-3 (continued)
142
I am fun to be around I am a nice person Most people like me There are lots of things I can do well I like to try new things I like myself a
Reverse keyed items
. Table 142-4 Brief multidimensional students’ life satisfaction scale (BMSLSS) items I would describe my satisfaction with my family life as: I would describe my satisfaction with my friendships as: I would describe my satisfaction with my school experience as: I would describe my satisfaction with myself as: I would describe my satisfaction with where I live as: Response options are a 7-point Likert-type scale based on anchors of terrible, unhappy, mostly dissatisfied, mixed (about equally satisfied and dissatisfied), mostly satisfied, pleased, and delighted
in age from elementary school to high school (Seligson et al., 2003, 2005). Furthermore, life satisfaction (via measurement with the BMSLSS) has been shown to be related to the six priority adolescent objective health behavior categories in the United States for premature morbidity and mortality as defined by the US Centers for Disease Control and Prevention. For instance, life dissatisfaction has been shown to be significantly associated with behaviors that lead to intentional and unintentional injuries (Valois et al., 2001, 2004a); increased use of tobacco, alcohol and other drugs (Zullig et al., 2001); increased risky sexual behavior (Valois et al., 2002); poor body image and dietary behavior (Valois et al., 2003); and physical inactivity (Valois et al., 2004b).
3
Future Directions for Research
In spite of the potential advantages of including multidimensional QOL measures in assessment strategies and prevention/intervention programs, several neglected areas remain. First, more theory-based research is needed. Although not necessarily problematic at the idiographic level, the selection of one QOL measure over another when examining QOL across groups and/or nations may only provide a partial explanation of the factors that contribute to QOL. Compounding this issue is the finding that not all QOL domains appear to have the same relevance across groups. For example, recent studies (Gilman et al., 2008; Park and Huebner, 2005) reported that mean scores were significantly different across QOL domains with the MSLSS, with these differences falling along the individualistic/collectivistic cultural continuum. Considering that many of the reviewed measures in this chapter have to date been administered to limited samples often representing only one cultural group, clearly additional studies are
2433
2434
142
Generic Quality of Life Measures for Children and Adolescents
necessary to establish generalization of their findings. Such studies are necessary to expand the empirical understanding of what domains should be included when investigating QOL among adolescents. Further, while the scales reviewed in this chapter have been administered to a variety of samples, many existing measures appear to be designed for a specific series of studies, or for a specific population. Few measures have attempted to expand their scope of focus to additional groups to address fundamental psychometric questions. For example, Gilman and Huebner (2000) recommended that existing child- or adolescent-focused QOL measures would benefit from rigorous studies of basic psychometric properties, including evaluations of normative samples, reliability, and validity. Additionally, research is needed to address limitations inherent in using self-reports, such as the potential effects of response distortions or social desirability. Although such strategies have been rarely used, other-raters (e.g., peers, teachers, parents) or the use of alternative methods of QOL assessment (such as using experiential time sampling) have been recommended as strategies to minimize response artifacts (see Huebner et al., 2007 for a comprehensive review of these strategies). Moreover, research on QOL has been limited mostly to cross-sectional research. Studies of the correlates of QOL offer a useful initial step, but advancement of QOL research will require longitudinal and/or experimental studies to clarify the directionality of relationships. Although in their infancy, studies have shown that low QOL precedes the occurrence of psychological distress and poor health (e.g., Suldo and Huebner, 2004). Such findings provide unique and necessary information regarding the directionality of QOL effects as well as potential support for dual-factor models of mental health (Greenspoon and Saklofske, 2001).
3.1
Quality of Life Measurement: How many Domains can and Should be Measured?
Another area of future research is the consideration of what actually constitutes QOL in the terms of optimal number and types of domains to measure. Some research has been conducted that has attempted to name a specific number of QOL domains noteworthy across multiple QOL studies. Whereas Felce and Perry (1995) outline five domains of QOL (Material, Physical, Social, Emotional, and Productive Well-Being), Cummins’ (1996) content analysis of 1,500 articles concerning the topic of QOL determined that at least seven domains compose subjective QOL: Material Well-Being, Health, Productivity, Intimacy, Safety, Community, and Emotional Well-Being. Of interest, Hagerty et al. (2001) suggest these domains should also be used to guide the use and selection of objective QOL assessments as well. However, translating these domains to children and adolescents presents some challenges. For instance, Cummins’ (1996) identified domains were based on studies of adults and support for the domains was primarily based on a content analysis procedures without following up with more psychometric analyses of particular scales (e.g., factor analyses, multitrait-multimethod tests of convergent and discriminant validity). The Multidimensional Students’ Life Satisfaction Scale (MSLSS; Huebner, 1994) has offered empirical support, but contains five domains. Were we to strictly abide by Cummins’ (1996) criteria, this would have precluded us from including the MSLSS in our review. What is more important to us is the appropriateness of the measure for the research topic. That is, if a researcher is interested in QOL among peers members, then the measure should contain a ‘‘peers’’ or ‘‘friends’’ domain; and each domain should be supported by research-based evidence of its validity.
Generic Quality of Life Measures for Children and Adolescents
142
Thus, continuing research is needed to empirically support the various multidimensional models available. There are also developmental issues we must acknowledge that challenge measurement in general, whether it is intelligence testing or QOL. If we borrow from Piaget’s influential work, we understand that as youth age, they move through a series of stages with progressing cognitive abilities. Consequently, what is an appropriate QOL measure for youth in preoperational development (ages 2–7) where reasoning is limited owing to egocentrism, centering, and transductive reasoning may not be useful for youth in formal operational development (ages 11 and older) where youth are better equipped to think more abstractly, solve problems mentally, and engage in hypothetico-deductive reasoning. In other words, there is reason to believe that the nature and number of QOL domains that children and adolescents are able to differentiate may change across age groups. Younger children may be unable to distinguish domains to the same extent as older adolescents. Nevertheless, some preliminary research is emerging in the United States that attempts to maximize available data to assess child and adolescent QOL in the form of national index.
4
National Adolescent Indexes to Match Objective and Subjective QOL Data
Countries have long held the belief that monitoring QOL at the national-level is of great importance resulting in the creation of numerous indexes which were recently systematically reviewed (see Hagerty et al., 2001). Noteworthy conclusions from this comprehensive review (p. 86) include the need for: (1) the formulation of a domain structure common to all indexes; (2) additional predictive validity or sensitivity studies to establish causality; and (3) testing convergent validity against other QOL indexes. In this regard, there has been some progress, specifically pertaining to adolescent QOL measurement. Promising efforts to match population objective and subjective QOL measurement into an overall index have been conducted by Land and colleagues (Land et al., 2001, 2007) using the Child and Youth Well-being Index (CWI). The CWI pulls from a variety of objective and subjective data sources (e.g., vital statistics and population-level survey data) and is based on 28 national-level indicators (see Land et al., 2001 for a full review and justification of these 28 indicators). Because 25 of the 28 indicators used in the CWI date back to 1975, while three (health insurance coverage, activity limitation reports, and subjective health assessments) were implemented in 1985, the CWI is calculated and indexed first by year: 1975 or 1985. According to Land et al. (2007, p. 112), ‘‘The base year value of the indicator is assigned a value of 100 and subsequent values of the indicator are taken as percentage changes in the index. The directions of the indicators are oriented so that a value greater (equal or lesser) than 100 in subsequent years means the social condition measured has improved (no change or deteriorated),’’ relative to its base year. Land and colleagues (2007) have demonstrated the usefulness of the CWI by each of Cummins’ (1996) QOL domains using 1975 as the referent year through 2003, indicating that from 1975 through 1993, overall adolescent QOL declined in the USA. However, from 1994 onward, the CWI indicated a sustained improvement in the QOL of American youth. When the CWI was computed by domain, Land et al. (2007) determined that much of the decline was in response to a decline in relationships and emotional/spiritual well-being. Even more interesting was an observed and persistent decrease in health from the mid-1980s through 2003. Land et al. (2007) attributes this decline to the increasing prevalence of obesity among youth in the United States and provides compelling sensitivity data showing that when the
2435
2436
142
Generic Quality of Life Measures for Children and Adolescents
obesity indicator is included in the Health domain, it decreases by 24% from 1981 to 2001. When the obesity indicator is excluded, the health domain increases from 1975 to 1983 and has since remained relatively stable. Equally compelling comparisons are made by race, gender, and even overall life satisfaction levels of twelfth grade students. However, as with any developing index or measure, some limitations are extended to the CWI.
4.1
Limitations of the CWI
First, the CWI is highly reliant on objective indicators and restricted to currently available, national data resources. For example, only three of the 28 indicators are based on subjective assessments (e.g., self-perceived excellent/very good, good, fair, or poor health; activity limitation; and the importance of religion). Second, as observed by Land and colleagues (2007), the CWI domains of Intimacy and Emotional and Spiritual Well-Being are underdeveloped. In fact, a direct measure of Emotional and Spiritual Well-Being is not available and is instead measured via indicators of suicide, church attendance, and the importance of religion in one’s life. As noted earlier, knowing the modest relationship between objective and subjective QOL indicators, despite the robust relationships identified between the CWI and several other adolescent health measures, additional adolescent QOL item development is recommended. Third, a comprehensive understanding of the relationship between mental ‘‘health’’ and mental illness needs to be made before such an index is realized. To date, quality of life studies have primarily focused on analyzing characteristics leading to positive outcomes, while the converse continues for those who research mental illness. There seems to be little discourse in the way of understanding the nexus between the two worlds. Although some efforts have been made in this area (e.g., Schwartz et al., 2007), much work remains in explaining adolescent psychosocial development.
4.2
Conclusions
Fortunately, there are quality measures already available that have yet to be incorporated into national-level youth surveys. For instance, all 15 identified generic adolescent QOL measures located for this chapter contain adequate measures of both intimacy and emotional wellbeing, elements which might be incorporated into national-level surveys to enhance the scope of measurement systems, such as the CWI. In conjunction with QOL researchers, local, regional, and national policy makers may wish to determine which particular QOL domains should be assessed to best serve their unique needs and interests, then select or develop various indices based on the research evidence available for the specific measures available, In short, they must address the question, ‘‘What should be measured for which specific purpose(s), with which particular population(s), under what specific conditions(s)?
Summary Points QOL appears to be a crucial psychological strength that is related to healthy adaptation in adolescents.
Generic Quality of Life Measures for Children and Adolescents
142
The use of QOL ratings may provide important information above and beyond that
what is traditionally measured in health care settings-typically symptom presence and severity. QOL measures can and should be used in conjunction with severity measures to monitor the effectiveness of a given intervention on enhanced life quality, in addition to symptom reduction. QOL reports can provide a multicontextual perspective because they gather important information across a number of life domains and may help monitor how adolescents’ perceive the quality of interventions, which is in turn a factor contributing to treatment compliance. This chapter reviewed generic measures of child and adolescent (ages 18 or less) QOL that have been applied in at least two samples (e.g., age groups, nationalities, etc.) and have yielded strong psychometric properties. Four qualifying measures were identified and divided into two categories: (1) measures that assessed both objective and subjective (n = 2) and (2) measures that assessed subjective QOL only (n = 2). Acceptable objective and subjective QOL measures are: (1) the Comprehensive Quality of Life Scale-School Version (ComQOL-S5), and (2) the Quality of Life Profile-Adolescent Version (QOLPAV). Acceptable subjective QOL measures are: (1) the Youth Quality of Life (YQOL), and b) the Multidimensional Students’ Life Satisfaction Scale (MSLSS). Eleven other measures [The Child Quality of life (C-QOL); Personal Wellbeing IndexSchool Children (PWI-SC); Child Health and Illness Profile (CHIP-AE); Child Health Questionnaire-Child Form (CHQ-CF80); Child Health Questionnaire-Parent Form (CHQ-PF50); Pediatric Quality of Life Inventory (PedsQL); Generic Children’s Quality of Life Measure (GCQ); Teddy Quality of Life (TedQL); Short Form-12 Health Survey (SF12); Short Form (SF-36); Sickness Impact Profile (SIP) were identified as, but are not being included. Although not necessarily problematic at the idiographic level, the selection of one QOL measure over another when examining QOL across groups and/or nations may only provide a partial explanation of the factors that contribute to perceived life quality. Few measures have attempted to expand their scope of focus to additional groups to address fundamental psychometric questions. Moreover, research on QOL has been limited mostly to cross-sectional research. Studies of the correlates of QOL offer a useful initial step, but advancement of QOL research will require longitudinal and/or experimental studies to clarify the directionality of relationships. Advocations for including well-being indicators as a national index of positive mental health have been made for adults and such an index would be salient for adolescents as well. Nevertheless, a comprehensive understanding of the relationship between mental ‘‘health’’ and mental illness needs to be made before such an index is realized. Efforts have been undertaken to address how many domains should QOL measure, but additional research is needed to empirically support a specific domain structure. The Child and Youth Well-being Index (CWI) currently being tested on youth in the United States is providing the first steps toward that empirical support.
2437
2438
142 Appendix A
Generic Quality of Life Measures for Children and Adolescents
Generic Quality of Life Measures for Children and Adolescents
142
2439
2440
142
Generic Quality of Life Measures for Children and Adolescents
Generic Quality of Life Measures for Children and Adolescents
142
2441
2442
142
Generic Quality of Life Measures for Children and Adolescents
Generic Quality of Life Measures for Children and Adolescents
142
2443
2444
142
Generic Quality of Life Measures for Children and Adolescents
Generic Quality of Life Measures for Children and Adolescents
142
2445
2446
142
Generic Quality of Life Measures for Children and Adolescents
Generic Quality of Life Measures for Children and Adolescents
142
2447
2448
142
Generic Quality of Life Measures for Children and Adolescents
Generic Quality of Life Measures for Children and Adolescents
142
2449
2450
142
Generic Quality of Life Measures for Children and Adolescents
References Bachman JG, Kahn RL, Mednick MT, Davidson TN, Johnston LD. (1967). Youth in Transition: Volume 1: Blueprint For a Longitudinal Study of Adolescent Boys. Institute for Social Research, University of Michigan, Ann Arbor. Bearsley C. (1997). No Place Called Home: Quality of Life and Meaning of Life of Homeless Youths. Honours Thesis, School of Psychology, Deakin University, Melbourne. Bradford R, Rutherford DL, John A. (2002). J Adolesc. 25: 261–274. Carlson MJ, Gabriel RM. (2001). Psychiatr Serv. 52: 1230–1236. Conners CK. (1997) Conners’ Auxillary ADHD/DSM-IV Adolescent Self-Report Scale (CADS-A). Multihealth systems, North Tonawanda, NY. Cummins RA. (1996). Soc Indic Res. 38: 303–332. Cummins RA. (1997). Comprehensive Quality of Life Scale-School Version (Grades 7–12), 5th (ed.). Deakin University.
Edwards TC, Huebner CE, Connell FA, Patrick DL. (2002). J Adolesc. 25: 275–28. Felce D, Perry J. (1995). Res Devel Disabil. 16: 51–74. Frisch MB. (1998). Clin Psychol: Sci Prac. 5: 19–40. Frisch MB, Clark MP, Rouse SV, Rudd MD, Paweleck J, Greenstone A, et al. (2003). Predictive validity and sensitivity to change in quality of life assessment and life satisfaction: Further studies of the Quality of Life Inventory or QOLI in mental health settings. In: Sirgy MJ et al. (eds.) Advances in Quality of Life Theory and Research. Kluwer Academic Publishers, The Netherlands, pp. 191–210. Frisch MB, Cornell J, Villanueva M, Retzlaff PJ. (1992). Psychol Assess. 4: 92–101. Gilman R, Ashby J, Sverko D, Florell D, Varjas K. (2005). Pers Individ Dif. 39: 155–166. Gilman R, Easterbrooks S, Frey M. (2004). Soc Indic Res. 66: 143–164. Gilman R, Handwerk ML. (2001). Resid Treat Child Youth. 18: 47–65.
Generic Quality of Life Measures for Children and Adolescents Gilman R, Huebner ES. (2000). Behav Change. 17: 178–183. Gilman R, Huebner ES. (2004). Resid Treat Child Youth. 21: 7–17. Gilman R, Huebner ES. (2006). J Youth Adolesc. 35: 311–319. Gilman R, Huebner ES, Laughlin JE. (2000). Soc Indic Res. 52: 135–160. Gilman R, Huebner ES, Park N, Tian L, O’Byrne J, Schiff M, et al. (2008). J Youth Adolesc. 2: 142–154. Gilman R, Laughlin JE, Huebner ES. (1999). Sch Psychol Int. 20: 300–307. Greenspoon PJ, Saklofske DH. (2001). Soc Indic Res. 54: 81–108. Gullone E, Cummins RA. (1998). Behav Change. 16: 127–139. Hagerty MR, Cummins RA, Ferriss AL, Land K, Michalos AC, Peterson M, Sharpe A, Sirgy J, Vogel J. (2001). Soc Indic Res. 55: 1–96. Huebner ES. (1994). Psychol Assess. 6: 149–158. Huebner ES, Dew T. (1993a). J Psychoeduc Assess. 11: 345–350. Huebner ES, Dew T. (1993b). Sch Psychol Int. 14: 355–360. Huebner ES, Laughlin JE, Ash C, Gilman R. (1998). 16: 118–134. Huebner ES, Gilman R. (2006). Appl Res Qual Life. 1: 139–150. Huebner ES, Gilman R, Suldo SM. (2007). Assessing perceived quality of life in children and adolescents. In: Handler L, Smith S (eds.) The Clinical Assessment of Children and Adolescents. A practitioner’s handbook. Lawrence Erlbaum, Mahwah, NJ, pp. 347–363. Huebner ES, Seligson JL, Valois RF, Suldo SM. (2006). Soc Indic Res. 79: 477–484. Huebner ES, Valois RF, Suldo SM, Smith LC, McKnight CG, Seligson JL, Zullig KJ. (2004). J Adolesc Health. 34: 270–278. Irwin CE, Burg SJ, Cart CU. (2002). J Adoles Health. 31: 91–121. Ito A, Smith DC. (2006). The Comm Psychol. 38: 19–21. Jirojanakul P, Skevington S. (2000). Br J Health Psychol. 5: 299–321. Jirojanakul P, Skevington S, Hudson J. (2003). Soc Sci Med. 57: 1277–1288. Keyes CLM. (2007). Am Psychol. 62: 95–108. Kovacs M. (1992). Children’s Depression Inventory (CDI). Multi-Health Systems, North Tonawanda, NY. Land KC, Lamb VL, Meadows SO, Taylor A. (2007). Soc Indic Res. 80: 105–132. Land KC, Lamb VL, Mustillo SK. (2001). Soc Indic Res. 56: 241–320. Marsh HW, Barnes J, Cairns L, Tidman M. (1984). J Educ Psychol. 940–976.
142
Meuleners LB, Lee AH. (2005). Qual Life Res. 14: 1057–1063. Meuleners LB, Lee AH, Binns CW, Lower A. (2003). Qual Life Res. 12: 283–290. Moore KA, Brown BV, Scarupa MS. (2003). The uses (and misuses) of social indicators: Implications for public policy. Child Trends Research Brief, Publication #2003–01. Child Trends, Washington, DC. Patrick DL, Connell FA, Edwards TC, Topolski TD, Huebner CE. (1998). Final report submitted to the Centers for Disease Control and Prevention: Special Interest Project #4 (96): Age Appropriate Measures of Quality of Life and Disability Outcomes among Children: Youth Quality of Life Study. Center for Disability Policy and Research, Seattle, WA. Patrick DL, Edwards TC, Topolski TD. (2002). J Adolesc. 25: 287–300. Park N, Huebner ES. (2005). J Cross-Cult Psychol. 36: 444–456. Park N, Huebner ES, Laughlin JE, Valois RF, Gilman R. (2004). Soc Indic Res. 66: 61–79. Raphael D, Rukholm E, Brown I, Hill-Bailey P, Donato E. (1996). J Adolesc Health. 19: 366–375. Rapley M. (2003). Should we ‘hang up quality of life as a hopeless term?. In: Rapley M (ed.) Quality of Life Research: A Critical Introduction. Sage Publications, Thousand Oaks, CA, pp. 212–226. Ravens-Sieberer U, Bullinger M. (1998). Qual Life Res. 7: 399–407. Recker GT. (1992). Life Attitude Profile-Revised. Trent University, Canada. Reynolds CR, Kamphaus RW. (2004). Manual for the Behavioral Assessment System for Children, 2nd version. American Guidance Services, Circle Pines, MN. Schalock RL, Parameter T. (2000). Preface. In: Quality of Life: Its Conceptualization, Measurement, and Application. IASSID, A Consensus Document. Washington. Schwartz SJ, Pantin H, Coatsworth JD, Szapocznik J. (2007). J Prim Prev. 28: 117–144 Selgison JL, Huebner ES, Valois RF. (2003). Soc Indic Res. 61: 121–145. Seligson JL, Huebner ES, Valois RF. (2005). Soc Indic Res. 73: 355–374. Suldo SM, Huebner ES. (2004). Sch Psychol Q. 19: 93–105. Suldo SM, Huebner ES. (2006). Soc Indic Res. 78: 179–203. Topolski TD, Edwards TC, Patrick DL, Varley P, Way ME, Buesching DP. (2004). J Atten Disord. 3: 163–173. Valois RF, Zullig KJ, Drane JW, Huebner ES. (2001). Am J Health Behav. 25: 353–366. Valois RF, Zullig KJ, Huebner ES, Drane JW. (2003). Eat Disord: J Treat Prev. 11: 271–288. ValoisRF, Zullig KJ, Huebner ES, Drane JW. (2004a). J Sch Health. 74: 59–65.
2451
2452
142
Generic Quality of Life Measures for Children and Adolescents
Valois RF, Zullig KJ, Huebner ES, Drane JW. (2004b). Soc Indic Res. 66: 81–105. Valois RF, Zullig KJ, Kammermann SK, Huebner ES, Drane JW. (2002). J Child Fam Stud. 11: 427–440. Walker LS, Greene JW. (1991). J Pediatr Psychol. 16: 39–58.
Zullig KJ, Valois RF, Drane JW, Huebner ES. (2001). J Adolesc Health. 29: 279–288. Zullig KJ, Valois RF, Huebner ES, Drane JW. (2005). Qual Life Res. 14: 1573–1584.
143 Quality of Life in Children with Cerebral Palsy A. Aran 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2454
2 Assessment Tools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2456 2.1 Parental and Child Perceptions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2456 2.2 Measures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2456 3
Generic Tools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2457
4 Specific Tools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2458 4.1 Function and Participation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2461 5 5.1 5.2 5.3
QOL in Children with CP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2461 Parent’s View . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2461 Children’s View . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2461 Factors Affecting HRQL in Children with CP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2461
6
Level of Motor Disability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2462
7
Cognitive Impairment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2464
8
Other Co-Morbidities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2464
9
Pain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2465
10
Socioeconomic Status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2465
11
Child Characteristics (Age, Gender) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2465
12
Parenting Style . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2465
13
Summary and Recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2465 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2466
#
Springer Science+Business Media LLC 2010 (USA)
2454
143
Quality of Life in Children with Cerebral Palsy
Abstract: > Cerebral Palsy (CP) is a non-progressive disorder of movement and posture caused by a defect or injury to the immature brain and its impact is further exacerbated by disabilities other than the motor impairments, such as epilepsy, learning disabilities, behavioral and emotional problems. Traditionally, the treatment of children with chronic diseases, as Cerebral Palsy (CP) was focused on the physical aspects of the disease and the treatment efficacy was measured primarily by physical improvement. In the past two decades, quality of life (QOL), defined, as well-being across various broad domains, has become an important treatment goal, especially in chronic diseases like CP. As studies on QOL, are highly subjective in nature and may have many inherent limitations, a combination of well valid, parent-based and child-based questionnaires as well as generic and disease specific tools is required. Although several studies have found that parents report lower QOL for their children with CP (in every aspect of QOL measured), the children themselves, usually rate their quality of life in the emotional and social domains, equal to their typically developed peers. Further more, parents of children with severe impairment, often reported better quality of life in the psychosocial domains compared to the reports for children with mild impairment. These consistent findings suggest that children with cerebral palsy can adapt well to their activity limitations and may have satisfactory quality of life in spite of significant deficits. These results also mean that factors other than the impairment severity may have a major influence on QOL in disabled children. The consensus that improving QOL is an important treatment goal in children with CP, mandates measures and treatments that enhance this goal. List of Abbreviations: CHQ, > Child Health Questionnaire; CP, Cerebral Palsy; GMFCS, Motor Function Classification System; PedsQL, 4.0, Pediatric Quality of Life Inventory Version 4.0; PODCI, Pediatric Outcomes Data Collecting Instrument; QOL, Quality of Life
> Gross
1
Introduction
Cerebral palsy (CP) is the most common cause of physical handicap in children, with an estimated rate of 2–2.5 per 1,000 live births (for Key Facts of cerebral palsy, please see > Table 143-1). The term CP represents a group of conditions that are caused by a permanent static lesion of the motor areas in the developing brain. CP is caused by either inborn developmental problem or by acquired injury, like ischemic or hemorrhagic stroke, anoxia, traumatic injury or infection. The lesion occurred before the birth, during the birth, or within the first 2–3 years of life. Even though the lesion itself does not change, the clinical manifestations of the lesion change as the child grows and develops. The motor skills of most children with cerebral palsy improve as they grow, but the rate of improvement is slower in children with cerebral palsy compared to unaffected children. As a result of the lesion in the brain’s motor areas, the child faces difficulties with stance and active motions. More than 50% of the patients are able to walk without arm assistance but cannot run or Jump; 20% can walk only with assistance and 25% cannot walk at all. If the lesion involves other brain areas in addition to the motor area, most commonly in large lesions that cause severe motor impairment, the physical disability is often associated with other problems such as epilepsy (35%) learning disabilities (50%), cerebral visual impairment (40%) and mental retardation (30%) (Koman et al., 2004). For research and clinical purposes, CP is usually graded according to the Gross Motor Function Classification System (GMFCS). This scale measures severity of CP and rates
Quality of Life in Children with Cerebral Palsy
143
. Table 143-1 Key Facts of cerebral palsy Definition CP is an ‘‘umbrella term’’ for a group of physical disabilities (stance and movements) that are caused by a permanent static damage to the motor control centers of the developing brain Causes
Common causes of CP are disturbances of the blood supply to the brain or infections that occur in most cases during pregnancy or in pre term infants and sometimes during infancy
Types
CP is commonly classified according to the anatomical distribution of the impairment: Diplegia – legs are involved much more than arms; Hemiplegia – involvement of arm and leg in the same side; Quadriplegia – all four limbs are involved. The dominant movement disorders are Spasticity (stiff muscles) and ataxia (lack of muscle coordination)
> Table
143-1 summarizes briefly the definition, classification and the common causes of cerebral palsy. CP cerebral palsy
. Table 143-2 Overview of Gross Motor Function Classification System (Palisano et al., 1997) Level I
Walks without restrictions; limitation in more advanced gross motor skills
Level II
Walks without assistive devices; limitation walking out doors and in community
Level III
Walks with assistive devices; limitation walking out doors and in community. (usually requiring wheelchair)
Level IV
Self mobility (usually with wheelchair) with limitations; children are transported or use power mobility out doors and in community
Level V
Self mobility (even with the use of power wheelchair) is very limited
> Table
143-2 describes the Gross Motor Function Classification System – a common classification of the motor severity in Cerebral Palsy into five levels from the milder impairment (level I) to the most severe (level V)
outcome of motor function in a scale ranging from I to V (> Table 143-2). CP is usually described as mild if the child can walk independently (corresponds to levels I/II), moderate if the child is ambulant with assistive devices (level III) and severe if the child is wheelchair dependent (levels IV/V) (Oeffinger et al., 2004). The impact of CP, the motor impairment and co-morbidities on most individuals and their families is usually substantial. The treatment is supportive (there is no cure for CP) and is often long term and complex. The exact treatment regimen depends on the specific type of CP and co-morbidities but usually includes physical and occupational therapy, special education, orthopedic surgery, and assistive devices. Frequently, further treatments are applied and the emotional and financial costs on the children, their parents, and the health services are substantial. Whereas treatment modalities of CP have been documented in a multitude of studies, their subjective impact on children with CP and their families is an area that has been relatively neglected (Bjornson and McLaughlin, 2001; Majnemer and Mazer, 2004). Considering the supportive rather than curative nature of the various treatment modalities, it is now well accepted that improving quality of life should be the main treatment goal for children with CP and specific assessment of QOL should be an integral part of any outcome measurement.
2455
2456
143
Quality of Life in Children with Cerebral Palsy
2
Assessment Tools
2.1
Parental and Child Perceptions
The challenge of assessing a highly subjective parameter like QOL in a quantitative tool led to the development of many different measurement instruments. The first studies of QOL in children, were based on the parent’s perspective in most cases (Bullinger and Ravens-Sieberer, 1995). With time, data based on children self-reports began to accumulate, driven by a strong recommendation of the World Health Organization and the International Association for Child Psychology and Psychiatry, that measures of QOL in children will use self-reports whenever possible (World Health Organization, 1993). The accumulating data demonstrated good psychometric properties even in young children and gradually, child based, self reported QOL measurements, became standard of care (Connolly and Johnson, 1999; Harding, 2001; Landgraf and Abetz, 1996). However, QOL assessment in children, especially those with developmental problems, cannot be based only on self-report tools. For many children, parent’s report, is the only option due to a young age, severe motor disability, or cognitive impairment. Even if the child is capable of self-reporting, some studies have found children’s reports to be less reliable (Goodwin et al., 1994). Several studies compared the parents’–proxy report on the QOL of their children to the children’s self-report. Interestingly, the correlation between the parents and the children reports, is depended considerably on the children’s health status. While parents generally perceive the QOL of their healthy children as higher than the children themselves (Cremeens et al., 2006; Theunissen et al., 1998; Waters et al., 2003), They score the QOL of their disabled children, lower than the self-reports of the disabled children (Bastiaansen et al., 2004; Havermans et al., 2006; Majnemer et al., 2007; Ronen et al., 2003; Russell et al., 2006; Shelly et al., 2008; Varni et al., 2005; White-Koning et al., 2007). While many factors can influence this discrepancy between parental and child’s views, a recent large study, found that parental stress and child’s pain are important contributors. When parents experience difficulties in their own lives they tend to report lower QOL for their children and conversely, when a child reports a severe pain, his self QOL report is lower compared to his parent’s report, suggesting that parents may underestimate the extent to which pain affect their child’s life (White-Koning et al., 2007). Considering the subjective nature of Quality of life, the question whose perspective reflects the child’s QOL more ‘‘reliably’’ (the child’s perspective or the parental), is somewhat irrelevant. Today, it is usually accepted that each type of instrument, highlights different aspects of the child QOL and both are important. Together they create a complementary picture of the child’s health status that helps us to define the child’s needs for better QOL (De Civita et al., 2005; Eiser and Morse, 2001a, b; Shelly et al., 2008; White-Koning et al., 2007).
2.2
Measures
QOL measurements are based on the use of standardized and validated questionnaires. For the assessment of > health-related quality of life (HRQL) in people with CP, two major types of questionnaires are available: Generic and disease specific. > Generic questionnaires were developed to measure the HRQL of the general population regardless of health condition. These tools are generally being used for discrimination
Quality of Life in Children with Cerebral Palsy
143
studies: the comparison of different aspects of QOL between patients and the healthy population or between different groups of patients, separated by severity of the disease or any other physical or psychosocial parameter. The advantage of these questionnaires is the general nature of their items, that allows them to be relevant for wide spectrum of health conditions. This general nature is also a disadvantage when it comes to detection of minor changes in HRQL following a disease specific intervention. To answer this need, > disease specific questionnaires have been developed. These questionnaires contain many items that are more relevant to patients with a certain type of disease, which enable them to be sensitive to changes across time and interventions. As mentioned above, when we deal with children, each type of questionnaire is further subdivided according to whether it is supposed to be answered by the children themselves or by a proxy of the children, usually a parent. Some generic questionnaires has both child and parent versions for the same items. While currently, most of the assessment instruments for QOL in children with CP, are generic questionnaires that approach the parents, more specific and child based questionnaires, appears recently. In the last two decades, studies of QOL in children with CP have used many different tools. A recent metanalysis identified 17 instruments, which have been used in more than 100 studies (Viehweger et al., 2008). It is beyond the scope of this article to describe all those different tools and only questionnaires used in numerous and/or large studies, will be discussed.
3
Generic Tools
Various aspects of some of the more common generic questionnaires used for assessment of QOL in children with CP, are summarized in > Table 143-3. Child Health Questionnaire (CHQ) is a generic questionnaire that taps physical and psychosocial aspects of QOL as well as impact on family (> Table 143-4). A score of zero represents the worst health state and 100 the best. Each scale has its own score and additionally there are two summary scores: physical and psychosocial. The CHQ has demonstrated very good reliability and validity in population and clinical studies all over the world (Landgraf and Abetz, 1996). As other generic questionnaires, the CHQ is less sensitive to changes over time or following interventions, and may fail to detect meaningful improvements in functioning (Vargus-Adams, 2006). Nevertheless, the CHQ appears to be a reliable, valid, and acceptable tool for children with CP across a range of severities (Vargus-Adams, 2005; Wake et al., 2003). It can reliably reflect how parents perceive the health and well-being of their children and is extensively used in both the research and clinical fields. A common clinical use is as a screening tool for lower QOL (especially in the psychosocial aspects of QOL) of the child with CP or for a higher burden of care carried by the family, than would be anticipated by the clinical assessment. The CHQ has three parent’s forms that contain 28, 50, or 87 items (CHQ PF – 28, CHQ PF – 50 and CHQ PF – 87 respectively) as well as a child’s form with 87 questions (CHQ CF – 87). > KIDSCREEN – KIDSCREEN is a relatively new, generic questionnaire that has child and adolescent (ages 8–18 years) as well as parent/proxy versions. Three KIDSCREEN instruments are available: KIDSCREEN-52 (long version) covering ten HRQOL dimensions, KIDSCREEN-27 (short version) covering five HRQOL dimensions and the KIDSCREEN-10 Index as a global HRQOL score. The questionnaire has been psychometrically validated recently on 22,110
2457
2458
143
Quality of Life in Children with Cerebral Palsy
. Table 143-3 Common generic questionnaires, for QOL assessment in children with CP Instrument CHQ
Versions Child
Parent Kidscreen
Child
Parent PedsQL
Child Parent
Age
Scales and number of items
10–18 Physical functioning; General health; Pain, Social limitations due to Physical disability and due to Emotional difficulties; Behavior; Mental health; Self esteem; Impact on family activities, on Parent’s emotions and on Parent’s free time; Family cohesion. 28/50 or 87 items 5–18 8–18 Physical well-being; Psychological well-being; Moods and emotions; Self-perception; Autonomy; Relationships with parents; Social support and peers; School environment; Social acceptance; Financial resources. 27/52 or 10 questions 8–18 5–18 Physical functioning; Moods and emotions; Social functioning, School functioning 5–18 23 items
> Table 143-3 expresses
important features of the three most common quality of life questionnaires, designed for the general population and used to assess quality of life in children with cerebral palsy. The ages approached by each version (the child and the parent versions), the scales covered by each tool and the total items in the various questionnaires are listed above. CHQ Child Health Questionnaire; QOL Quality of Life; PedsQL Pediatric Quality of Life Inventory; CP Cerebral Palsy
European children (Ravens-Sieberer, 2005).The version that has been used for disabled children, the KIDSCREEN-52 covers the following 10 dimensions: physical well-being, psychological wellbeing, moods and emotions, self-perception, autonomy, relationships with parents, social support and peers, school environment, social acceptance and financial resources. For each domain, item responses are summed and a score out of 100 is computed with higher scores indicating better QOL. The time required for administration is 15–20 min. This questionnaire have been used in the largest study so far, on QOL of children with CP (818 parents have answered the parent’s version and 500 children have answered the child’s version). > Pediatric Quality of Life Inventory (PedsQL) – The PedsQL is a 23 items generic questionnaire, which was designed to enable integration of disease-specific modules (its PedsQL-CP Module is described in the next paragraph). It has parent’s form and three versions of child’s form (5–7 years, 8–11, 12–18) with good psychometric properties both in healthy children and children with chronic conditions (Varni et al., 2007, 2006). The PedsQL covers four domains: Physical functioning (eight items), Emotional functioning (five items), Social functioning (five items) and School functioning (five items) and the mean of the last three domains creates the Psychosocial summary score (the sum of the items divided by the number of items answered).
4
Specific Tools
Various aspects of disease specific questionnaires for assessment of QOL in children with CP, are summarized in > Table 143-5.
Quality of Life in Children with Cerebral Palsy
143
. Table 143-4 Description of the CHQ scales – Adapted from the US CHQ manual (Landgraf and Abetz, 1996) Domain Physical
Scale Physical Functioning
The extent of limitations in physical activities, including self-care, due to health problems
General Health Perceptions and Global health
Parent subjective assessment of overall child’s health; past, present, and future
Role/Social LimitationsPhysical
Limitations on school work and activities with friends as a result of physical health problems
Bodily Pain Psychosocial Role/Social LimitationsEmotional/Behavioral
Family
Description
Intensity and frequency of general pain and discomfort Limitations on school work and activities with friends due to emotional or behavioral difficulties
Behavior and Global Behavior
Frequency of behavior problems: improper, aggressive or delinquent behavior
Mental Health
Frequency of depressive or anxiety versus happy states
Self-esteem
Satisfaction with school and athletic abilities, looks, relationships, and life overall
Family Activities
Frequency of interruption of child’s general health to family activities
Family Cohesion
How well the family gets along with one another
Parent Impact-Emotional
Level of distress experienced by parent due to child’s health and well-being
Parent Impact-Time
Limitation on parental time for personal needs due to child’s health and well-being
> Table 143-4 explains each of the 12 scales of the child health questionnaire, covering three domains of quality of
life: Physical, Psychosocial and impact on family. CHQ Child Health questionnaire
> Pediatric
Outcomes Data Collection Instrument (PODCI) is a 55 questions, disease specific questionnaire that is available as self- or proxy report forms for Adolescents (11–18 years) and as parent’s form for children 2–10 years old. The American Academy of Orthopedic Surgeons (AAOS) developed the PODCI to measure general health and problems related to bone and muscle conditions in children (AAOS, 1997). The questionnaire contains six domains: Upper extremity and physical function, transfers and basic mobility, sports and physical function, pain/comfort, expectations from treatment and happiness with physical condition (Daltroy et al., 1998). Zero is the poorest score and 100 reflects best health status. A general function and symptom score is computed as a composite of the first four domains (the three physical function domains and the pain and comfort domain). This instrument allows orthopedic surgeons to assess the functional health and efficacy of treatment of their patients at baseline and follow-up, but provides less information on the impact of the condition on the family and the individual’s self-esteem. PedsQL-CP Module – The PedsQL CP Module was designed to measure HRQOL dimensions specific to CP. The module is applicable for both child self-report and parent proxy-report. Child self-report versions include ages 5–7 years, 8–12 years, and 13–18 years. Parent proxy
2459
2460
143
Quality of Life in Children with Cerebral Palsy
. Table 143-5 Specific instruments for QOL assessment in children with CP Instrument Versions PODCI
Child
DISABKIDS
Child
CP-QOL
Child
Parent
Parent
Parent PedsQL-CP Child
Parent
Age
Scales and number of items
11–18 Upper Extremity and physical function; Transfers and basic mobility; Sports; Pain; Expectations from treatment; Happiness with physical condition. (55 items) 2–18 4–16 Autonomy; Moods and emotions; Social integration and Social exclusion; Physical limitations; Treatment. (37 or 12 items) 4–16 9–12 Friends and family; Participation; Communication; General health; Equipment; Pain. (53 items). The parents form includes also access to treatment and parental health. (66 items) 4–12 5–18 Daily activities; School activities; Movement and balance; Pain and hurt; Fatigue; Eating activities; Speech and communication. (35 items, the toddler’s form – less items) 2–18
> Table
143-5 describes four common questionnaires, which were designed specifically to tap health related quality of life in Children with cerebral palsy. All four instruments have both a child version, for self-report and a parent version. The ages approached by each version and the scales covered by each tool are listed above. PODCI Pediatric Outcomes Data Collecting Instrument; QOL Quality of Life; PedsQL Pediatric Quality of Life Inventory; CP Cerebral Palsy
report versions includes ages 2–4 years (toddler), 5–7 years (young child), 8–12 years (child), and 13–18 years (adolescent). This 35-item’s questionnaire encompasses seven scales: Daily activities (nine items), School activities (four items), Movement and balance (five items), Pain and Hurt (four items), Fatigue (four items), Eating Activities (five items); and Speech and communication (four items) (Toddlers’ questionnaire contains fewer items). Higher scores indicate better HRQOL. The PedsQL 3.0 CP Module, demonstrated good psychometric properties in children with CP (Varni et al., 2006). > CP-QOL – This disease specific questionnaire was developed based on views of children with CP and their parents regarding what the child needs in order to have a good QOL. The parent/proxy version assesses seven domains of QOL including social well-being and acceptance, feelings about functioning, participation and physical health, emotional well-being, access to services, pain and feeling about disability, and family health. The child self-report version assesses all of the above domains except access to services and family health. The parent proxy form (parents of children aged 4–12 year) comprises 66 items and the child selfreport form (9–12 year) comprises 52 items. Both the child’s and the parent-proxy’s forms has demonstrated good psychometric properties (Waters et al., 2007). > DISABKIDS – DISABKIDS is a new family of instruments, developed in Europe for the assessment of HRQOL in Children and adolescents with various chronic conditions. The instruments are available as self- or proxy report forms for three age groups: 4–7, 8–12 and 13–16, in three versions: long form (37 items), short form (12 items) and Smiley Measure.
Quality of Life in Children with Cerebral Palsy
143
These questionnaires can be used to assess HRQOL in patients with any type of chronic medical condition and have specific modules for some common childhood chronic conditions, including cerebral palsy. After completing the validation process (Simeoni et al., 2007) theses instruments will probably have a major role in QOL assessment of children with CP.
4.1
Function and Participation
Over the years, various instruments have been developed for the assessment of different aspects of function in children with CP; their description is beyond the scope of this chapter. Likewise, Participation is another emerging model for the evaluation of children with CP, although related to quality of life this is a different concept that will not be discussed in this chapter.
5
QOL in Children with CP
5.1
Parent’s View
Several studies have found that parent’s perceive the QOL of their children with CP to be lower than the QOL of their healthy siblings or other typically developed kids, regardless of CP severity (> Table 143-6). These findings are consistent across various populations and measurement tools and apply to almost any aspect of QOL (Aran et al., 2007; Arnaud et al., 2008; Shelly et al., 2008; Vargus-Adams, 2006; Wake et al., 2003). Siblings of children with CP had normal QOL regardless of disease severity (Aran et al., 2007).
5.2
Children’s View
In spite of their parents view, and the common assumption that disabled children have a lower quality of life, the opinion of the disabled children themselves as reflected in two recent large studies (Dickinson et al., 2007; Shelly et al., 2008) as well as several smaller studies before (Grue, 2003; Varni et al., 2005; Watson and Keith, 2002), seems to be quite different (> Table 143-7). In the largest study ever, to tap self reported QOL in children with CP (Dickinson et al., 2007), 500 children aged 8–12 years, from six European countries, reported their QOL using KIDSCREEN. Over all, children with cerebral palsy had similar scores to children in the general population in all the KIDSCREEN’s domains except schooling, in which evidence was equivocal, and physical wellbeing, in which comparison was not possible. It should be noticed though, that pain was more common then expected based on the proxy reports and it was associated with lower QOL on all domains.
5.3
Factors Affecting HRQL in Children with CP
Various factors have been found to affect quality of life in children with CP (see > Table 143-8 for details) but most of them are child’s characteristics (like disease severity) that are less treatable.
2461
2462
143
Quality of Life in Children with Cerebral Palsy
. Table 143-6 QOL in children with CP, as perceived by their parents Study
Participants
Instruments
Major outcome
Arnaud et al., 2008 (SPARCLE)
Parents of 818 European children, age 8–12 year, mild to severe CP
Kidscreen
Parents reported lower scores in most domains, compared to the general population. The scores in the physical, but not the psychosocial domains, were lower in more disabled children (see > Table 143-8 for more details)
Vargus-Adams et al., 2005
Parents of 177 children age 5–18; mild to severe CP
CHQ-PF50
Parents reported lower scores in all domains, compared to the general population. The scores in the physical, but not the psychosocial domains correlated with level of disability
Kennes et al., 2002
Parents of 408 children age 5–13; mild to severe CP
HUI-3
Rates of functional limitations were lower for more disabled children but scores in emotion and pain were not correlated with disability
Wake et al., 2003
Parents of 80 children, mild to moderate CP, age 5–18
CHQ-PF 50
Parents reported lower scores compared to the general population in all domains. The scores in the physical, but not the psychosocial domains, correlated with the level of disability
> Table 143-6 summarizes the results of recent four large studies, which assessed the quality of life in children with
Cerebral Palsy, by asking their parents to complete the Kidscreen, CHQ or HUI-3 questionnaires. All four studies found that children with CP have lower QOL scores (that reflect poorer quality of life), compared to the general population in every aspect of quality of life. In the physical domains, children with mild motor disability had higher scores but in the psychosocial domains, the scores were low regardless of motor impairment. SPARCLE Study of Participation of Children with Cerebral Palsy Living in Europe; QOL Quality of Life; CHQ-PF 50 Child Health Questionnaire Parent Form 50 (items); CP Cerebral Palsy; HUI-3 Health Utilities Index – Mark 3
Furthermore, those factors primarily influence the physical aspects of QOL, which usually simply reflect the physical functioning rather than the more variable psychosocial aspects.
6
Level of Motor Disability
Level of motor disability is a proven determinant of QOL in children with CP (Arnaud et al., 2008; Kennes et al., 2002; Liptak et al., 2001; Livingston et al., 2007; Majnemer et al., 2007; Shelly et al., 2008; Vargus-Adams, 2005; Varni et al., 2005; Wake et al., 2003). In all of these studies, the physical scales scores were significantly lower in children with severe motor disability compared to those with mild or moderate disability. However, in most of the studies the degree of motor disability did not affect the social or emotional functioning in a similar manner. In fact, the scores on the psychosocial scales were often lower for children with mild CP than for those with moderate CP. A possible explanation is that children with mild CP measure themselves and are compared by their parents to the healthy population while children with moderate to severe CP are less likely to attend mainstream classes.
Parents of 205 children age 4–12; 53 children age 8–12; mild to severe CP
Parents of 235 children age 2–18; 77 children age 5–18. mild to severe CP
81 youth with mild to moderate CP CHQ-CF87 age 10-13y and 30 typically developing youth age 10–13y
Parents of 95 children age 6–12; 55 children age 6–12; mild to severe CP
Parents of 148 children age 5–18; 69 children age 6–18; mild to severe CP
Shelly et al., 2008
Varni et al., 2006
Bjornson, 2008
Majnemer, 2007
Varni et al., 2005
Parents and children reported lower QOL in all domains, compared to healthy kids with correlation to disease severity. Children reported higher scores than their parents.
Parents and children reported lower QOL in physical domains, compared to healthy kids, but half reported equal scores in the psychosocial domains, regardless of disease severity. Children scores correlated with the parents’ scores but they were higher.
Disabled youth reported lower QOL scores in the physical domains of the CHQ and equal scores in the psychosocial domains, compared to the typically developing youth.
Parents and disabled children reported lower QOL scores, in all the PedsQL domains, compared to typically developed children, with correlation to disease severity. Children reported higher scores than their parents in both tools.
Parents reported all domains of the CP-QOL to be lower in more disabled children. The children themselves reported high psychosocial QOL scores regardless of disability.
Children with varying levels of disability have similar Quality of Life scores. Children rate themselves higher, for most of the PODCI and PedsQL scales, than their parents do and the difference increases with the level of disability.
Children with CP reported similar QOL scores compared to children in the general population in most domains of the Kidscreen, regardless of physical impairment.
Major outcome
quality of life higher (better QOL) compared to their parents and in most of them the QOL scores in the psychosocial domains of the children self reports were equal to those of the general population. SPARCLE Study of Participation of Children with Cerebral Palsy Living in Europe; QOL Quality of Life; PODCI Pediatric Outcomes Data Collecting Instrument; CP Cerebral Palsy; PedsQL 4.0 Pediatric Quality of Life Inventory Version 4.0; CHQ-PF 50 Child Health Questionnaire Parent Form 50 (items); CHQ-CF 87 Child Health Questionnaire Child Form 87 (items)
> Table 143-7 summarizes recent studies of quality of life in children with Cerebral Palsy, that used both child’s self-reports and parent’s reports. In all seven studies, the children rated their
PedsQL 4.0
CHQ-PF50 PedsQL 4.0
PedsQL PedsQL-CP
CP-QOL
PODCI PedsQL 4.0
Parents of 562 children age 4–18; 495 children age 5–18 (PedsQL); 247 children age 11–18 (PODCI); mild to moderate CP
Oeffinger, 2007
Instruments
500 European children, age: 8–12, mild KIDSCREEN to severe CP
Participants
Dickinson, 2007 (SPARCLE)
Study
. Table 143-7 QOL in children with CP, as reported by the children themselves (with comparison to their parents view when applicable)
Quality of Life in Children with Cerebral Palsy
143 2463
2464
143
Quality of Life in Children with Cerebral Palsy
. Table 143-8 Elements affecting QOL in children with CP in a recent large, population based study with 818 participants, using the Kidscreen (Arnaud et al., 2008) Element
Effect
Severity of motor impairment
Domains affected
Negative Physical well being, autonomy Positive
Social acceptance, school environment
Cognitive impairment
Negative Social support
Pain
Negative Physical well being, psychological well being, Self perception
Epilepsy
Negative Social support
Positive
Moods and emotions, self perception
High parental education
Negative Parent relations/home life
Single parent household
Negative Moods and emotions
Speech difficulties
Negative Physical well being
Visual impairment
Negative Physical well being
Hearing impairment, age, gender, parental occupation
No effect Any domain of kidscreen
> Table 143-8 describes the impact of different aspects of CP (right column) on various domains of the QOL (left column), as reflected in this large European study. Positive effect (middle column) means that children with higher disability have higher scores (which reflect better QOL) and Negative effect means that children with higher disability had lower QOL scores. QOL quality of life; CP cerebral palsy
7
Cognitive Impairment
As children with cognitive impairment are usually unable to reliably answer the QOL questionnaires, data regarding the impact of cognitive impairment on QOL is based on parental perception. In children with CP, this data is inconclusive. In an Australian study with 80 participants, CHQ scores were similar for children with and without cognitive delay (Wake et al., 2003) and this was also the finding in the study of McCarthy et al. (2002). Another large study, which used the KIDSCREEN, found that children with CP and cognitive impairment have lower QOL scores in the Social support domain but higher scores in the Moods and emotions and Self-perception domains, compared to children with CP and normal cognition (Arnaud et al., 2008).
8
Other Co-Morbidities
Parents of children with CP and epilepsy reported lower QOL on the Self-esteem and Family Cohesion scales of the CHQ (Wake et al., 2003) compared to parents of children with CP per se, the same pattern was found in the much larger study of Arnaud et al. (2008), for the social support domain of the Kidscreen. Other co-morbidities of CP including Learning, behavioral, vision, hearing and communication difficulties were occasionally reported to affect a single, related aspect of QOL in these children (Arnaud et al., 2008; Majnemer et al., 2007).
Quality of Life in Children with Cerebral Palsy
9
143
Pain
Chronic pain correlates with reduced social contacts and activities with peers (Houlihan et al., 2004; Kennes et al., 2002) and with lower scores in the Physical well being, Psychological well being and Self perception (Arnaud et al., 2008) in the parent’s reports. In the child’s reports, pain was the most consistent influencing factor on QOL (Dickinson et al., 2007; Shelly et al., 2008).
10
Socioeconomic Status
As oppose to typically developed children and children with other chronic conditions, there is no evidence that disabled children from lower socioeconomic background have lower QOL scores compared to disabled children from higher socioeconomic background (Arnaud et al., 2008; Majnemer et al., 2007).
11
Child Characteristics (Age, Gender)
Many studies demonstrated that age and gender are not correlated with QOL in children with CP (Arnaud et al., 2008; Majnemer et al., 2007), However, such a correlation was found in some smaller studies (Bottos et al., 2001) and for other chronic conditions.
12
Parenting Style
In a recent study, the QOL of children with CP and their siblings were measured along with the parenting style of their mother (Aran et al., 2007). For the children with CP, autonomy allowing parenting style was found to correlate with higher CHQ scores in both the physical and the psychosocial domains. Accepting parenting style also positively correlated with the psychosocial domain. The impact of parenting style was greater than factors, such as severity of illness, IQ, SES and anxiety level. For the siblings there was no such a correlation. As studies suggest that both the physical and psychosocial well being of children with CP deteriorate with the transition from youth to adulthood (Bottos et al., 2001), autonomy allowing parenting style during childhood may prepare children with CP for more independent lives as adults.
13
Summary and Recommendations
Good quality of life is the main goal for all children: disabled or typically developed. The World Health Organization defined QOL as ‘‘an individual’s perception of their position in life in the context of the culture and value systems in which they live, and in relation to their goals, expectations, standards and concerns’’ (WHO, 1995) and naturally, this should be reported by the individual himself. However, when dealing with children, the parent’s opinion must be considered as well, especially for young or mentally impaired children.
2465
2466
143
Quality of Life in Children with Cerebral Palsy
Several studies on QOL in children with CP revealed the dismal fact that parents perceive the QOL of their children with CP (even mild CP), to be lower in almost any aspect, compared to their healthy siblings and the general population. QOL scores in the physical domains correlate well with the level of disability, but the scores in the psychosocial scales are low regardless of the impairment severity. This finding emphasizes the importance of factors, other than the level of disability that may affect QOL in these kids. In two recent studies, family variables as Parenting style and Family functioning were found to be important factors affecting the psychosocial aspects of QOL of children with CP (Aran et al., 2007; Majnemer et al., 2007). Autonomy allowing and accepting parenting styles, in contrast to controlling and rejecting parenting styles, were reflected in improved mental health, higher self-esteem, better behavior and less social and emotional limitations. These findings can and should be translated to implementation of family interventions, particularly those focusing on parenting style, early in the disease course. As other potentially treatable factors probably affect the psychosocial aspects of QOL in children with CP, any effort should be made to identify and treat those factors.
Summary Points Cerebral Palsy (CP) is a term that describes various motor problems that their cause is a
permanent and stable brain damage that happened in early life. CP is the most common cause of motor disability in childhood. The treatment of CP is supportive (there is no cure for CP) and is often long term and complex. The exact treatment regimen depends on the specific type of CP and usually includes physical and occupational therapy, special education, orthopedic surgery, and assistive devices. Considering the supportive rather than curative nature of the various treatment modalities, it is now well accepted that improving quality of life should be the main treatment goal for children with CP and specific assessment of QOL should be an integral part of any outcome measurement. As studies on QOL, are highly subjective in nature and as such have many inherent limitations, a combination of well valid, parent-based and child-based questionnaires as well as generic and disease specific tools is required. Parents of children with CP usually report lower quality of life scores, for their disabled children, in every aspect of the quality of life and regardless of disease severity (even parents of children with mild CP reports lower scores). Parents of children with severe impairment often reported better quality of life in the psychosocial domains compared to the reports for children with mild impairment. As oppose to typically developed children, disabled children usually report higher quality of life scores compared to their parents. Children with CP, usually rate their quality of life in the emotional and social domains, equal to their typically developed peers. The findings listed above, suggest that children with cerebral palsy can adapt well to their activity limitations and may have satisfactory quality of life despite of significant deficits. These findings also mean that factors other than the impairment severity may have a major influence on QOL in disabled children.
Quality of Life in Children with Cerebral Palsy
143
The consensus that improving QOL is an important treatment goal in children with CP mandates measures and treatments that enhance this goal.
For example: the recent findings that family variables as Parenting style and Family functioning have major impact on the psychosocial aspects of QOL of children with CP, can and should be translated to implementation of family interventions, early in the disease course.
References American Academy of Orthopedic Surgeons. (1997). Pediatric Outcomes Data Collection Instrument version 2.0. In: American Academy of Orthopedic Surgeons. Boston, Massachusetts. Aran A, Shalev RS, Biran G, Gross-Tsur V. (2007). J Pediatr. 151(1): 56–60, e1. Arnaud C, White-Koning M, Michelsen SI, Parkes J, Parkinson K, Thyen U. (2008). Pediatrics. 121(1): 54–64. Bastiaansen D, Koot HM, Ferdinand RF, Verhulst FC. (2004). J Am Acad Child Adolesc Psychiatry. 43: 221–230. Bjornson KF, Belza B, Kartin D, Logsdon RG, McLaughlin J. (2008). Arch Phys Med Rehabil. 89 (1): 121–127. Bjornson KF, McLaughlin JF. (2001). Eur J Neurol. 8(Suppl 5): 183. Bottos M, Feliciangeli A, Sciuto L, Gericke C, Vianello A. (2001). Dev Med Child Neurol. 43: 516–528. Bullinger M, Ravens-Sieberer U. (1995). Revue Europeenne de Psychologie Appliquee. 45: 245–254. Connolly MA, Johnson JA. (1999). Pharmacoeconomics. 16: 605–625. Cremeens J, Eiser C, Blades M. (2006). Health Qual Life Outcomes. 4: 58. Daltroy LH, Liang MH, Fossel AH, Goldberg MJ. (1998). J Pediatr Orthop. 18: 561–571. De Civita M, Regier D, Alamgir AH, Anis AH, Fitzgerald MJ, Marra CA. (2005). Pharmacoeconomics. 23: 659–685. Dickinson HO, Parkinson KN, Ravens-Sieberer U, Schirripa G, Thyen U, Arnaud C, et al. (2007). Lancet. 369(9580): 2171–2178. Eiser C, Morse R. (2001a). Qual Life Res. 10: 347–357. Eiser C, Morse R. (2001b). Health Technol Assess. 5: 1–157. Grue L. (2003). Dev Med Child Neurol. 45(October Suppl): 8 (abstr). Goodwin DA, Boggs SR, Graham-Pole J. (1994). Psychol Assess. 6: 321–328. Harding L. (2001). Clin Psychol Psychother. 8: 79–96. Havermans T, Vreys M, Proesmans M, De Boeck C. (2006). Child Care Health Dev. 32: 1.
Houlihan CM, O’Donnell M, Conaway M, Stevenson RD. (2004). Dev Med Child Neurol. 46: 305–310. Kennes J, Rosenbaum P, Hanna SE, Walter S, Russell S, Raina P, Bartlett D, Galuppi B. (2002). Dev Med Child Neurol. 44: 240–257. Koman LA, Smith BP, Shilt JS. (2004). Lancet. 363: 1619–1631. Landgraf J, Abetz LN. (1996). In: Spilker B. (ed.) Quality of Life and Pharmacoeconomics in Clinical Trials. Lippincott-Raven, Philadelphia, PA, pp. 793–802. Liptak GS, O’Donnell M, Conaway M, Chumlea WC, Wolrey G, Henderson RC. (2001). Dev Med Child Neurol. 43: 364–370. Livingston MH, Rosenbaum PL, Russell DJ, Palisano RJ. (2007). Dev Med Child Neurol. 49(3): 225–231. Majnemer A, Mazer B. (2004). Semin Pediatr Neurol. 11: 11–17. Majnemer A, Shevell M, Rosenbaum P, Law M, Poulin C. (2007). J Pediatr. 151(5): 470–475, 5 e1–3. McCarthy ML, Silberstein CE, Atkins EA, Harryman SE, Sponseller PD, Hadley-Miller NA. (2002). Dev Med Child Neurol. 44(7): 468–476. Oeffinger D, Gorton G, Bagley A, Nicholson D, Barnes D, Calmes J. (2007). Dev Med Child Neurol. 49(3): 172–180. Oeffinger DJ, Tylkowski CM, Rayens MK, Davis RF, Gorton GE, 3rd, D’Astous J. (2004). Dev Med Child Neurol. 46: 311–319. Palisano R, Rosenbaum P, Walter S, Russell D, Wood E, Galuppi B. (1997). Dev Med Child Neurol. 39: 214–223. Ravens-Sieberer U, Gosch A, Rajmil L. (2005). Expert Rev Pharmacoeconomics Outcomes Res. 5: 353–364. Ronen GM, Streiner DL, Rosenbaum P. (2003). Epilepsia. 44: 598–612. Russell KM, Hudson M, Long A, Phipps S. (2006). Cancer. 106: 2267–2274. Shelly A, Davis E, Waters E, Mackinnon A, Reddihough D, Boyd R. (2008). Dev Med Child Neurol. 50(3): 199–203. Simeoni MC, Schmidt S, Muehlan H, Debensason D, Bullinger M, Group tD. (2007). Qual Life Res. 16 (5): 881–893.
2467
2468
143
Quality of Life in Children with Cerebral Palsy
Theunissen NC, Vogels TG, Koopman HM. (1998). Qual Life Res. 7: 387–397. Vargus-Adams J. (2005). Arch Phys Med Rehabil. 86: 940–945. Vargus-Adams J. (2006). Dev Med Child Neurol. 48(5): 343–347. Varni JW, Burwinkle TM, Berrin SJ, Sherman SA, Artavia K, Malcarne VL. (2006). Dev Med Child Neurol. 48(6): 442–449. Varni JW, Burwinkle TM, Sherman SA. (2005). Dev Med Child Neurol. 47: 592–597. Varni JM, Limbers CA, Burwinkle TM. (2007). Health Qual Life Outcomes. 3: 1. Viehweger E, Robitail S, Rohon MA, Jacquemier M, Jouve JL, Bollini G. (2008). Ann Readapt Med Phys. 51(2): 129–137.
Wake M, Salmon L, Reddihough D. (2003). Dev Med Child Neurol. 45: 194–199. Waters E, Davis E, Boyd R. (2007). Dev Med Child Neurol. 49: 49–55. Waters E, Stewart-Brown S, Fitzpatrick R. (2003). Child Care Health Dev. 29: 501–509. Watson SMR, Keith KD. (2002). Ment Retard. 40: 304–312. White-Koning M, Arnaud C, Dickinson HO, Thyen U, Beckung E, Fauconnier J. (2007). Pediatrics. 120(4): e804–e814. World Health Organization. (1993). Geneva, Switzerland: Division of Mental Health, World Health Organization, Geneva, Switzerland. World Health Organization. (1995). Soc Sci Med. 41: 1403–1409.
144 Quality of Life Measures in Children with Cancer C. H. Yeh . Y.-P. Kung . Y.-C. Chiang 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2470
2 2.1 2.2 2.3
The Criteria for Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2471 Conceptual Definitions of HRQL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2471 Disease Specific and Generic Instruments to Measure HRQL . . . . . . . . . . . . . . . . . . . . . 2474 Respondent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2477
3 3.1 3.2 3.3
Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2477 Problems with the Conceptual Definition of HRQL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2477 Proxy Agreement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2478 Developmental Concerns of HRQL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2479
4
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2479 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2480
#
Springer Science+Business Media LLC 2010 (USA)
2470
144
Quality of Life Measures in Children with Cancer
Abstract: Health related quality of life (HRQL) in pediatric oncology patients has gained much attention in the past three decades due to the advances in medical technology which has dramatically increased survival rates. The purpose of this review is to identify, describe and critique HRQL measures developed or applied in pediatric oncology patients. Considerable progress has been made in the development of HRQL for pediatric oncology patients in the past two decades. Measures reviewed in this chapter have been examined for many factors: multidimensional constructs, patient’s subjective assessment, as well as supplemental information provided by parents or healthcare providers. Evidence of strong psychometric qualities has also been provided for the studies reviewed here. Still, there are problems regarding HRQL measures for pediatric oncology patients which need to be solved, including additional consensus of conceptual and operation definitions of HRQL. Issues related to concordances of proxy agreement remain. Few HRQL measures have taken developmental issues into account, especially for younger children. Interpretation of quantitative HRQL may ignore the qualitative meaning to children’s perception of HRQL. List of Abbreviations: ARM, adolescent resilience model; CHI, child health and illness; CHQ, child health questionnaire; CHRIs, child health rating inventories; FS-II, functional status II; HCP, health care professionals; HRQL, health related quality of life; HUI, health utilities index; PCQL-32, pediatric cancer quality of life inventory; PEDQOL, quality of life in children and adolescents with cancer; PedsQL™, pediatric quality of life inventory™; POQOLS, pediatric oncology quality of life scale; PPSC, play performance scale for children; QOLCC, quality of life for children with cancer; QWB, quality of well being; WHO, World Health Organization
1
Introduction
The survival rate for pediatric oncology patients has dramatically increased due to the advances in medical technology over the past two decades. The overall 5 year survival rate for all forms of pediatric cancer has improved, and is now up to 80% for some types of cancer (Adamson et al., 2005). Along with the improved outcomes for pediatric oncology, much attention has been directed toward health related quality of life (HRQL) issues for both cancer patients and long term cancer survivors. The co-morbidity of cancer treatment continues to be a major factor in providing care for oncology patients further increasing the need for Health Care Professionals (HCP) to attend to quality of life issues. The purpose of this chapter is to review and critique the research that has been published in the area of assessing and measuring HRQL for pediatric oncology patients and their families. HRQL has become an important outcome indicator for assessing the effectiveness of these recent advances in cancer treatment (Hinds et al., 2006; Pickard et al., 2004). Consensus in the field of HRQL has been achieved for both the subjective and objective perspectives of the conceptual and theoretical underpinnings (Cella, 1994; Eiser, 2004; Hinds et al., 2006; Pickard et al., 2004). In the past three decades, a significant amount of literature has been published in the area of HRQL associated with pediatric oncology patients. The objective of this review is to identify, describe and critique HRQL measures developed or applied in pediatric oncology patients. Measures related to HRQL for pediatric oncology patients, have especially burgeoned over the past 20 years. The organization of this chapter is as follows: after a brief historical overview of HRQL, the current consensus of conceptual definitions for HRQL will be presented, the remaining sections of the chapter will describe the methods used to select
Quality of Life Measures in Children with Cancer
144
and review the most rigorous research studies that have developed and tested measures for assessing HRQL in pediatric oncology patients and their families. Generic and disease specific HRQL instruments used for pediatric oncology patients will be described, issues associated with self-report and proxy measurements will be addressed as will concerns regarding reliability and validity of measures. The chapter will end with conclusions for further research.
2
The Criteria for Selection
The following section describes the process used for selecting the specific research studies that will be reviewed and critiqued in this chapter. An initial computer search (MEDLINE) was conducted using the headings ‘‘quality of life,’’ ‘‘children,’’ and ‘‘cancer.’’ The bibliographies of other related review articles (Eiser and Morse, 2001; Hinds et al., 2006; Nathan et al., 2004; Pickard et al., 2004; Varni et al., 2007) were also reviewed and each selected article provided a partial cross-check on the completeness of the library-based search. Only articles published in the English-language, and in nationally circulated publications were included in the review. When the titles were appropriate, the articles were retrieved and examined to determine if they met the selection criteria for this review. First, the article had to qualify as a research report. To qualify as a research report, the manuscript had to reflect the five major research elements recommended by Duffy (1985), which include: evidence of problem formation, data collection, evaluation of data points, analysis and interpretation, and presentation of results. The scientific merit of each study was reviewed according to criteria elaborated by Duffy (1985), according to title, abstract, problem, literature review, methodology, data analysis and discussion. The following is based on the searches and criteria just described. The introduction of the concept of quality of life in medical research begin in the definition delineated by the World Health Organization (WHO) (1947) as ‘‘the complete state of physical, mental and social well being and not merely the physical, mental, and social dimensions (p.29).’’ Since that time, over 53,836 publications have appeared related to quality of life. Not surprising, only 0.5% (n = 285) of the publications identified are related to children (with cancer). Using a weighting for the search terms ‘‘quality of life,’’ ‘‘children,’’ and ‘‘cancer’’ resulted, only 156 publications that specifically focused on quality of life in children. In the 156 articles examined, the majority referred to theoretical or conceptual work (n = 95, 60%), empirical quality of life research in children with specific chronic conditions (n = 48, 30%), while the testing of assessment instruments (quality of life in children) or empirical quality of life research in general was less prominent.
2.1
Conceptual Definitions of HRQL
Until recently there was no consensus on the conceptual definitions associated with HROL. That has now changed and the accepted definitions are presented in > Table 144-1 for conceptual definition and > Table 144-2 for theoretical definition. In their article Hinds and Haase (1998), highlight function-based and meaning-based models, as well as Substantive models for quality of life in pediatric oncology patients. Each of these has contributed to the development of different scientific knowledge. Function-based models focus primary on objective assessment of one’s functioning while subjective based models examine the meaning of the illness and the meaning of quality of life to the individual. A typical definition of
2471
2472
144
Quality of Life Measures in Children with Cancer
. Table 144-1 Conceptual definition of quality of life for pediatric oncology patients References
Conceptual definition
Dimensions
Hinds et al. (2004)
An overall sense of well-being based on Symptoms, usual activities, social being able to participate in usual activities; and family interactions, health to interact with others and feel cared about; status, mood, meaning of being ill to cope with uncomfortable physical, emotional, and cognitive reactions; and to find meaning in the illness experience
Woodgate and Degner, (2003)
Patient’s changing perceptions of themselves across the trajectory of treatment were a major influence on their QOL or well-being. An inductively identified process labeled by the researchers as ‘‘keeping the spirit alive’’ represented the patients’ efforts to live with cancer. Cancerrelated symptoms precipitated changes in self-perceptions; the same symptoms were viewed as part of the entire cancer experiences, in part, represented by a process labeled ‘‘getting through all the rough spots.’’ Cancer was described as profoundly affecting adolescents’ sense of self, particularly with physical changes
Yeh et al. (2004a)
QOL is defined as the impact of disease and treatment on the child’s appraisal and satisfaction of functioning as measured in the following scales: (1) physical function, defined as functional status in the activities of daily living; (2) psychological dysfunction, defined as the degree of emotional distress; (3) social function, defined as interpersonal functioning in peer/school relationships; (4) treatment/disease-related symptoms, defined as anxiety and worry about the illness and treatments; and (5) cognitive function, defined as cognitive performance in problem solving
Physical function, psychological dysfunction, social function, treatment/disease-related function and cognitive function
function-based quality of life is ‘‘a multidimensional construct’’ that includes but is not limited to social, physical, and emotional functioning (Varni et al., 1998). Most of the measures represented in > Table 144-2 are based on a function-based model. Hinds and Haase (1998) believe the function-based definition may not detect patients’ experiences which are meaningful to them. For example, adolescents may choose to answer an item asking if he or she is able to see friends at school with a ‘‘no’’ when in fact they could see friends but choose not to because they have lost their hair and choose to stay home due to their
Quality of Life Measures in Children with Cancer
144
. Table 144-2 Theoretical definition of health related quality of life for pediatric oncology patients References
Models
Theoretical definition
Haase et al. (1999)
Adolescent resilience model
Illness-related risk (i.e., symptom distress and uncertainty), social integration (i.e., relationships with peers and healthcare providers), and spiritual perspective meaning, positive coping, defensive coping, and family atmosphere, resilience and self-transcendence. The process of identifying or developing resources and strengths to flexibly manage stressors to gain a positive outcome, a sense of confidence, mastery and self-esteem
Hinds and Martin, (1988)
Self-sustaining model for adolescent
It is defined as a natural progression adolescents experiencing serious health threats move through to comfort themselves and to achieve competence in resolving health threats
Woodgate (1999)
Resiliency model
This model includes stressors or risk situations (i.e., cancer diagnosis, hair lose), vulnerability factors or protective factors, and outcomes (adaptation and maladaptation as a continuum)
Taieb et al. (2003)
Post-traumatic stress disorder
The symptoms of PTSD are grouped into three categories: reexperiencing (recurrent distressful memories or dreams), avoiding or numbing (shunning talking or thinking about the event), and increased arousal (behavioral or cognitive problems, sleep disturbances)
Hinds et al. (2005)
Pediatric QOL at end-of_life model
This model depicts the transition from curative to end-of-life care and reflects a dual focus on the QOL of the terminally ill child or adolescent and that of the family
Nuss et al. (2005)
Relational decision making at end It depicts the centrality of three interdependent of life in pediatric oncology perspectives in end-of-life decision making: those of the child or adolescent, the parent or guardian, and the healthcare provider. Relational decision making is depicted as being influenced by communication skills, competence, emotions, faith, and hope
Table 144-2 was adapted from Hinds and Haase (1998) and updated
appearance and this alters their attendance at school, rather than the severity of their illness. Thus their quality of life score may be misinterpreted. Hinds (1990) proposed a model of environmental influence on quality of life: that the internal environment (i.e., the child’s feelings about him or herself), the immediate environment (i.e., significant others such as family or health care providers) and the institutional
2473
2474
144
Quality of Life Measures in Children with Cancer
environment (i.e., financial support) all influence quality of life. The Adolescent Resilience Model (ARM) also indicates that the individual, family, and societal characteristics directly influence quality of life (Haase et al., 1999). The different definitions of health-related quality of life in the current literature have resulted in an international consensus about the components of quality of life. An operational definition of the term has evolved. Varni et al. have recently identified this consensus definition as: health-related quality of life is viewed as a psychological construct which describes the physical, mental, social, psychological and functional aspects of well-being and function from the patient perspective (Varni et al., 2007). This operational definition stresses the multidimensionality of the quality of life concept as well as the relevance of patients’ self-report.
2.2
Disease Specific and Generic Instruments to Measure HRQL
Two categories of instruments have been used to measure HRQL (> Table 144-3). This table is adapted from Pickard et al. (2004) and updated. Those developed specifically for children with cancer and generic instruments designed to measure quality of life in other chronic illnesses. These are separated and identified in > Table 144-2. Generic measures refer to instruments used to assess functional status or quality of life across disease groups which are not designed specifically for the assessment of cancer patients, but they have been used in some of the reports that will be presented here, such as Child Health and Illness (CHI), Child Health Questionnaire (CHQ), Functional status II (FS-II), Quality of Well Being (QWB), Health Utilities Index (HUI) Mark 2 (HUI2) and Mark 3 (HUI3), Play Performance Scale for Children (PPSC). The PPSC and FS-II were designed to measure physical functioning for pediatric oncology patients. The Health Utilities Index (HUI) Systems (HUI 2and HUI 3) (Feeny et al., 1996, 1999) have been used extensively to examine health status in pediatric cancer patients. The HUI is a multi-attribute instrument, with objective scoring criteria and brevity (Feeny et al., 1996, 1999). The Child Health Questionnaire (CHQ) (Landgraf et al., 1998) is designed to reveal a multidimensional profile of OQL for children, which measures various concepts not normally included in adult assessment tools, such as self-esteem, behavior and, the impact of the children’s QOL on the family and parents. Measurements developed to assess cancer-specific HRQL include the Pediatric Cancer Quality of life Inventory (PCQL-32) (this measure has also been expanded to the Pediatric Quality of life Inventory™, PedsQL™), Miami Pediatric Quality of Life Questionnaire, the Pediatric Oncology Quality of Life Scale (POQOLS), the Child Health Rating Inventories (CHRIs), the Quality of Life in Children and Adolescents with Cancer (PEDQOL), the Royal Marsden Hospital Pediatric quality of life questionnaire, and the Quality of Life for Children with Cancer (QOLCC). Psychometric testing with establish reliability and validity of these measures are well documented. Disease-specific HRQL instruments are believed to enhance sensitivity for health (physical and disease) domains specifically for cancer. Cancer-specific HRQL measures described above allow us to evaluate psychological factors in addition to physical functioning and have sound psychometric qualities. Both the Child Health Rating Inventories (CHRIs) (Parsons et al., 1999) and the Pediatric Cancer Quality of Life Inventory-32 (PCQL-32) (Varni et al., 1998a) employ parent/child proxy to assess QOL and report variability in different domains of QOL. While these studies represent the significant advancement in QOL research on children with
144
Quality of Life Measures in Children with Cancer
. Table 144-3 Health-related quality of life measures in pediatric oncology patients Instrument
Domains/items
Respondent
Items
Age
Child Health and Illness Profile – Adolescent Edition (CHIP-AE) (Starfield et al., 1995)
Activity, disorders, discomfort, satisfaction with health, achievement of social roles, resilience
Child
153
11–17
Child Health Questionnaire (CHQPF50) (Landgraf et al., 1998)
Physical functioning, role/social physical/ emotional behavior, general health, parental time/emotional impact, self-esteem, mental health, general behaviors, family activities, family cohesion
Parent and child
50
5–16
Functional status II (R) (Stein and Jessop, 1990)
Physical, psychological, social
Long: 43
0–16
KINDL (Ravens-Sieberer and Bullinger, 1998)
Psychological well-being, social relationships, physical function, everyday life activities
Generic
Parent
Short: 14 Child
40
10–16
Parent proxy Parent (interview) (4–7 years)
Sixteen-dimensional health-related measures (16-D, 17-D) (Apajasalo et al., 1996a,b)
Mobility, vision, hearing, breathing, sleeping, eating, elimination, speech, mental function, discomfort and symptoms, school and hobbies, friends, physical appearance, depression, distress, vitality (may varies due to different items included)
Child
16
12–15
Health Utilities Index (HUI)-Mark 2 and 3 (Barr et al., 1994; Feeny et al., 1996)
Mark 2: sensation, mobility, emotion, pain, fertility, cognition and self-care
Parent, child, physician
Interviewer:40;
6
Mark 3: vision, hearing, speech, ambulation, dexterity, emotion, cognition and pain
Self-complete: 15
2475
2476
144
Quality of Life Measures in Children with Cancer
. Table 144-3 (continued) Instrument
Domains/items
Respondent
Items
Age
Child Health Rating Inventories (CHRIs) (Parsons et al., 1999)
Physical function, role function, mental health, overall quality of life, social/personal resources, cognitive function and energy
Child
30
5–18
Quality of well-being scale (QWB) (Bradlyn et al., 1993)
Mobility, physical functioning, social activity, current symptomatology
40
0–18
Behavioral, affective, and somatic experiences scale (BASES) (Phipps et al., 1994)
Somatic distress, compliance, mood/ behavior, interactions, activity
38
2–20
Child
34
Adolescent
Parent
1–16
1
Child, adolescent
8–12 (child)
Child: 84
Parent
13–18 (adolescent)
Adolescent: 87
Child
8–18
34
Parent Doctor
Parent Child Adolescent Nurse (BASE-N) Parent (BASE-P) Child (BASE-C)
Perceived illness experience (PIE) (Eiser et al., 1995)
Physical functioning (symptoms, functional disability, and restrictions), psychological functioning (symptoms)
Play performance Scale Level of activity for children (Lansky et al., 1987) Cancer specific measures Pediatric cancer quality of life inventory (PCQL), PCQL-32, PedsQL™ (Varni et al., 1998a, 1999)
Disease/treatment related symptoms, physical, psychological, social, and cognitive functioning
PEDQOL: quality of life in children and adolescents with cancer (Calaminus et al., 2000)
Physical functioning, autonomy, emotional functioning, cognition, social functioning/friends, social functioning/family, body image
Miami Pediatric Quality of life questionnaire (Armstrong et al., 1999)
Social competence, emotional stability, self-competence
Parent
1–18
40
Pediatric oncology quality of life scale (POQOLS) (Goodwin et al., 1994)
Physical functions and role restriction, emotional distress, reaction to current medical treatment
Parent
Not stated
21
Parent
144
Quality of Life Measures in Children with Cancer
. Table 144-3 (continued) Instrument Royal Marsden Hospital Pediatric quality of life (Watson et al., 1999)
Domains/items
Respondent
Items
Age
Functional status, global quality of life, physical symptoms, emotional status, social functioning, cognitive functioning, behavioral problems, school/educational progress
Parent
Not stated
78
Child
7–18
34
Quality of life for children Physical, social, cognitive with cancer (QOLCC) disease/treatment, and (Yeh et al., 2004a) psychological functioning
Adolescent Parent
Table 144-3 was adapted from Pickard et al. (2004) and updated
cancer, most have been conducted in western countries. Only one measure (Quality of Life in Children with Cancer, QOLCC) originated in Taiwan, is published in English, and was developed using a qualitative approach to explore life experiences of Taiwanese children with cancer. Based on the items from PCQL-32, QOLCC is a multidimensional measure to assess HROL (Yeh et al., 2004a,b).
2.3
Respondent
Information collected regarding HRQL includes patient, parent or healthcare providers. Among generic measures (n = 11), 4 includes respondent from patient and parent assessment, 3 from child only, 4 from parent only, and 1 from healthcare provider as well. Among cancerspecific measures (n = 6), 3 include responses from patient and parent assessment, and 3 from parent only.
3
Discussion
Considerable progress has been made in the development of HRQL for pediatric oncology patients in the past two decades. Measures reviewed in this article have considered many factors: multidimensional constructs, patient’s subjective assessment, as well as supplemental information provided by parents or healthcare providers. Evidence of strong psychometric qualities has also been provided for the studies reviewed here. Still, there are problems regarding HRQL measures for pediatric oncology patients which need to be solved.
3.1
Problems with the Conceptual Definition of HRQL
Several conceptual definitions and models of HRQL for pediatric oncology patients have been developed or proposed. The development of a valid measure of a theoretical construct is the cornerstone of QOL research, which ensures that the construct under investigation is
2477
2478
144
Quality of Life Measures in Children with Cancer
measurable in an applied setting (Hendrick and Hendrick, 1986). The importance of accuracy in measurement has been emphasized in the discipline of nursing. However, this may not be true in other disciplines. For example, many HRQL measures used in the literature are not supported by conceptual definitions, but only include the domains assessed in the measures. The lack of theoretical development for QOL concepts in pediatric cancer patient research may result in false conclusions, and subsequently, improper clinical interventions (Hinds and Haase, 1998). Some conceptual definitions are proposed, but limited HRQL measures are grounded under their conceptual definitions. For example, the meaning-based model of HRQL is very important because HRQL is a subjective construct (Hinds and Haase, 1998). However, few HRQL measures have considered the meanings from patients’ perspective. Indeed, assessment of qualitative HRQL in addition quantitative measures has its challenge of how to collect both qualitative and quantitative data simultaneously. Triangulated data collection and data analysis are sometimes employed in nursing research in order to capture the meaning for the whole individual. Nursing focuses on caring and sensitivity for the individual and the holistic nature of that care to the point of also including the family and understanding the meaning of an event to all of them. Such methodological challenge of triangulation of data collection and analysis still limits many empirical studies.
3.2
Proxy Agreement
One of the challenges of a valid and reliable measure for children with cancer is that the cognitive capacity for self-evaluation changes with the normal process of maturation. One of the overall definitions of HRQL assessment as a subjective construct should be from patients themselves. Assessing the QOL of children is complex due to developmental differences in understanding the content being measured (Landgraf and Abetz, 1996). A few HRQL measures have been developed considering the perspective of children. Different versions of the same measures have been developed especially for children of different ages; child and adolescent self-report versions are available for both the PCQL-32 (Varni et al., 1998a,b) and the QOLCC (Yeh et al., 2004a,b). Proxy reports may be the only available source of data when children are too young to understand the content of self-report measures or too sick to answer a questionnaire. However, inconsistencies between children’s self-reports and parentproxy reports have been reported frequently in the literature (Chang and Yeh, 2005; Cotterill et al., 2000; Feeny et al., 1999; Varni et al., 1998a). Proxy reports tend to be more valid on objective assessments (such as hyperactivity, acting out) than on subjective assessments (depression and anxiety) (Varni et al., 1999). Differences between the degree of concordance between parents and young children and parents and adolescents have been established (Cotterill et al., 2000). Parents may have limited knowledge regarding children’s HRQL, however, parents can still provide useful information (Jokovic et al., 2004). The understanding of differences between patients and parent proxy report may also be limited by cross-sectional study design. Agreement of HRQL among patient and parent may change during a course of cancer treatment. In a study examining the agreement on quality of life measures between children’s self-reports and parent-proxy reports at different points in time, parent proxy consistently overestimated QOL. They tended to report better QOL (lower score of QOLCC) than did children (in both the on- and off-treatment groups) on baseline and
Quality of Life Measures in Children with Cancer
144
6-month follow-up assessments, with the one exception being the cognitive subscale at 6-month follow-up (Yeh et al., 2005). In a review study of proxy agreement in research conducted between 1980 and 1999 on QOL in pediatric populations, the authors concluded that before implementing QOL measures, the relationship between child and proxy rating must be clarified (Eiser and Morse, 2001).
3.3
Developmental Concerns of HRQL
HRQL for pediatric oncology patients include a broad range of age and developmental concerns up to age 18. It has been emphasized that HRQL should be able to detect the issues related to growth and development for children and adolescents. However, few measures have been designed to address developmental concerns. Some measures have been developed and tested using separate version of HRQL for children and adolescents (i.e., PCQL-32, QOLCC) (Varni et al., 1998a; Yeh et al., 2004a,b). The items included in these two versions are only different in wording levels regarding different cognitive maturity, but there are no differences in actual item content for children and adolescents. Issues regarding HRQL for children and adolescent may be different. For example, adolescents may focus on their peer relationships and career development but children are more concerned about their limited ability to actively play with others (Yeh, 2001). How the HRQL measures consider developmental changes across childhood is a very important issues in assessing children. In addition, another important issue is whether the same HRQL measure used to collect data when a patient is newly diagnosed is appropriate for patients during follow up courses of cancer treatment or even for survivors.
4
Conclusion
It is evident that the development and testing of HRQL measures for pediatric oncology patients have made considerable progress over the past 20 years. Measures of HRQL specific for pediatric oncology patients have been developed and tested with established reliability and validity. Agreements of proxy report are also examined. Nevertheless, there are still issues in measures of HRQL that need to be solved. First, the development of conceptual definitions for children and adolescents’ perspectives of HRQL need to clearly defined according to cognitive differences. The perceptions of HRQL across the cancer treatment trajectory for survivors need to be addressed. Rigorous qualitative studies can be employed to explore children and adolescent’s perception of HRQL. Second, the conceptual and operational definition of HRQL should be explicitly stated and matched with underlying conceptual assumptions. Measures of HROL should be tested in cross-sectional studies as well to examine their appropriateness for longitudinal follow up studies. Third, a measure of HRQL for young children (younger than 6 years) needs to be developed. Due to the limited cognitive development of young children, few empirical studies report the HRQL in younger children. In addition patient self-report data, parent proxy assessment should be used as complementary information to increase understanding of HRQL. Finally, HRQL measures need to consider the meaning of HRQL measures to children and adolescents. How we can interpret HRQL correctly is important for health care providers.
2479
2480
144
Quality of Life Measures in Children with Cancer
Summary Points The need for accurate measurement of Health related quality of life (HRQL) for assess-
ment and evaluation of pediatric cancer treatment is essential due to the dramatic increase in the survival rate. Several conceptual definitions associated with HRQL have been proposed but many HRQL measures still lack explicit conceptual definitions. The meaning of HRQL measures may be misinterpreted and more work remains to solve how the interpretation of quantitative of HRQL can capture the meaning of HRQL to pediatric oncology patients. Differences between the degree of concordance between parents and young children and parents and adolescents have been established but are still limited by the use of cross sectional study designs. Developmental issue of HRQL measures have been addressed, but few measures have been designed to consider developmental concerns. Due to the limited cognitive development of young children, few empirical studies report the HRQL in younger children.
Acknowledgments This review was supported by a grant to Dr. Yeh from National Health Research Institutes, Taiwan (Grant number: NHRI-EX95–9302PI) and National Science Council (NSC94–2314-B182–014). Special thanks to Dr. Susan Jay for the manuscript editing.
References Adamson P, Law G, Roman E. (2005). Lancet. 365: 753. Apajasalo M, Rautonen J, Holmberg C, Sinkkonen J, Aalberg V, Pihko H, Siimes MA, Kaitila I, Makela A, Erkkila K, Sintonen H. (1996a). Qual Life Res. 5: 532–538. Apajasalo M, Sintonen H, Holmberg C, Sinkkonen J, Aalberg V, Pihko H, Siimes MA, Kaitila I, Makela A, Rantakari K, Anttila R, Rautonen J. (1996b). Qual Life Res. 5: 205–211. Armstrong FD, Toledano SR, Miloslavich K, LackmanZeman L, Levy JD, Gay CL, Schuman WB, Fishkin PE. (1999). Int J Cancer Suppl. 12: 11–17. Barr RD, Pai MK, Weitzman S, Feeny D, Furlong W, Rosenbaum P. (1994). Int Oncol. 4: 639–648. Bradlyn AS, Harris CV, Warner JE, Ritchey AK, Zaboy K. (1993). Health Psychol. 12: 246–250. Calaminus G, Weinspach S, Teske C, Gobel U. (2000). Klin Padiatr. 212: 211–215. Cella DF. (1994). J Pain Symptom Manag. 9: 186–192. Chang PC, Yeh CH. (2005). Psychooncology. 14: 125–134. Cotterill SJ, Parker L, Malcolm AJ, Reid M, More L, Craft AW. (2000). Brit J Cancer. 83: 397–403.
Duffy ME. (1985). Nurs Health Care. 6: 538–547. Eiser C. (2004). Children with Cancer: The quality of life. Lawrence Erlbaum Associates, Mahwah. Eiser C, Havermans T, Craft A, Kernahan J. (1995). Arch Dis Child. 72: 302–307. Eiser C, Morse R. (2001). Arch Dis Child. 84: 205–211. Feeny D, Furlong W, Mulhern RK, Barr RD, Hudson M. (1999). Int J Cancer Suppl. 12: 2–9. Feeny DH, Torrance GW, Furlong WJ. (1996). In: Spilker B (ed.) Quality of Life and Pharmacoeconomics in Clinical Trials. Lippincott-Raven Publishers, Philadelphia, PA, pp. 239–252. Goodwin DAJ, Boggs SR, Graham-Pole J. (1994). Psychol Assess. 6: 321–328. Haase JE, Heiney SP, Ruccione KS, Stutzer C. (1999). Int J Cancer Suppl. 12: 125–131. Hendrick C, Hendrick S. (1986). J Pers Soc Psychol. 50: 392–402. Hinds PS. (1990). Semin Oncol Nurs. 6: 285–291. Hinds PS, Gattuso JS, Fletcher A, Baker E, Coleman B, Jackson T, Jacobs-Levine A, June D, Rai SN, Lensing S, Pui CH. (2004). 13: 761–772.
Quality of Life Measures in Children with Cancer Hinds PS, Burghen EA, Haase JE, Phillips CR. (2006). Oncol Nurs Forumm. 33: 23–29. Hinds PS, Haase JE. (1998). In: King CR, Hinds PS (eds.) Quality of Life: from Nursing and Patient Perspective: Theory, Research, Practice. Jones and Bartlett Publishers International, London, pp. 54–63. Hinds PS, Martin J. (1988). Nurs Res. 37: 336–340. Hinds PS, Oakes LL, Hicks J, Anghelescu DL. (2005). Semin Oncol Nurs. 21: 53–62. Jokovic A, Locker D, Guyatt G. (2004). Qual Life Res. 13: 1297–1307. Landgraf JM, Abetz LN. (1996). In: Spilker B (ed.) Quality of Life and Pharmacoeconomics in Clinical Trials. Lippincott-Raven Publishers, Philadelphia, PA, pp. 793–802. Landgraf JM, Maunsell E, Speechley KN, Bullinger M, Campbell S, Abetz L, Ware JE. (1998). Qual Life Res. 7: 433–445. Lansky SB, List MA, Lansky LL, Ritter-Sterr C, Miller DR. (1987). Cancer. 60: 1651–1656. Nathan PC, Furlong W, Barr RD. (2004). Pediatr Blood Cancer. 43: 215–223. Nuss SL, Hinds PS, LaFond DA. (2005). Semin Oncol Nurs. 21: 125–134. Parsons SK, Barlow SE, Levy SL, Supran SE, Kaplan SH. (1999). Int J Cancer Suppl. 12: 46–51. Phipps S, Hinds PS, Channell S, Bell GL. (1994). J Pediatr Oncol Nurs. 11: 109–117. Pickard AS, Topfer LA, Feeny DH. (2004). J Natl Cancer Inst Monographs. 33: 102–125. Ravens-Sieberer U, Bullinger M. (1998). Qual Life Res. 7: 399–407.
144
Starfield B, Riley AW, Green BF, Ensminger ME, Ryan SA, Kelleher K, Kim-Harris S, Johnston D, Vogel K. (1995). Med Care. 33: 553–566. Stein RE, Jessop DJ. (1990). Med Care. 28: 1041–1055. Taieb O, Moro MR, Baubet T, Revah-Levy A, Flament MF. (2003). Eur Child Adolesc Psychiatry. 12: 255–264. Varni JW, Katz ER, Seid M, Quiggins DJ, FriedmanBender A. (1998a). Cancer. 82: 1184–1196. Varni JW, Katz ER, Seid M, Quiggins DJ, FriedmanBender A, Castro CM. (1998b). J Behav Med. 21: 179–204. Varni JW, Limbers CA, Burwinkle TM. (2007). Health Qual Life Outcomes. 5: 43. Varni JW, Seid M, Rode CA. (1999). Med Care. 37: 126–139. Watson M, Edwards L, Von EL, Davidson J, Day R, Pinkerton R. (1999). Int J Cancer Suppl. 12: 65–70. Woodgate RL. (1999). J Pediatr Oncol Nurs. 16: 35–43. Woodgate RL, Degner LF. (2003). J Pediatr Oncol Nurs. 20: 103–119. World Health Organization. (1947). The constitution of the World Health Organization. WHO Chron 1: 29. Yeh CH. (2001). Nurs Sci Q. 14: 141–148. Yeh CH, Chang CW, Chang PC. (2005). Nurs Res. 54: 354–362. Yeh CH, Chao KY, Hung LC. (2004a). Psychooncology. 13: 161–170. Yeh CH, Hung LC, Chao KY. (2004b). Psychooncology. 13: 171–176.
2481
145 Quality of Life in Healthy and Chronically Ill Icelandic Children: Agreement Between Child’s Self-Report and Parents’ Proxy-Report E. K. Svavarsdottir 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2484 1.1 HRQOL from the Child’s, the Adolescent’s and the Parent’s Perspective . . . . . . . . . . 2485 1.2 Level of Agreement Between Child’s Self-Report and the Parents’ Report on HRQOL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2486 2 2.1 2.2 2.3 2.4
Methodological Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2487 Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2487 Subjects/Sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2488 Instruments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2490 Data Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2490
3
3.5
An Analysis of the Agreement Between Pre-Teenagers’ Self Report and Parents’ Proxy Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2491 Mean Differences between Pre-Teenagers’ Self-Report and Mothers’ and Fathers’ Proxy-Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2491 Level of Agreement Between Pre-Teenagers’ Self-Report and Parents’ Proxy-Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2493 Mean Differences and Agreements Between Pre-Teenagers’ Self-Report and Single Parents’ and Dual Parents’ Proxy-Reports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2495 Mean Differences and Agreement Between Pre-Teenagers’ Self-Report and Parents’ Proxy-Report Based on the Gender of the Child and the Gender of the Parent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2496 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2498
4
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2500
3.1 3.2 3.3 3.4
Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2501
#
Springer Science+Business Media LLC 2010 (USA)
2484
145
Quality of Life in Healthy and Chronically Ill Icelandic Children
Abstract: The purpose of this chapter is to report on findings on level of agreement between 10 and 12-year-old healthy and chronically ill Icelandic pre-teenagers’ reports and their mothers’ and fathers’ proxy-reports on the children’s > health related quality of life (HRQOL) and to view the findings in the light of the international literature. The research is cross-sectional and was introduced to 1,079 children in 5th and 6th grade and their parents. Out of those, 480 children (209 boys and 271 girls) and 912 parents (510 mothers and 402 fathers) gave their written consent and participated in the study. Data were collected from March to early June 2004 in 12 randomly selected public elementary schools in Reykjavik, Iceland. Descriptive statistics and dependent t-test were used to answer the research questions and to test the hypotheses. The main findings were that Icelandic pre-teenagers’ > self-report differed significantly from their mothers’ and fathers’ proxy-report on the social and school functioning subscales of the HRQOL measure, as well as on the overall HRQOL score. Further findings were that mothers of healthy children and both parents of chronically ill children differed significantly from their children in their perception of the physical functioning of the child. These findings emphasize that > parents’ proxy-report cannot be substituted for the pre-teenagers’ report on their own HRQOL. However, within the subscale of emotional functioning, the children and their parents were found to agree. Agreement was also found between fathers and their healthy children on the physical functioning subscale. This suggests that Icelandic parents can provide valid information on their children’s physical and emotional functioning, which can be used as a substitute for the children’s own response. Although parents’ proxy-report can only substitute a child’s self-report within the emotional and physical functioning subscale, a proxy-report can add needed and valid information regarding parents’ perspective on their pre-teenagers’ HRQOL. List of Abbreviations: HRQOL, health related quality of life; QOL, quality of life
1
Introduction
Parents of healthy as well as chronically ill pre-teenagers play a crucial role in their children’s life. In their pre-teenage years, children who receive support and encouragement from their parents, other close family members, and school personnel are developing an identity, independency, and a sense of accomplishment when achieving normal growth and development. Parents are expected to encourage, support and stimulate children’s development towards adulthood by providing a safe environment and by guiding and strengthening the development of educational, emotional, relational and behavioral skills as the children progress through the adolescent years. Early information on health related quality of life (HRQOL) among school children can help identify health related problems and give school nurses an opportunity to develop interventions, but one factor that greatly influences developmental outcomes and quality of life in children is their health status. Quality of life (QOL) has been defined broadly as life satisfaction (Vila et al., 2003). Further, Varni et al. (2004) have emphasized the importance that HRQOL instruments are sensitive to cognitive development and include both the child’s self-report and parents’ proxyreport to reflect both perspectives. However, lack of agreement between self- and proxy-report has been documented in the literature (Annett et al., 2003; Guyatt et al., 1997; le Coq et al., 2000; Varni et al., 2004). No correlation was found between asthma control as viewed by clinician and the quality of life scores of the children (Williams and Williams, 2003). In addition, low correlation was found between the children’s and the parents’ rating of HRQOL.
Quality of Life in Healthy and Chronically Ill Icelandic Children
145
Self-report is considered the standard for measuring perceived HRQOL among adolescents (Janse et al., 2005). HRQOL measures include both physical and psychological functioning but, in order to evaluate the impact of illness on day-to-day functioning, investigators have, in a range of clinical studies, included both disease specific and generic HRQOL questionnaires. Nevertheless, it is typically the parents’ perceptions of their children’s HRQOL that influence health care utilization. Therefore, it is important to determine whether proxy-reports are valid and whether they can be used to assess children’s QOL for example, when children’s self-report data is not possible to obtain. Little is known about quality of life of pre-teenagers and about parental perception of HRQOL when a child in the general population is diagnosed with a chronic health condition or illness. The purpose of this study is to evaluate agreement between pre-teenagers’ selfreports on HRQOL and their parents’ proxy-report (both the mothers’ and the fathers’ perspective) among 10- to 12-year-old healthy and chronically ill Icelandic children.
1.1
HRQOL from the Child’s, the Adolescent’s and the Parent’s Perspective
Health related quality of life has been studied among children and adolescents with variety of chronic illnesses, both from the child’s or the adolescent’s own perspective (CharronProchownik, 2002; Faulkner, 2003; Faulkner and Chang, 2007; Fiese et al., 2005; Mednick et al., 2004; Varni et al., 2003) as well as from the parents’ or caregivers’ perspective (Bothwell et al., 2002; Hays et al., 2006; Knowles et al., 2007; Landolt et al., 2002; Markham and Dean, 2006; Sheppard et al., 2005; Whitney, 2005). In a recent study on quality of life among children with asthma, the children’s perception of their own quality of life was found to be associated with the parents’ report of routine burden (Fiese et al., 2005). The authors emphasized that pediatric asthma management is a multifaceted activity and that in family environments where parents are overwhelmed or burdened by care there is a cost to the child as well. However, when families are able to create predictable routines around daily care, their children may be better equipped to follow doctor’s orders, less likely to have anxiety-related symptoms, and have a better overall quality of life (Fiese et al., 2005). Ungar et al. (2006) pointed out in a study on analysis of a dyad approach to health related quality of life measurement in children with asthma that by bringing parent and child together, the dyad provides a forum for discussion and elaboration of perceptions of HRQOL. The authors highlighted that through skilled interviewer facilitation, parent and child may learn from each other’s judgments and the child may be enabled to provide more reliable responses to structured HRQOL questionnaire items. International literature on quality of life among families of children with asthma has among other things focused on life satisfaction. In a sample of Swedish families, Rydstro¨m et al. (2004) found that a desire to be like other families and hope and longing for a better life was the families’ strategy to handle uncertainty regarding the disease. However, Dalheim-Englund et al. (2004) found in a sample of 371 Swedish parents of children with asthma that most of the parents evaluated their own quality of life as being close to the positive end of the scale, and there was close agreement in the scoring between parents within the same family. In that study, children with asthma were not found to influence the parents’ QOL to a great degree. Further, in a US sample of children with diabetes, the parents of 10- to 12-year-old school age children were found to experience greater life satisfaction (within the health related quality of life subscale) than parents of adolescents (13–18 years) (Faulkner and Clark, 1998). In another
2485
2486
145
Quality of Life in Healthy and Chronically Ill Icelandic Children
study on 99 African-American families of school age children and adolescents with type 1 diabetes, Faulkner and Chang (2007) found that emphasizing open family communication and providing emotional support for diabetic management would benefit the families regarding developmentally appropriate levels of self-care and quality of life. For families of children with cancer, the number of days from central venous catheter placement in the child and the child’s coping was found to significantly predict the children’s quality of life, which in turn predicted parental trust in the medical care (Tremolada et al., 2005). Further, a difference in perception of quality of life has been found between parents of chronically ill children and pediatricians (Janse et al., 2005) not only at diagnosis but also within 6 months follow-up measures. Few researches have focused on quality of life among healthy school age children. Chen et al. (2005) conducted a study on the association of lifestyle factors with quality of life in 7,794 Japanese children aged 9–10 years and found unfavorable lifestyles in childhood, like almost never participating in exercise or having seldom breakfast, to be associated with poor quality of life in early adolescence. However, comparing quality of life among healthy as well as chronically ill children has received some attention in the literature. Meuleners et al. (2002) conducted a Delphi study regarding teachers’, parents’ and health care professionals’ perception of the relative importance of different aspects of quality of life for adolescents with chronic illnesses. Half of the panels of parents and health professionals, and 68% of teachers, perceived the chronically ill adolescents to have worse QOL than their healthy counterparts. Reasons cited by the panels included: a poorer attitude as result of the chronic illness, the adolescent can be limited by what he or she can do, poor physical health, lack of independence, having greater obstacles to overcome than a healthy adolescent, the illness makes the chronically ill children different to everyone else and that the health condition prevents chronically ill children from making long term plans.
1.2
Level of Agreement Between Child’s Self-Report and the Parents’ Report on HRQOL
Little research has been published on the agreement between children’s and parents’ proxyreport on quality of life, neither in the literature on chronic illnesses nor in the general population literature. Upton et al. (2005) conducted a study on 69 children (aged 8–18 years) in public care and their carers (requited through routine pediatric assessments) and 662 children not in public care (recruited from local schools). The results indicated significant correlation with generic module scores between proxy and self-report scores. However, in a study on health related quality of life among 1,760 French adolescents and their parents (Simeoni et al., 2001), the parents’ score was found to be significantly higher than the corresponding scores from the adolescents; this was true for the total index and for each dimension score except for the subscales of inaction and relationships with friends. In addition, in a study on families of children with cystic fibrosis (Epker and Maddrey, 1998), the parents were found to perceive the QOL of their children to experience greater impairment in the domains of behavior, self-esteem and general health perception than the children themselves reported. Further, Noyes (2007) found, in a study on health related quality of life among seventeen 4- to 18-year-old children that were dependent on ventilator, that the children and their parents reported the children’s overall health-related quality of life the same, but parents reported significantly lower scores for their child’s relationships with friends and their disease. The author emphasizes the need for both child and parent perspectives in order to understand the impact of ventilator-dependency on the child’s life.
Quality of Life in Healthy and Chronically Ill Icelandic Children
145
The level of agreement between children’s perception and their parents’ perception on quality of life has, over the last few years, received increased attention, both among health care professionals and researchers. Contradictory findings have nevertheless been reported regarding the child’s perception and the parents’ proxy-report. Chang and Yeh (2005) conducted a study on the agreement between children’s self-reports and their parents’ proxy-reports to evaluate quality of life in 141 Taiwanese children and adolescents with cancer (age 7–18 years) and their parents. Different statistical approaches were employed to evaluate convergence of self-reports and proxy-reports, such as product-moment correlation coefficient and the comparison of group means on the seven subscales (physical, psychological, social, disease symptom, cognitive, understanding and communication) as well as on the total score of the Quality of Life for Children with Cancer scale (QOLCC) (Yeh and Hung, 2003; Yeh et al., 2004a,b). The results indicated that children and parents had the largest discrepancy on the understanding subscale. For the adolescents (children age 12 and older), the mean differences were largest for the subscales of psychological factors and understanding, which indicates less agreement for these subscales than for the other subscales. The results suggested that for children who were younger and not able to evaluate QOL assessment due to their developmental limitation or severity of illness, the parents can provide valid information about the children’s QOL. However, parent-proxy of QOL for adolescents provided significantly different information than the adolescents’ self-report. The authors concluded that proxy-data of QOL for adolescents should be used with caution. However, Varni et al. (2003) conducted a study among 10,241 families of healthy as well as chronically ill children, age 2–16 years, and found a trend toward higher inter-correlations among parents proxy-report and increasing age of the children. Little is known about agreement on HRQOL between healthy and chronically ill preteenagers (10–12 years old children) and their parents’ proxy-report. Further, no study was found that evaluated agreement between pre-teenagers’ perception on their own HRQOL and their parents’ perception based on the children’s family type (single parent families or dual parent families) or on the gender of the child and the gender of the parent. Based on the review of the literature, it was hypothesized that there would be no significant difference on the total HRQOL scale nor on any of the subscales of the HRQOL measure between: (1) > healthy pre-teenagers and their parents; (2) chronically ill pre-teenagers and their parents; (3) children and their mothers’ proxy-report not matter what type of families the children lived in (single parent families or dual parent families); and (4) there would be no significant difference on the total HRQOL scale nor on any of the subscales of the HRQOL scale based on the gender of the pre-teenager nor on the gender of the parent. The following research questions were asked: (1) Is there a difference between healthy and chronically ill preteenagers’ levels of agreement on HRQOL (both the total score and within all subscales) and their parents’ proxy-report? (2) What is the association between pre-teenagers’ self-reports and parents’ proxy-reports on the total HRQOL scale as well as on all the subscales of the HRQOL measure?
2
Methodological Considerations
2.1
Procedure
This is a cross sectional study that is based on a larger study on generic health related quality of life (HRQOL) (from the children’s and the parents’ perspective) among 10- to 12-year-old
2487
2488
145
Quality of Life in Healthy and Chronically Ill Icelandic Children
Icelandic school children, which included both healthy and chronically ill pre-teenagers (Svavarsdottir and Orlygsdottir, 2006a,b). Data were collected from March to early June 2004 in 12 randomly selected public elementary schools in Reykjavı´k, Iceland. The study was introduced to 1,079 children in 5th and 6th grade and their parents. Out of those, 480 children (209 boys and 271 girls) and 912 parents (510 mothers and 402 fathers) gave their written consent (both the child and its parents signed the consent) and participated in the study (45% participation of parent(s)-child pairs). The inclusion criteria for the study were: (1) the children needed to be in 5th or 6th grade, and (2) needed to read and write Icelandic. The parents had to be able to read and write Icelandic. The research was approved by the Institutional Review Board of the Reykjavik Health Care Services, the Reykjavik Council of Education, the principals in the 12 elementary schools, the National Bioethics Committee, and the study was reported to the Data Protection Committee. All children in 5th and 6th grade of the participating schools were informed about the study by the data collection personnel and the school nurse in each school, who gave the children a package of documents to bring home to their parents. The package included an introduction letter, a form for informed consent, two questionnaire booklets (one for the mother and one for the father), and empty envelopes to bring back to the school nurse with a signed consent form (from both parents (if applicable) and the child) and completed questionnaires from the parents. The data collection personnel and the school nurse met with the children who had brought back a signed consent form from their parents. The children who participated were taken out of the classroom to answer the questionnaires. The data collection personnel and the school nurse gave verbal instructions and answered questions from the children. If the children had reading difficulties, either the data collection personnel or the school nurse read the questions out loud for the students.
2.2
Subjects/Sample
A majority of the families were dual parent families (n = 414; 72.71%) and 18.9% (n = 99) one parent families. Parents’ mean age was 40.27 years (SD = 5.45), ranging from 24 to 63 years. The mean age of the children was 10.95 years (SD = 0.645) (range 10–12 years), 56.5% were girls and the remainder boys. As reported by the parents, 24.6% (n = 118) of the children had a chronic health condition (chronic illnesses). Of the children, 94 were diagnosed with one health condition (disease), 20 children had two diseases, and 4 children had three diseases. The chronic health condition varied among the 118 children, from being a physical illness (79 health conditions, 54.11%; e.g., allergy/eczema, n = 23 (15.8%); migraine, n = 19 (13%); asthma, n = 15 (10.3%)) to mental and learning disabilities (67 health conditions, 45.89%; e.g., learning disabilities, n = 23 (15.8%); hyperactivity/attention deficit disorder, n = 23 (15.8%); dyslexia, n = 6 (4.1%) (see > Table 145-1). Range in the children’s visits to the school nurse varied from being ‘‘every day’’ (n = 2; 3%), up to ‘‘once that week’’ (n = 20; 27%). The children who indicated that they were bullied in school last week reported being bullied from ‘‘often a day’’ (n = 3; 9%) to ‘‘once that week’’ (n = 8; 23%). Further, sample information regarding the mean score of the total HRQOL scale as well as all the subscales of HRQOL (physical, emotional social and school functioning) are listed in the Figure, for the parents and the pre-teenagers (see > Figure 145-1).
145
Quality of Life in Healthy and Chronically Ill Icelandic Children
. Table 145-1 Demographics of Icelandic school age children and their parents Variables
Children (N = 480) (n) Mean
% SD
Female
(271)
(56.5)
Male
(209)
(43.5)
Age
10.5
0.665
Parents (N = 911) (n) mean
% SD
Range
Gender of the children
Gender of the parents Female
(510)
(59.8)
Male
(401)
(44.02)
Age
41.11
5.77
Single parent families
(99)
(18.9)
Dual parent families
(424)
(81.2)
Yes
(142)
(29.58)
No
(338)
(70.42)
24–63
Martial status
Child with chronic health condition
Gender of the pre-teenagers’ and their parents’ gender, age and marital status. The parents’ perception of their children’s chronic illnesses
. Figure 145-1 Mean scores of healthy (N = 117) and chronically ill pre-teenagers’ report (N = 360) and their parents’ proxy-report
2489
2490
145 2.3
Quality of Life in Healthy and Chronically Ill Icelandic Children
Instruments
The questionnaires were translated and culturally adapted to Icelandic from English. Two Icelandic health care providers independently translated the questionnaires into Icelandic (forward translation), and, after discussion and consultation with specialists, came to an agreement on a reconciled version. A professional bilingual translator, who is a native speaker of English and an Icelandic linguistic scholar, performed a back translation. Together, the back and forward translators decided on a new conceptually equivalent Icelandic version, which was pilot tested and validated on seven school children and their families. The instruments were finally proofread by an Icelandic professional. Validity of all the instruments that were used in the study was established by translating the instruments into Icelandic according to the steps described above. Demographic and background information. Demographic information such as age, gender, grade, after school activities, visits to school nurse and experience of bullying victimization was reported by the children themselves. In addition, the parents answered questions regarding their own background and about their child’s chronic health condition/illness(es). Children’s quality of life. The children’s health related quality of life (HRQOL) was measured by the 23-item instrument ‘‘Pediatric Quality of Life Inventory’’ (PedsQL), the Generic Core Scale for 8- to 12-year-old children (children’s self-report scale), which is to be administered to healthy school children, community population, or children with acute or chronic illnesses. The instrument has four subscales, physical functioning (eight items), emotional functioning (five items), social functioning (five items), and school functioning (five items). The children are asked how much of a problem each item has been for them for the past 1 month, ranging from ‘‘never’’ to ‘‘almost always a problem.’’ The items are scored reversely and put on a 0–100 scale, where higher scores point to a higher HRQOL (0 = 100, 1 = 75, 2 = 50, 3 = 25, 4 = 0) (Varni et al., 2002). The internal consistency of the total scale is 0.88 for US children. For the Icelandic children, the alpha reliability was 0.90 for boys and 0.86 for girls. The internal consistency scores for the subscales ranged from 0.68 to 0.80 for the US sample (Varni et al., 2001), 0.70–0.79 for the Icelandic boys and 0.66–0.77 for the Icelandic girls. Children’s quality of life as perceived by their parents. The children’s HRQOL as perceived by parents was measured by the 23-item instrument Pediatric Quality of Life Inventory (PedsQL), the Generic Core Scale for 8- to 12-year-old children (parent proxy-report; Varni et al., 1999, 2001). The instrument is a multidimensional parent proxy-report scale that is to be administered to parents of healthy school children or children with acute or chronic illness. The questionnaire is identical to the PedsQL (children’s self-report) except that it is phrased in the third person. As described above the instrument has four subscales: physical functioning (eight items), emotional functioning (five items), social functioning (five items) and school functioning (five items). The parents were asked to rate how much of a problem each item had been for their child in the past month, ranging from ‘‘never a problem’’ to ‘‘almost always a problem.’’ The items are scored reversely on a 0–100 scale (Varni et al., 2002). The internal consistency of the total scale is 0.90 for US parents. For the Icelandic parents, the alpha reliability for the total scale was 0.86 for mothers and 0.87 for fathers.
2.4
Data Analysis
The data on major study variables (average score for each participant on the total HRQOL score as well as for all the subscales) were normally distributed according to Stem and Leaf
Quality of Life in Healthy and Chronically Ill Icelandic Children
145
plots and histograms. Descriptive statistics, such as means and standard deviations or frequency distributions, were calculated for the major study variables on HRQOL, both the total score and the subscales, as well as for the demographic variables. Descriptive statistics and dependent t-tests were used to answer the research questions and to test the hypotheses of no significant difference in agreement between pre-teenagers’ reports on HRQOL (both for the total score and subscales) and their parents’ proxy-report. Alpha level of significance was set at 0.05 to reduce the likelihood of committing a Type I error. In order to compare the differences between the child’s self-report and parent’s proxyreport, the mean scores and standard deviations were summarized separately. Further, the means of the differences (mean bias) (the children’s group – parents’ group) and the standard deviation of difference were computed (Chang and Yeh, 2005). A mean difference of less than zero indicates that the parents tend to overestimate their children’s HRQOL and a mean difference greater than zero indicates that parents tend to underestimate it. The > effect size d was used to examine the magnitude of this difference, which was found by dividing the mean difference by the SD of the mean score (Chang and Yeh, 2005). The value of the d was judged by the guideline provided by Cohen (1992) and as presented in a research by Chang and Yeh (2005), where d = 0.2 was categorized as a small effect size, d = 0.5 a medium effect size and d = 0.8 a large effect size. In addition, the agreement between the children and parents was further quantified using Person correlation coefficients. The Person product correlation coefficient’s effective size was categorized further. When the correlation coefficient was smaller than 0.3 the effective size was labeled as small, when the correlation coefficient was between 0.3 and 0.5 it was evaluated to be medium, and when the correlation coefficient was equal or larger than 0.5 the effective size was considered to be large (Chang and Yeh, 2005; Cohen, 1992).
3
An Analysis of the Agreement Between Pre-Teenagers’ Self Report and Parents’ Proxy Report
3.1
Mean Differences between Pre-Teenagers’ Self-Report and Mothers’ and Fathers’ Proxy-Report
For healthy pre-teenagers, a significant difference was found between the children’s self-report and the mothers’ proxy-report on the subscale of physical functioning. The mothers were found to report significantly lower physical functioning of their children than the children themselves. A significant difference was also found between healthy pre-teenagers’ self-report and their mothers’ proxy-report and between the children and their fathers perception on social functioning and school functioning. Thus, the mothers and the fathers were found to report significantly lower social functioning and significantly lower school functioning than the pre-teenagers. Further, a significant difference was found between the children’s total health related quality of life score and the mothers’ and the fathers’ response; both the mothers and the fathers reported significantly lower health related quality of life (total score) than their healthy pre-teenagers. Interestingly, however, no difference was found between the children’s self-report and their fathers’ proxy-report on the physical functioning subscale, or between the parents’ and the children’s subscale of emotional functioning (see > Table 145-2). For children with chronic illness(es) (both physical illnesses and psychological illnesses), a significant difference was found between the pre-teenagers’ self-report and their parents’ proxy-report (both the mothers’ and the fathers’ proxy); the parents reported significantly
2491
2492
145
Quality of Life in Healthy and Chronically Ill Icelandic Children
. Table 145-2 Icelandic schoolage children’s (N = 480) mean differences in self-report and parents’(N = 524) proxy report regarding HRQOL (subscales and total score) Variables
Mean
SD
88.41
9.48
Mothers’-proxy report (n = 330)
86.40
14.66
Self-report (n = 251)
87.98
9.36
Fathers’-proxy report (n = 251)
86.42
14.46
77.89
14.92
t-valuea
df
p-value
2.308
329
0.02
1.568
250
0.12
1.142
324
0.25
1.869
247
1.63
9.194
323
0.00
8.293
248
0.00
8.508
323
0.00
7.696
247
0.00
326
0.00
5.999
249
0.00
2.074
113
0.04
2.114
87
0.04
0.724
111
0.47
1.26
86
0.21
Healthy children (n = 362) Physical functioning Self-report (n = 330)
Emotional functioning Self-report (n = 325) Mothers’-proxy report (n = 325)
78.92
13.46
Self-report (n = 248)
77.54
14.95
Fathers’-proxy report (n = 248)
79.64
12.88
Self-report (n = 324)
86.50
14.84
Mothers’-proxy report (n = 324)
74.25
22.94
Self-report (n = 249)
85.64
15.18
Fathers’-proxy report (n = 249)
72.76
22.91
Self-report (n = 324)
83.66
12.84
Mothers’-proxy report (n = 324)
73.60
19.00
Self-report (n = 248)
83.28
13.21
Fathers’-proxy report (n = 248)
72.05
20.19
Social functioning
School functioning
Total HRQOL score Self-report (n = 327)
84.77
9.96
Mothers’-proxy report (n = 327)
79.45
79.45
Self-report (n = 250)
84.36
10.19
Fathers’-proxy report (n = 250)
78.87
12.87
79.45
Children with chronic illness(es) (n = 118) Physical functioning Self-report (n = 114)
82.60
13.52
Mothers’-proxy report (n = 114)
79.27
16.08
Self-report (n = 88)
82.72
14.03
Fathers’-proxy report (n = 88)
78.61
18.03
Self-report (n = 112)
72.83
17.28
Mothers’-proxy report (n = 112)
71.70
14.44
Self-report (n = 87)
73.35
16.65
Fathers’-proxy report (n = 87)
71.14
14.48
Emotional functioning
145
Quality of Life in Healthy and Chronically Ill Icelandic Children
. Table 145-2 (continued) Variables
Mean
SD
Self-report (n = 110)
80.74
16.92
Mothers’-proxy report (n = 110)
64.26
20.89
Self-report (n = 86)
81.86
16.19
Fathers’-proxy report (n = 86)
64.00
19.69
Self-report (n = 112)
77.51
15.94
Mothers’-proxy report (n = 112)
66.62
17.35
Self-report (n=85)
79.03
14.22
Fathers’-proxy report (n=85)
64.58
18.13
78.73
12.86
t-valuea
df
p-value
7.560
109
0.00
7.371
85
0.00
5.403
111
0.00
6.162
84
0.00
5.697
112
0.00
5.84
85
0.00
Social functioning
School functioning
Total HRQOL score Self-report (n = 113) Mothers’-proxy report (n = 113)
71.53
12.14
Self-report (n = 87)
79.56
12.33
Fathers’-proxy report (n = 87)
71.00
12.71
Difference between pre-teenagers own report and their parents’report regarding health related quality of life. n = varies due to missing data a Paired t-tests
lower physical functioning than the > pre-teenagers with chronic illnesses. Further, significant difference was found between the self-report of the children with chronic illness(es) and their parents’ proxy on the subscales of social functioning, school functioning and on the total health related quality of life score. Parents of children with a chronic health condition reported lower social functioning, school functioning and lower total health related quality of life score than the children themselves. Nevertheless, no differences were found between the children’s report and their parents’ proxy-report on the subscale of emotional functioning. Hypotheses (1) and (2) were therefore only partly supported (see > Table 145-2).
3.2
Level of Agreement Between Pre-Teenagers’ Self-Report and Parents’ Proxy-Report
When the > level of agreement between the children’s and their parents’ proxy-report was further evaluated (to answer the research questions), healthy children and their parents’ proxy differed the most on the social and the school functioning subscales. The mean differences on these two subscales varied from 10.06 to 12.88 (see > Table 145-3). Similarly, for the children with chronic illness(es), their self-report and their parents’ proxy-report differed the most on the subscales of social and school functioning, where the mean difference varied from 10.89 to 17.86 (see > Table 145-3). As introduced by Chang and Yeh (2005), a mean difference below zero between self- and proxy-report equals > overestimation, whereas a mean difference greater than zero equals > underestimation. For these Icelandic pre-teenagers, both healthy children and children dealing with chronic illness(es), mean difference on all the subscales as well as on the total HRQOL score was greater than 0 for both the mothers’ (varied from 1.03 to
2493
2494
145
Quality of Life in Healthy and Chronically Ill Icelandic Children
. Table 145-3 Level of agreement between children’s self report (N = 480), and the parents’ proxy report (mothers N = 510; fathers N = 401) of HRQOL (subscales and total score) Variables
Mean difference
Pearson’s correlation
SD of difference
Effect size d
2.01
0.20*
15.80
0.13
*
16.27
0.06
Healthy children Mothers versus children Physical functioning (n = 330) Emotional functioning (n = 325)
1.03
0.35
Social functioning (n = 324)
12.25
0.25*
23.98
0.51
School functioning (n = 324)
10.06
0.15*
21.29
o.47
Total HRQOL score (n = 327)
5.32
0.24*
14.14
0.38
1.56
0.18*
15.77
0.10
*
17.64
0.12
Fathers versus Children Physical functioning (n = 251) Emotional functioning (n = 248)
2.09
0.20
Social functioning (n = 249)
12.88
0.22*
24.50
0.53
School functioning (n = 248)
11.23
0.10
22.99
0.49
Total HRQOL score (n = 250)
5.49
0.23*
14.46
0.38
Physical Functioning (n = 114)
3.33
0.34*
17.13
0.19
Emotional Functioning (n = 112)
1.14
0.46*
16.64
0.07
Social Functioning (n = 110)
16.48
0.28
*
22.86
0.72
School Functioning (n = 112)
10.89
0.18
21.32
0.51
Total HRQOL score (n = 113)
7.20
0.42*
13.43
0.54
Physical Functioning (n = 88)
4.11
0.38*
18.23
0.23
Emotional Functioning (n = 87)
2.21
0.45*
16.38
0.14
Social Functioning (n = 86)
17.86
0.23
**
22.47
0.79
School Functioning (n = 85)
14.45
0.12
21.61
0.67
13.67
0.63
Children with chronic illness(es) Mothers versus Children
Fathers versus Children
Total HRQOL score (n = 87)
8.56
0.40
*
Agreement between children’s report and their parents report on health related quality of life * p < 0.01 ** p < 0.05
12.25 for healthy children; and from 1.14 to 16.48 for children with chronic illness(es)) and the fathers’ proxy-report (varied from 1.56 to 12.88 for healthy children; and from 2.21 to 17.86 for children with chronic illness(es)) (see > Table 145-3). These results indicate that the Icelandic pre-teenagers and their parents had the largest discrepancy on the subscales of social and school functioning, which shows less agreement on these two subscales than the subscales of physical and emotional functioning. Further, these findings indicate that the parents of both health and chronically ill pre-teenagers had a tendency to underestimate their children’s HRQOL when compared to the report of the children’s themselves.
Quality of Life in Healthy and Chronically Ill Icelandic Children
145
In order to evaluate further the level of difference between the children’s self-report and the parents’ proxy-report, the magnitude of the difference was found by dividing the mean difference with the SD of the mean score (Chang and Yeh, 2005; Cohen, 1992). For the healthy pre-teenagers, the effect size d was small for the physical and the emotional functioning subscales as well as the total HRQOL score, but medium for the social and the school functioning subscales. However, for the pre-teenagers with physical and psychological chronic illness(es), the effect size was small for the physical and emotional subscales, but medium or high for the total HRQOL score and the subscales of social and school functioning (see > Table 145-3). The association between the children’s self-report and their parents’ proxy-report was further evaluated by Pearson r correlation. As indicated by Chang and Yeh (2005) and Cohen (1992), Pearson product correlation coefficient smaller than 0.3 equals small effect size, correlation coefficient between 0.3 and 0.5 equal medium effect size, and correlation coefficient equal to or larger than 0.5 equals large effect size. For the families of the healthy children, Pearson product correlation between the children and their parents’ proxy-report, ranged from 0.15 (school functioning, small correlation) to 0.35 (emotional functioning, medium correlation) (p < 0.01), indicating a small to medium degree of association (see > Table 145-3). Similarly, Pearson product correlation between children with chronic illness(es) and their parents’ proxy-report, ranged from 0.23 (social functioning, small correlation) to 0.46 (emotional functioning, median correlation) (p < 0.01), indicating small to medium association (see > Table 145-3).
3.3
Mean Differences and Agreements Between Pre-Teenagers’ Self-Report and Single Parents’ and Dual Parents’ Proxy-Reports
A significant difference was found between single mothers’ proxy-report and their preteenagers on the social and school functioning subscales and on the total HRQOL scale. The single mothers were found to report significantly lower total HRQOL score as well as significantly lower social and school functioning score than the pre-teenagers themselves. For mothers and children living in dual parent families, a significant difference was also found on the mothers’ proxy-report and their children’s self-report on the total HRQOL scale as well as on the physical, social and school functioning subscales. Thus, mothers’ proxy-report in dual parent families was significantly lower on the subscales of physical, social and school functioning as well as on the total HRQOL score than the pre-teenagers’ self-report scores. Hypothesis (3) was therefore only partly supported (see > Table 145-4). For all the children who were living with a single parent or living in dual parent families, mean difference on the subscales as well as on the total HRQOL score was greater than 0. Difference scores between children and their single parent varied from 1.98 to 15.38 (see > Table 145-4). Difference scores between children and their mothers in dual parent families varied from 1.02 to 12.71, indicating that Icelandic school age children, no matter what family type they were living in, had the largest discrepancy with their mother’s proxy-report on the subscales of social and school functioning of the HRQOL scale (see > Table 145-4). Interestingly, the magnitude (effect size) of the difference was small for the subscales of physical and emotional functioning, both for children in single and dual parent families, but medium for the social an school functioning in both family types. However, for children living in single parent families, the magnitude (effect size) of the difference between the child and their single mothers was medium for the total HRQOL score but small for children living in dual parent families. The association between the children’s self-report and their parents’ proxy-report
2495
2496
145
Quality of Life in Healthy and Chronically Ill Icelandic Children
. Table 145-4 Mean differences and level of agreement between Icelandic school age children (N = 480), single parent families’ (n = 99) and dual parent families’ proxy report (n = 424) on HRQOL (both subscales and total score) Variables
Pearsons Mean diff. correlation SD of diff.
t-valuea
Effect size d
3.95
0.25*
19.21
1,848
0.21
Schoolage children (N = 480) Single mothers versus children Physical functioning (n = 82) Emotional functioning (n = 81)
1.98
0.44**
15.39
1,155
0.13
Social functioning (n = 79)
15.38
0.35**
22.30
6,130***
0.69
School functioning (n = 80)
12.59
0.22*
19.90
5,622***
0.63
Total HRQOL score (n = 81)
7.99
0.35**
13.79
5,216***
0.58
1.98
0.29**
15.41
2,445*
0.13
Dual parent families versus Children Physical functioning (n = 361) Emotional functioning (n = 354)
1.02
0.39**
16.61
1,160
0.06
Social functioning (n = 353)
12.71
0.27**
23.97
9,961***
0.53
School functioning (n = 315)
9.67
0.18**
21.58
8,427***
0.45
Total HRQOL score (n = 354)
5.29
0.34**
13.95
7,164***
0.38
Difference in agreement between pre-teenagers and their single parent families and their dual parents families on their health related quality of life *paired t-tests; p < 0.05 **p < 0.01 ***p < 0.001
ranged from 0.18 (school functioning, small correlation) to 0.44 (emotional functioning, medium correlation) (p < 0.01), indicating a small to medium degree of association for both family types (see > Table 145-4).
3.4
Mean Differences and Agreement Between Pre-Teenagers’ Self-Report and Parents’ Proxy-Report Based on the Gender of the Child and the Gender of the Parent
A significant difference was found between the mothers’ proxy-report and their daughters’ self-report on the physical, social and school functioning subscales and on the total HRQOL scale. The mothers were found to report significantly lower total HRQOL score as well as significantly lower physical, social and school functioning score than the pre-teenagers themselves. Further, a significant difference was found between the fathers’ proxy-report and their daughters’ self-report on the physical, social and school functioning subscales and on the total HRQOL scale. The fathers reported significantly lower total HRQOL score as well as significantly lower physical, social and school functioning score compared to the preteenagers themselves. Hypothesis (4) was therefore partly supported (see > Table 145-5).
145
Quality of Life in Healthy and Chronically Ill Icelandic Children
. Table 145-5 Mean difference and the level of aggrement between parents proxy report (N = 524) of HRQOL and the child’s self report (N = 580) based on the gender of the school age child Variables
Pearson’s Mean diff. correlation SD of diff.
t-valuea
Effect size d
School age children (n = 524) Mothers versus daughters Physical functioning (n = 250)
3.03
0.27*
16.02
*
2,987*
0.19
0.30
0.45
15.30
0,305
0.02
Social functioning (n = 245)
15.68
0.30*
22.82
10,756**
0.69
School functioning (n = 247)
13.08
0.14***
21.78
9,436**
0.60
7.19
0.35
*
13.25
8,568
**
0.54
2.44
0.32*
14.90
2,294***
0.16
*
Emotional functioning (n = 247)
Total HRQOL score (n = 249) Fathers versus Daughters Physical functioning (n = 196)
1.63
0.33
16.74
1,362
0.10
Social functioning (n = 195)
14.29
0.25*
23.46
8,504**
0.61
School Functioning (n = 195)
13.28
0.09
23.28
7,964**
0.57
6.70
0.33
*
13.60
6,920
**
0.49
Physical functioning (n = 194)
1.47
0.30*
16.30
1,255
0.09
Emotional functioning (n = 190)
0.71
0.34*
17.72
0.549
0.04
10.26
0.28*
24.62
5,729**
0.42
School functioning (n = 189)
6.61
0.25
*
20.07
4,529
**
0.33
Total HRQOL score (n = 191)
3.98
0.34*
14.70
3,749**
0.27
Physical functioning (n = 143)
1.92
0.25*
18.43
1,247
0.10
Emotional functioning (n = 140)
0.06
0.27*
18.43
0,38
0.00
0.20***
18.30
6,624**
0.56
**
0.48 0.37
Emotional functioning (n = 195)
Total HRQOL score (n = 197) Mothers versus Sons
Social functioning (n = 189)
Fathers versus sons
Social functioning (n = 140)
13.97
School functioning (n = 138)
10.32
0.15
24.96
5,587
Total HRQOL score (n = 140)
5.69
0.29*
21.71
4,403**
Agreement between parents and the pre-teenagers on health related quality of life based on the gender of the child a p < 0.01 b p < 0.001 c p < 0.05
When the parent-son dyads were evaluated, a significant difference was also found between the mothers’ proxy-report and their sons’ self-report on the social and school functioning subscales and on the total HRQOL scale. The mothers were found to report significantly lower total HRQOL score as well as significantly lower social and school functioning score than their sons’ themselves. In addition, a significant difference was also found between the fathers’ proxy-report and their sons’ self-report on the social and school functioning subscales and on
2497
2498
145
Quality of Life in Healthy and Chronically Ill Icelandic Children
the total HRQOL score, indicating that fathers reported significantly lower total HRQOL score as well as significantly lower social and school functioning score than their sons’ in their selfreport. Hypothesis (4) was therefore partly supported (see > Table 145-5). When all the school age children were put together in one group (healthy as well as chronically ill children), mean difference based on the gender of the parents as well as on the gender of the child was greater than 0 for both the total score of HRQOL as well as all the subscales of HRQOL (see > Table 145-5). Difference score between mothers and their daughters varied from 0.30 to 15.68; difference score between fathers and their sons varied from 0.06 to 13.97; difference scores between fathers and daughters varied from 1.63 to 14.29 and difference scores between mothers and sons varied from 0.71 to 10.26; indicating that the parents and their pre-teenagers disagreed the most on the social and school functioning subscales of the HRQOL scale, no matter what gender combination of the ‘‘parent-child dyad’’ was evaluated, which shows less agreement on these two subscales than the subscales of physical and emotional functioning (see > Table 145-5). Further, these findings indicate that the parents had a tendency to underestimate their children’s HRQOL no matter what parent-child gender pair was evaluated (see > Table 145-5). The magnitude (effect size) of the difference was small for the physical and emotional functioning subscales both for the mother-daughter pairs and the father-daughter pairs, but medium for these same gender pairs regarding social and school functioning as well as for the total HRQOL scale. However, for the mother-son pairs and the father-son pairs, the magnitude of the difference was none or small on the physical and emotional functioning, but small for the school functioning subscale as well as for the total HRQOL score (see > Table 145-5). For the father-son pairs, the magnitude of difference was medium regarding the social functioning subscale but small for the mother-son pairs. The association between the mother-daughter pairs ranged from 0.14 (school functioning, small correlation) to 0.45 (emotional functioning, medium correlation) (p < 0.01), indicating small to medium degree of association. For the father-daughter pairs, the association ranged from 0.25 (social functioning, small correlation) to 0.33 (emotional functioning; total HRQOL score, medium correlation) (p < 0.01); indicating small to medium degree of association. The association between the mother-son pairs ranged from 0.25 (school functioning, small correlation) to 0.34 (emotional functioning, total HRQOL score, small correlation) (p < 0.01), indicating small to medium correlation. For the father-son pairs, the association ranged from 0.20 (social functioning, small correlation) to 0.29 (total HRQOL score, small correlation) (p < 0.01), indicating a small degree of association (see > Table 145-5).
3.5
Discussion
Key findings from the hypotheses testing add to our understanding of health related quality of life among healthy as well as chronically ill Icelandic pre-teenagers. Knowing that both healthy and chronically ill Icelandic pre-teenagers’ perspective differs significantly from their mothers’ and their fathers’ perspective on their social and school functioning as well as on the overall HRQOL measure, and that the mothers of healthy children and both parents of chronically ill children differed significantly from their children in their perception of physical functioning, is important information for health care professionals and school personnel. These findings emphasize that parents’ proxy-report cannot be substituted for the pre-teenagers’ perspective on their own HRQOL. In all of these cases the parents reported significantly lower mean scores
Quality of Life in Healthy and Chronically Ill Icelandic Children
145
of social, school and physical functioning than their healthy or chronically ill pre-teenagers; thus, Icelandic parents underestimated their pre-teenagers on these subscales of HRQOL. These findings on 10- to 12-year-old Icelandic school age children and their parents are contradictory to the findings of Chang and Yeh (2005), who reported that parents could provide valid information of QOL for their children under 12 years of age and that parents’ response could be used as substitute for the children’s response. In this Icelandic study, it was only within the subscale of emotional functioning that the children and their parents were found to agree. A good agreement was also found between the fathers and their healthy children on the physical functioning subscale. This finding has not been previously reported in the literature, but indicates that Icelandic parents can provide valid information on their children’s emotional functioning, and that that information can be used as a substitute for the healthy children’s or the chronically ill children’s own response. Interestingly, however, a gender difference was found between the parents. Icelandic fathers were found to be in agreement with their healthy pre-teenagers regarding their children’s physical functioning. Fathers in Iceland might therefore be a good source of information for school personnel such as school nurses regarding physical functioning of their healthy children and could be used as a valid source or as a supplement for the children’s own response. Nevertheless, the magnitude of the mean bias for healthy children and their parents on the physical and the emotional functioning scales was found to be small, but medium on the social and school functioning subscales. However, for the pre-teenagers with chronic illnesses and their parents, the magnitude of the mean bias was small for the physical and emotional functioning subscales, but medium to high for the social and school functioning subscales. For the total HRQOL scale, the magnitude of the difference between the pre-teenagers’ report and their parents’ proxy-report was found to be small for the healthy children and their parents but medium for the pre-teenagers with chronic illness and their parents’ proxy-report. In this sample of pre-teenagers and their mothers and fathers, there was a consistent bias for parental proxies to underestimate physical, social and school functioning as well as the total HRQOL scale score for both the healthy as well as chronically ill pre-teenagers. This finding emphasizes the need for school personnel and health care professionals to gather both parents’ and the pre-teenagers’ perspectives on children’s HRQOL. Knowing both the children’s and their parents’ perception is important in order to gain a more holistic information from family members regarding both healthy as well as chronically ill children’s HRQOL. Parents’ proxyreport needs to be used with caution as a substitute information regarding 10- to 12-year-old children’s HRQOL. However, the parents’ perception of their pre-teenagers’ HRQOL is valid additional information for health care professionals and school personnel. At the same time, this information needs to be used as such, that is, as additional information and not as substitute information for the pre-teenagers’ own response. According to Varni et al. (2003) and Chang and Yeh (2005), it is useful to evaluate the level of agreement between children’s self-report and their parents’ proxy-reports by examining the Pearson product correlation coefficient. In this study, all subscales of the HRQOL scale as well as the total scale score were found to be low or moderately correlated between mothers and their healthy children. Correlation was also low for the fathers and their healthy children except on the school functioning subscale, where no correlation was found. Similarly, for both parents and their chronically ill children, a moderate or low correlation was found on all subscales of the HRQOL scale as well as on the total scale score except for the school functioning subscale, where no correlation was found between the parents and their chronically ill pre-teenagers. According to Chang and Yeh (2005), Pearson correlation has been the
2499
2500
145
Quality of Life in Healthy and Chronically Ill Icelandic Children
most used approach to examine the proxy-validity and has been referred to as the level of agreement. However, according to these authors, this method has been criticized and they point out that Pearson correlation coefficient only measures the strength of a relation between two variables, rather than agreement. These authors emphasize that, since the Pearson correlation coefficient is insufficient to evaluate agreement between children’s self-report and parents’ proxy-report, group differences should be used to further supplement information provided by correlation coefficients. In this Icelandic study, significantly lower mean differences were found between single mothers’ proxy-report and the pre-teenagers’ self-report on the social and school functioning subscales and on the total HRQOL. Similarly, for mothers and children living in dual parent families, a significantly lower mean difference was also found on the mothers’ proxy-report and their pre-teenagers self-report and the total HRQOL scale as well as on the physical, social and school functioning subscales. This finding is new and has not been reported previously in the literature. Interestingly, however, for the subscales of emotional functioning, there was an absence of between-group difference for the pre-teenagers and their mothers’ proxy-response, indicating that the children and their mothers had a high level of agreement on the emotional functioning subscale no matter what family type the children lived in (single or dual parent families). Similarly, no difference was found between the children and their single mothers on the physical functioning subscale, indicating good agreement between the single mothers and their pre-teenagers regarding the child’s physical functioning. This was however not the case for children living in dual parent families. When gender difference was evaluated, it was interesting to notice how alike the mothers’ and the fathers’ proxy responses were. A possible explanation of this similarity in the parents’ responses regarding their perception of their pre-teenagers’ HRQOL is the age of the parents; the mothers mean age was 40.07 and the fathers’ mean age was 42.15, which indicates that the parents were experienced individuals. Nevertheless, and contrary to expectations, a significant difference was found between mothers and their daughters and the fathers and their daughters on the subscales of physical, social and school functioning as well as on the total HRQOL scale score. Interestingly, however, for the mothers and the fathers and their sons, a significant difference was only found on the subscale of social and school functioning and the total HRQOL scale score. These findings emphasize a gender difference between the preteenagers and their parents regarding the level of agreement on HRQOL. The Icelandic parents, both the mothers and the fathers, showed a higher level of agreement with their sons than their daughters on the pre-teenagers’ HRQOL.
4
Conclusion
International researchers have reported contradictory findings on agreement between parents’ proxy-reports and children’s own response on HRQOL. Most of these researchers have focused on a variety of different chronic illnesses in different cultures and on how developmentally capable children are, based on their age, to report themselves on their HRQOL and whether their parents’ report can serve as substitute report for the children. In order to achieve normal growth and development, healthy and chronically ill children need support from families, school personnel and health care professionals. Children’s health status has been found to influence developmental outcomes and QOL in children. Therefore, information regarding children’s HRQOL early on can help identify health related problems and give school nurses
Quality of Life in Healthy and Chronically Ill Icelandic Children
145
and other professionals opportunity to intervene appropriately. However, contrary to expectation, in this study on Icelandic pre-teenagers and their parents, lack of agreement was found between the pre-teenagers and their parents’ proxy-report on the children’s HRQOL. This finding emphasizes that for 10- to 12-year-old Icelandic healthy as well as chronically ill children, parents can only provide valid substitute report regarding their children’s emotional functioning and the Icelandic fathers can provide valid substitute report for their healthy children’s physical functioning. Nevertheless, parents’ proxy-report can add valid information regarding both the mothers’ and the fathers’ perspective on their pre-teenagers’ HRQOL and need therefore to be treated as such. Interestingly, in this study, single mothers were found to have a higher level of agreement with their pre-teenagers than mothers in dual parent families; and both mothers and fathers were found to have a higher level of agreement with their male pre-teenager than with their female pre-teenager. These findings indicate that within the Icelandic culture, family type and gender of the child influence the level of agreement between the pre-teenagers’ self-report and the parents’ proxy-report.
Summary Points Children’s health status influences developmental outcome and quality of life. Therefore,
early information regarding health related quality of life (HRQOL) can help identify health related problems. Lack of agreement between pre-teenagers’ self report and parents’ proxy report has been documented in the international literature. The instrument used in this study, ‘‘Pediatric Quality of Life Inventory’’ (PedsQL) developed by Varni et al. (2001, 2002), is a self-report intended for 8- to 12-year-old children has four subscales: physical functioning, emotional functioning, social functioning, and school functioning. Parents’ perspective on their children’s HRQOL was measured by a questionnaire identical to the children’s self-report (but addressed to the parents). The pre-teenagers were in their self-report found to differ significantly from their mothers’ and fathers’ proxies on the social and school functioning subscales of the HRQOL measure as well as on the overall HRQOL score. These findings emphasize that parents’ proxy report cannot be totally substituted for the pre-teenagers’ report on their own HRQOL. Children and their parents showed a high level of agreement on the subscale of emotional functioning within the HRQOL scale. Agreement was also found between fathers and their healthy children on the physical functioning subscale. Even though parents’ proxy report can only be substituted within the emotional and physical functioning subscale it can add needed and valid information regarding parents’ perspective on their pre-teenagers HRQOL.
References Annett RD, Bender BG, DuHamel TR, Lapidus J. (2003). J Asthma. 40: 577–587. Bothwell JE, Dooley JM, Gordon KE, MacAuley A, Camfield PR, MacSween J. (2002). Clin Pediatr. 41: 105–109. Chang PC, Yeh CH. (2005). Psychooncology. 14: 125–134.
Charron-Prochowinik D. (2002). J Pediatr Nurs. 17: 407–413. Chen X, Sekine M, Hamanishi S, Yamagami T, Kagamimori S. (2005). Child Care Health Dev. 31: 433–439. Cohen J. (1992). Psychol Bull. 112: 155–159.
2501
2502
145
Quality of Life in Healthy and Chronically Ill Icelandic Children
Dalheim-Englund AC, Rydstro¨m I, Rasmussen BH, Mo¨ller C, Sandman PO. (2004). J Clin Nurs. 13: 386–395. Epker J, Maddrey AM. (1998). Int J Rehabil Health. 4: 215–222. Faulkner MS. (2003). Pediatr Nurs. 29: 362–368. Faulkner MS, Chang LI. (2007). J Pediatr Nurs. 22: 59–68. Faulkner MS, Clark FS. (1998). Diabetes Educ. 24: 721–727. Fiese BH, Wamboldt FS, Anbar RD. (2005). J Pediatr. 146: 171–176. Guyatt GH, Juniper EF, Griffith LE, Feeny DH, Ferrie PJ. (1997). Pediatrics. 99: 165–168. Hays RM, Valentine J, Haynes G, Geyer JR, Villareale N, Mckinstry B, Varni JW, Churchill, SS. (2006). J Palliat Med. 9: 716–728. Janse AJ, Sinnema G, Uiterwaal CSPM, Kimpen JLL, Gemke RJBJ. (2005). Arch Dis Child. 90: 486–491. Knowles RL, Griebsch I, Bull C, Brown J, Wren C, Dezateux C. (2007). Arch Dis Child. 92: 388–393. Landolt MA, Grubenmann S, Meuli M. (2002). J Trauma. 53: 1146–1151. Le Coq EM, Boeke AJ, Bezemer PD, Colland VT, van Eijk JT. (2000). Qual Life Res. 9: 625–636. Markham C, Dean T. (2006). Int J Lang Commun Disord. 41: 189–212. Mednick L, Cogen FR, Streisand R. (2004). Child Health Care. 33: 169–183. Meuleners LB, Binns CW, Lee AH, Lower A. (2002). Child Care Health Dev. 28: 341–349. Noyes J. (2007). J Adv Nurs. 58: 1–10. Rydstro¨m I, Dalheim-Englund AC, Segesten K, Rasmussen BH. (2004). J Pediatr Nurs. 19: 85–94. Sheppard L, Eiser C, Kingston J. (2005). Child Care Health Dev. 31: 137–142.
Simeoni MC, Sapin C, Antoniotti S, Auquier P. (2001). J Adolesc Health. 28: 288–294. Svavarsdottir EK, Orlygsdottir B. (2006a). J Sch Nurs. 22: 178–185. Svavarsdottir EK, Orlygsdottir B. (2006b). Scand J Caring Sci. 20: 209–215. Tremolada M, Axia V, Pillon M, Scrimin S, Capello F, Zanesco L. (2005). J Pain Symptom Manage. 30: 544–552. Ungar WJ, Mirabelli C, Cousins M, Boydell KM. (2006). Soc Sci Med. 63: 2354–2366. Upton P, Maddocks A, Eiser C, Barnes PM, Williams J. (2005). Child Care Health Dev. 31: 409–415. Varni JW, Burwinkle TM, Rapoff MA, Kamps JL, Olson N. (2004). J Behav Med. 27: 297–318. Varni JW, Burwinkle TM, Seid M, Skarr D. (2003). Ambul Pediatr. 3: 329–341. Varni JW, Seid M, Kurtin PS. (2001). Med Care. 39: 800–812. Varni JW, Seid M, Knight TS, Uzark K, Szer IS. (2002). J Behav Med. 25: 175–193. Varni JW, Seid M, Rode CA. (1999). Med Care. 37: 126–139. Vila G, Hayder R, Bertrand C, Falissard B, de Blinc J, Mouren-Simeoni MC, Scheinmann P. (2003). Psychosomatics. 44: 319–328. Whitney B. (2005). Fam Community Health. 28: 176–183. Williams J, Williams K. (2003). Pediatr Pulmonol. 35: 114–118. Yeh CH, Chao, Hung LC. (2004a). Psychooncology. 13: 161–170. Yeh CH, Hung LC. (2003). Psychooncology. 12: 345–356. Yeh CH, Hung LC, Chao, KY. (2004b). Psychooncology. 13: 171–176.
146 Health-Related Quality of Life in Obese Children and Adolescents M. de Beer . R. J. B. J. Gemke 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2504
2
Concepts and Definitions of HRQOL Instruments for Children and Adolescents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2505 2.1 Generic HRQOL Measures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2505 2.2 (Disease) Specific Instruments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2507 3 Obesity and HRQOL in Children and Adolescents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2508 3.1 Summary of Effect of Interventions on HRQOL of Obese Children and Adolescents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2510 4
Directions for Further Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2513 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2514
#
Springer Science+Business Media LLC 2010 (USA)
2504
146
Health-Related Quality of Life in Obese Children and Adolescents
Abstract: Prosperity is accompanied by a rapidly increasing prevalence of overweight and obesity, especially among children and adolescents. This has major implications for the occurrence of cardiovascular (e.g., hypertension, stroke) and metabolic (e.g., insulin resistance and dyslipidemia) diseases in adulthood, not only regarding their prevalence but also as they increasingly occur at a relatively younger (adult of even adolescent) age. Obesity is not only responsible for these slowly and sub clinically presenting somatic conditions but is also closely associated with development of adverse psychological and social conditions. This burden of obesity has both short and long term effects while children and adolescents are particularly vulnerable for adverse psychosocial consequences. Health related quality of life is a concept that enables to assess overall well-being in large groups of healthy subjects and those with a (acute or chronic) condition. In this chapter the concepts of measuring health related quality of life in children and adolescents will be reviewed followed by the introduction and discussion of generic and condition (obesity) specific measures for assessment of health related quality of life. Subsequently the relation of obesity with (impairment of) health related quality of life will be reviewed by a number of relevant studies. Rather than to provide an exhaustive review we sought to provide a representative sample of studies that (1) were methodologically sound; (2) comprised an adequate sample size; (3) comprised individuals across a broad range of relative body weight (body mass index, > BMI [kg/m2];) and (4) used generally well-validated measures of HRQL. Thereafter the effects of interventions on health related quality of life in obese children and adolescents will be discussed and finally a perspective for future research on this topic is proposed. List of Abbreviations: BMI, body mass index; CHQ, child health questionnaire; CHQ-CF, child health questionnaire child form; CHQ-PF, child health questionnaire parent form; HRQOL, health related quality of life; HUI, health utilities index; IWQOL, impact of weight on quality of life; KINDL, kinderlebensqualita¨t fragenbogen; MHS, municipal health service; PedsQL, pediatric quality of life inventory; QOL, quality of life; SF-36, short form 36; SG, standard gamble; TACQOL, TNO-AZL children quality of life; TAPQOL, TNO-AZL preschool children quality of life; TTO, time trade-off; VAS, visual analogue scales
1
Introduction
In the past decade the increasing prevalence of children and adolescents with overweight or obesity has reached alarming proportions (Hedley et al., 2004; Van den Hurk et al., 2007). Although currently this is primarily a problem in the Western world, a huge expansion in the developing countries, especially in those with a rapidly emerging market economy is expected. There is growing awareness of the long term health consequences of this condition, not only in adults, but also in children (Ten and Maclaren, 2004; Weiss et al., 2004). Particularly type 2 diabetes and cardiovascular diseases, developing at an early age are among the most important obesity associated comorbid conditions (Weiss et al., 2004). It has increasingly become clear that the problems associated with obesity are not restricted simply to causing or exacerbating medical conditions. Obesity also appears to have a substantial impact on a person’s functional capacity and quality of life. It is important to know more about quality of life in obesity in order to understand which impairments of function and well-being are associated with this condition and what kind of needs for medical and/or psychosocial care may emerge. Also, assessment of changes in quality of life can be used as an outcome measure to evaluate the effects of interventions.
Health-Related Quality of Life in Obese Children and Adolescents
146
> Health-related quality of life (HRQOL) is a multidimensional construct of an individual’s subjective evaluation of his/her health, encompassing physical (e.g., sensory and locomotor function), psychological (e.g., emotional functioning) and social domains (e.g., occupational or, in children and adolescents, school functioning) (Schipper et al., 1996). HRQOL is commonly measured using instruments including several domains and/or (sub)scales and/or items. In these instruments usually an item is a single question, a scale contains the available categories for expressing the response to the question, a domain (also called dimension or attribute) identifies a particular focus of attention and may comprise the response to a single item or to several related items, and an instrument is the aggregated collection of all items. In this chapter, we will review the current knowledge on the impact of obesity on HRQOL and briefly address the effect of interventions (e.g., weight reduction) on the HRQOL of obese adolescents.1 We will also discuss issues regarding the assessment of HRQOL in obese youngsters, and highlight potential directions for future research.
2
Concepts and Definitions of HRQOL Instruments for Children and Adolescents
Assessment of HRQOL in children and adolescents poses unique problems. Especially at younger age, rapid changes in physical and mental performance occur as part of the normal development in children. In contrast to adolescents, whose ability to respond to questionnaires is similar to adults, assessment of HRQOL in younger children needs special attention and requires separate instruments that (also) allow for proxies (parents) as respondents. The ability of younger children to use rating scales, understand the language, and generally complete lengthy questionnaires of the type used in adult work, is largely affected by age and cognitive development. Children are often regarded as unreliable respondents, and for this reason, early attempts to rate children’s QOL were solely based on data provided by mothers. However, children and parents do not necessarily share similar views about the impact of illness, and therefore there are appropriate calls to involve children more directly in decisions about their own care and treatment (Schipper et al., 1996). As a consequence, any evaluation of current approaches to measuring children’s QOL needs to consider the provision made for children to rate their own QOL (Eiser and Jenney, 1996; Eiser and Kopel, 1997). There are two ways in which HRQOL can be assessed, namely with generic and with disease specific instruments.
2.1
Generic HRQOL Measures
Generic instruments are measures that attempt to assess all important aspects of HRQOL. They can be applied to a variety of populations allowing for broad comparisons of the relative impact of health and disease. They are classified into health profiles and health indexes (or utilities). Health profiles consist of multiple items that are grouped together into specific domains of health and functioning, providing patients, clinicians and researchers with information on the
1
Refer to chapter by Jonda et al. Implementing interventions to enhance QOL in overweight children and adolescents
2505
2506
146
Health-Related Quality of Life in Obese Children and Adolescents
impact of diseases or treatments on different aspects of QOL. In addition, summary scores may been derived for some health profile instruments by averaging across domains. For adults several robust generic instruments are available in numerous languages. The Medical Outcomes Study 36-Item Short Form (SF36, developed in 1992) is today probably the most widely used generic health status questionnaire around the world for adults (Ware, 1996). Although in adolescents generic health status measures for adults may (also) be used, for younger children health status and HRQOL measures are still in a developmental stage (Eiser and Morse, 2001). At present the following questionnaires are available for younger children. The Child Health and Illness Profile-adolescent edition,(Starfield et al., 1993) Child Health Questionnaire,(Landgraf et al., 1996) Functional Status II-R,(Stein and Jessop, 1990) KINDL, (Ravens Sieberer and Bullinger, 1998) TACQOL,(Vogels et al., 1998) TAPQOL,(Fekkes et al., 2000) the Warwick Child Health and Morbidity Profile,(Spencer and Coe, 1996) Health Utilities Index (HUI) mark 2 and 3 (Boyle et al., 1995). The Child Health Questionnaire (CHQ) is a frequently used instrument which has the same structure as the SF-36 does, but has been developed specifically for children and adolescents (Landgraf et al., 1996). Preference-based measures of HRQOL particularly the HUI2 and HUI3 that have been developed especially for children provide a single overall score representing the net aggregate impact of physical, emotional and social functioning on QOL. Traditionally the preference or utility scale in health extends from 1 (perfect health) to 0 (death) (Boyle et al., 1995). Preference scores are advantageous in the application of economic evaluations of the costutility of alternative programs. Two main types of preferences exist: values and utilities. Values are preferences measured under certainty (meaning that there is no risk or uncertainty in the preference measurement question). The time trade-off (TTO) and the visual analogue scales (VAS) are preference measurement instruments that produce values. Utilities are preferences measured under uncertainty (there is risk or probability involved in the preference based measure). The standard gamble (SG) is a measurement that produces utilities. The SG resembles a kind of structured lottery in which the respondent is presented two alternatives, one of which is uncertain. The probability p is varied systematically until the respondent is indifferent between the two alternatives i.e., based on the probability p, the utility for the health state in each alternative can be calculated (Bennett and Torrance, 1996; Boyle et al., 1995; Apajasalo et al., 1996a,b) (> Table 146-1).
. Table 146-1 Commonly used generic HRQOL measures Child health and illness profile-adolescent edition Child health questionnaire Functional status II-R KINDL kinderlebensqualita¨t fragenbogen TACQOL TNO-AZL children quality of life, TAPQOL TNO-AZL pre-school children quality of life The warwick child health and morbidity profile Health utilities index (HUI) mark 2 and 3 Pediatric quality of life inventory
Health-Related Quality of Life in Obese Children and Adolescents
146
Perhaps the best known and validated (Eiser and Morse, 2001) generic HRQOL instruments for children and adolescents are the Child Health Questionnaire (CHQ) (Landgraf et al., 1996) and the Pediatric Quality of Life Inventory (PedsQL) (Varni et al., 1999). These measures are applicable for both children (including proxy versions for young children) and adolescents in a wide variety of populations and allow for comparisons across a diversity of medical conditions. The CHQ Child Form 87 (CHQ-CF87) questionnaire encompasses 12 domains of which each item contains 4, 5 or 6 response alternatives. Per scale the items are summed up (some recoded/recalibrated) and transformed to a 0 (worst possible score) to 100 (best possible score) scale (Landgraf et al., 1996). Also parental (proxy) versions of the CHQ are available (CHQ-PF50 and CHQ-PF28, respectively) for application in younger children. The 23-item PedsQL 4.0 questionnaire encompasses physical functioning (8 items), emotional functioning (5 items), social functioning (5 items) and school functioning (5 items). A 5-point Likert scale is used for response (0 = never a problem; 4 = almost always a problem). Items are reversely scored and, if appropriate, linearly transformed to a 0–100 scale, so that higher scores indicate better HRQOL. A total scale score, a physical summary score and a psychosocial health summary score may be calculated (Varni et al., 1999). Commonly used generic HRQOL measures are listed in > Table 146-1.
2.2
(Disease) Specific Instruments
(Disease) specific HRQOL measures focus on aspects of health status that are specific to the area of primary interest. The instruments may be specific to a condition or disease (e.g., obesity, asthma, heart failure), to a population, to a certain function or to a problem. The major advantage of (disease) specific measures is that they are frequently more responsive to changes after treatment than generic instruments (Guyatt et al., 1993, 2007). Another feature is that specific measures have the advantage of relating closely to areas routinely analyzed by physicians. Until recently there was no disease-specific HRQOL measurement available for obese children or adolescents. In 2006 Kolotkin et al. (2006) developed a measure of weightrelated quality of life for adolescents, the Impact of Weight on Quality of Life (IWQOL)-Kids. This questionnaire consists of 27 items which were modeled after weight-related quality of life measures in adults, namely the IWQOL-Lite (Kolotkin and Crosby, 2002) and the original (adult) IWQOL (Kolotkin et al., 1997). Four factors were identified: physical comfort (6 items), body esteem (9 items), social life (6 items), and family relations (6 items). Response options range from always true (1) to never true (5). Preliminary analyses provide support for the measures strong psychometric properties, discrimination among BMI groups and between clinical and community samples, and responsiveness to a weight loss/social support interaction (Kolotkin et al., 2006) (> Table 146-2).
. Table 146-2 Commonly used condition (obesity) specific HRQOL measures Impact of weight on quality of life (IWQOL)-adults Impact of weight on quality of life (IWQOL)-kids Impact of weight on quality of life (IWQOL)-lite
2507
2508
146 3
Health-Related Quality of Life in Obese Children and Adolescents
Obesity and HRQOL in Children and Adolescents
So far, a limited number of studies assessing HRQOL in juvenile obesity have been performed. Probably due to a relation between the severity (and obviousness) of obesity and physical and social problems, a negative association between HRQOL and BMI has been described in adults, which may also apply to adolescents and to a lesser extend to children (Kortt and Clarke, 2005; Lee et al., 2005). In clinical studies on chronic health problems, including juvenile obesity, HRQOL is an important indicator of outcome complementary to clinical (e.g., biometric) and laboratory (e.g., biochemical) markers. The published studies concerning this issue are briefly discussed below and also summarized in > Table 146-1. We did not intend to provide an exhaustive review of all published studies that have estimated the effect of obesity on HRQOL in children and adolescents. Rather, we sought to provide a representative sample of studies that (1) were methodologically sound; (2) comprised an adequate sample size; (3) comprised individuals across a broad range of relative body weight (body mass index, BMI [kg/m2];) and (4) used generally well-validated measures of HRQOL. Studies which exclusively address younger children will not be discussed here. A longitudinal study published in 2001 put forward that obese children and adolescents (n = 584, mean age 12 years) experience more restriction in quality of life compared to children and adolescents with asthma and atopic dermatitis (Ravens-Sieberer et al., 2001). Of the total score, 37% of the variance was explained by stress level, coping, lack of emotional support and poor global health. However, it was unclear which criteria were used for the assessment of obesity and whether possible co-morbid diseases were accounted for. Furthermore a comparison with a normal weight control group is lacking. Therefore the generalizability of these findings is limited. Nine studies have used the Pediatric Quality of Life inventory (PedsQL) or a derived measure to assess self-report and/or parent–proxy report of HRQOL in obese youth. In 2003, Schwimmer et al. reported a decrease in all domains of PedsQL (physical, emotional, social and school), in a group of 106 severely obese children and adolescents aged 5–18 years (Schwimmer et al., 2003). In this study, obese children and adolescents appeared to have a similar reduction in HRQOL as those with cancer. The confounding factors that were adjusted for (age, gender, socio-economic status and ethnicity) were not accountable for these findings. BMI z-score among obese children and adolescents was inversely associated with physical functioning. Furthermore obesity-related comorbid conditions were not responsible for differences in HRQOL (Schwimmer et al., 2003). Comparable impairments in all domains of functioning were found by Zeller et al. (Zeller and Modi, 2006). In a clinical study among 166 children and adolescents aging 8–18 years they specifically addressed possible racial differences between obese African Americans and Caucasians. No differences between these groups were found. Furthermore, similar to findings from Schwimmer et al. (2003) they found discrepancies between parents and children in reported HRQOL; parents reported significantly worse HRQOL than their children across many dimensions, especially on emotional and social domains. The authors put forward that maternal distress, which is often reported by mothers of treatment-seeking obese youth, can be a possible explanation for this finding. Furthermore, previous research suggests that parents and youth tend to show greater agreement for observable (‘‘objective’’) behavior than internal (‘‘subjective’’) states reflected in reported emotional and social domains (Matza et al., 2004). Again BMI z-score was inversely related with physical and social HRQOL. Furthermore, the
Health-Related Quality of Life in Obese Children and Adolescents
146
study elucidated robust relations among impaired HRQOL, depressive symptoms, lower perceived social support, and degree of overweight. The strongest predictors of HRQOL for obese youth were depressive symptoms and perceived social support from class mates. Thus, obese youth who feel supported by their school-based peer group and who are less depressed may have better HRQOL (Zeller and Modi, 2006). In 2007 Janicke et al. (2007) demonstrated an association between increased parent distress, and peer victimization with lower HRQOL in overweight youth, aged 8–17 years. Depressive symptoms mediated the relationships between these (potential) external risk factors and child-reported HRQOL. Hence, the authors emphasize the important impact of those elements. Incorporating peer support in treatment for example, may improve weight loss in overweight youth and moreover interventions which address parents’ personal coping and distress will likely be beneficial in the treatment of their overweight children (Jelalian et al., 2006). In contrast with the studies mentioned earlier, BMI z-score was not related to HRQOL, indicating that the degree of overweight made no difference in perceived HRQOL. Similar to findings of Zeller et al. (2006), there was no relationship between weight status and HRQOL within different racial groups (Janicke et al., 2007). Another study (Ingerski et al., 2007), with partially the same patient population as that in the previously mentioned study, (Janicke et al., 2007) extends the findings of Zeller et al. (Zeller and Modi, 2006) and Janicke et al. (Janicke et al., 2007). Taken together, these studies suggest that social support is especially important to obese youth’s HRQOL. Again no relation between the degree of overweight (as measured by BMI z-score) with HRQOL was found in this study (Ingerski et al., 2007). By the same token Stern et al. (2007) found that both teasing (partially mediated by self esteem) and low self esteem were associated with obesity in a clinical population. The main focus of this study was to investigate the influence of gender and race on psychosocial factors and dietary habits associated with pediatric overweight. Few gender or ethnic differences on psychosocial variables were found (Stern et al., 2007). Results from a study from Doyle et al. suggest that overweight adolescents at high risk for the development of eating disorders also experience elevated levels of negative affect, impairment in HRQOL (physical, emotional and social aspects), and eating disturbances. (Doyle et al., 2007). In a group of 33 extremely obese adolescents who presented for evaluation of bariatric surgery marked decreases in HRQOL across all domains were found (Zeller et al., 2006). The level of impairment characterizing this population was worse than that in less extreme obese subjects (Schwimmer et al., 2003). The authors speculate that the severity of HRQOL impairment in these extremely obese adolescent was due to of the visible nature of extreme obesity, the social stigma attached to it, and the cumulative impact of medical comorbidities. Furthermore, their analysis revealed gender differences in HRQOL indicating that obese adolescent girls may be at particular risk in the social domain (Zeller et al., 2006). This gender difference was confirmed in several studies (Ingerski et al., 2007; Janicke et al., 2007; Jelalian et al., 2006; Matza et al., 2004; Schwimmer et al., 2003; Stern et al., 2007; Zeller and Modi, 2006) although others failed to demonstrate these differences (Zeller and Modi, 2006).. A cross-sectional study of children and adolescents from community pediatrics clinics (39 obese and 94 normal weight controls) and a hospital-based obesity clinic (49 children), showed significant differences in physical and social domain scores even in moderate obesity (Pinhas-Hamiel et al., 2006). Both physical and social domain scores decreased progressively with increased BMI z-scores. The perceived HRQOL of obese children treated
2509
2510
146
Health-Related Quality of Life in Obese Children and Adolescents
in a hospital setting was similar to that of obese children in the community (Pinhas-Hamiel et al., 2006). In a series by the authors of this chapter, HRQOL was measured by the PedsQL and CHQ, among 31 obese adolescents (Mean BMI = 34.9 kg/m2) in weight-loss program in a tertiary centre (De Beer et al., 2007). For every patient, two age and sex matched controls, with normal weight for height, were enrolled. These controls were randomly selected from adolescents at their regular visit to the preventive school physician from the regional Municipal Health Service (MHS). Substantial differences in HRQOL were found between obese and control adolescents, mainly in the physical and social domains (De Beer et al., 2007). Furthermore we found an inverse relation between BMI z-score and HRQOL. Obesity-related comorbidity appeared to be an important explanatory variable in the impairments found in physical and social domains, experienced by obese adolescents. This is in contrasts with findings by Schwimmer et al. (2003) and Zeller et al. (2006). Based on cross-sectional data from a sample of 110 treatment seeking extremely overweight adolescents and 34 normal weight controls, Fallon and al. found a decreased reported HRQOL regarding Social/Interpersonal, Self-esteem, and Daily living HRQOL similar to adults seeking weight loss treatment (Fallon et al., 2005). Interestingly, overweight had a greater impact among heavier whites, compared with blacks, with regard to social and psychological well-being, aspects of daily living, health efficacy, and physical appearance. This is the first study among overweight adolescents which used the disease specific instrument IWQOL-A, and therefore may have been more sensitive to distinctive impairments experienced by obese youth (Fallon et al., 2005). The only population based study among adolescents was published in 2005 by Swallen et al. (Swallen et al., 2005). Using a US nationally representative sample of 4,743 adolescents, they demonstrated that obesity in > adolescence is linked with poor physical HRQOL. In the general population, adolescents with above normal body mass did not report poorer emotional, school, or social functioning (Swallen et al., 2005). In summary, studies attempting to estimate the effect of obesity on HRQOL in children and adolescents indicate that: (1) obese children and adolescents report significantly impaired HRQOL; (2) obesity appears to have a greater impact upon physical and social functioning than on mental health domains; (3) there appears to be a relevant association between BMI and the degree of HRQOL impairment; (4) race may have a modulating effect that needs to be further examined; (5) social support is especially important to HRQOL in children and adolescents; and (6) obesity-related comorbid conditions may have a contribution to the negative effect obesity seems to have on HRQOL (> Table 146-3).
3.1
Summary of Effect of Interventions on HRQOL of Obese Children and Adolescents
Until now, most studies which evaluate the effect of obesity treatment focus on the effect of interventions on children’s and adolescent self esteem. Only few studies used HRQOL as an outcome measure, these studies are briefly reviewed here. A multidisciplinary behavioral program, published in 1990, produced significant weight losses in black female adolescents, which were associated with improvement in psychological status (Wadden et al., 1990). End-of-treatment scores of participating subjects on standardized tests indicated increased self-esteem and decreased feelings of depression (Wadden et al., 1990).
Health-Related Quality of Life in Obese Children and Adolescents
146
. Table 146-3 Summary of relevant studies assessing HRQOL in children overweight or obesity References
Study and sample
HRQOL measure
General finding
RavensSieberer et al. (2001)
Longitudinal study of 584 obese children and adolescents (mean age: 12 years)
German KINDLR
Decreased HRQOL in all domains (except physical functioning)
Schwimmer et al. (2003)
Cross-sectional data from clinically severe obese children and adolescents (5–18 yrs, n = 106)
PedsQL 4.0
Decreased HRQOL in all domains (physical, emotional, social and school domains)
Swallen et al. (2005)
Community based, cross-sectional study of 4,743 adolescents
Derivative measure of PedsQL 4.0
Decreased HRQOL in physical domain
Fallon et al. (2005)
Cross-sectional study of 110 overweight adolescents and 34 nonoverweight controls
IWQOL-A
Overweight is associated with poorer HRQOL, regardless of race
Zeller et al. (2006)
Cross-sectional data of 33 extremely PedsQL 4.0 obese adolescents, presenting for evaluation at a bariatric surgery program
HRQOL CHQ-PF50
Marked impairments in HRQOL across all domains
PinhasCross-sectional study of 182 Hamiel et al. children and adolescents from (2006) community pediatrics clinics and a hospital-based obesity clinic
PedsQL 4.0
Obese children had lower HRQOL scores in the physical, social and school domains. Severity of obesity affected the pattern of the HRQOL scores. No difference between community and clinical obese children
Zeller et al. (2006)
Cross-sectional study, children and adolescents of 8–18 yrs (n = 166)
PedsQL 4.0
Depressive symptoms, perceived social support from classmates, degree of overweight, and socioeconomic status seem tot be strong predictors of HRQOL
Janicke et al. Cross-sectional data, 96 at-risk-for(2007) overweight and overweight youth, 8–17 years of age
PedsQL 4.0
Increased parent distress, child depressive symptoms, and peer victimization were associated with lower HRQOL
Ingerski et al. (2007)
Cross-sectional study of 107 clinically overweight youth, ages 12–17 years
PedsQL 4.0
HRQOL obese adolescents lower than in healthy children. Degree of overweight was not related to HRQOL
Stern et al. (2007)
Cross-sectional study of 100 treatment seeking overweight adolescents
PedsQL 4.0
HRQOL among overweight adolescents is associated with both teasing and low selfesteem. The relation between teasing and HRQOL is partially mediated by self-esteem
2511
2512
146
Health-Related Quality of Life in Obese Children and Adolescents
. Table 146-3 (continued) References
Study and sample
HRQOL measure
General finding
Doyle et al. (2007)
Cross-sectional study of 81 treatment seeking adolescents
PedsQL 4.0
Overweight adolescents at high risk for the development of eating disorders also experience impairment in physical, emotional and social aspects of HRQOL
De Beer et al. (2007)
Cross-sectional study of 31 treatment seeking adolescents, compared to 62 matched normalweight controls
PedsQL 4.0, Substantial differences between CHQ CF-87 obese and normal weight adolescents in physical and social domains, partially explained by obesity-related comorbidity
CHQ-CF child health questionnaire child form; CHQ-PF child health questionnaire parent form; HRQOL healthrelated quality-of-life; IWQOL impact of weight on quality of life; KINDL kinderlebensqualita¨t fragenbogen); PedsQL pediatric quality of life inventory
Another treatment protocol which incorporated peer social support proved to be beneficial for weight loss and its maintenance in adolescents (Jelalian and Mehlenbeck, 2002). Group activities included both mental and physical challenges that fostered development of trust, social skills and self-confidence. Measures of height and weight, as well as questionnaires assessing self-concept, physical self-worth, and social functioning, were obtained prior to treatment, immediately following the 16-week intervention, and 6 months after completion of active treatment. They observed significant decreases in percent overweight during the course of the intervention. Furthermore, decreases in weight were accompanied by improvements in self-confidence related to physical appearance and physical self-worth. These findings provided some preliminary support for the application of a peer-based program as an adjunctive treatment for adolescent weight management intervention (Jelalian and Mehlenbeck, 2002). In a large systematic review a comprehensive assessment of pediatric weight management programs was conducted to evaluate the impact of these programs on child and adolescent self-esteem (Walker Lowry et al., 2007). The results of this review suggest overall positive effects on self esteem by pediatric weight management programs across a variety of settings and common treatment components, although the authors question the methodological soundness of the studies. They assume that components related to self-esteem improvements include weight status change, consistent parent involvement, the use of a peer group format to target self-esteem and develop positive peer interactions. The authors put forward that studies indicate that certain components of self-esteem may be affected first (such as body image) and subsequently lead to global self-esteem improvements (Walker Lowry et al., 2007). According to Ravens-Sieberer et al. in-patient rehabilitative treatment for obesity in children is associated with an increase in HRQOL. After a rehabilitation program, psychosocial factors explained 28% of the variance compared to 37% before treatment (Ravens-Sieberer et al., 2001). In a 6-month, randomized, double-bind, placebo-controlled trial, the use of sibutramine in 40 obese Mexican adolescents was evaluated (Garcı´a-Morales et al., 2006). HRQOL was
Health-Related Quality of Life in Obese Children and Adolescents
146
assessed at the study start and end using the 36-item Short-Form Health Survey (SF-36) questionnaire. Interestingly, no differences were found between the sibutramine and the placebo group; both groups showed significant increases in total HRQOL scores (Garcı´aMorales et al., 2006). As seen in the majority of studies in adult populations, weight reduction induced surgically, produced dramatic improvements in the majority of HRQOL indices (Fontaine and Barofsky, 2001). However, it is unclear whether weight reduction would have beneficial effects on HRQOL among persons who have lower degrees of obesity and participate in less invasive forms of weight-loss treatment. Inge et al. performed early observations in a sample of 16 extremely obese adolescents (mean BMI 59.9 kg/m2) presenting for bariatric surgery, which revealed that at 12-months post surgery adolescents experience significant improvement in HRQOL and depressive symptomatology despite continued obesity (mean BMI 36.9 kg/m2) (Inge et al., 2007). The authors suggest that these psychosocial changes may substantively alter the psychological health and socioeconomic trajectory of the adolescent to a greater extent than if surgery is performed later in adulthood (Inge et al., 2007). In summary, studies attempting to estimate the effect of intervention in obese children and adolescents indicate on HRQOL that: (1) obese adolescents report significantly improved self esteem and less feelings of depression after obesity treatment; (2) peer social support appears to be additionally advantageous; (3) HRQOL significantly improves after various types of obesity treatment (e.g., mediation or placebo based, surgery); (4) obesity treatment should commence early, preferably well before adulthood, because the negative impact on psychological health and the socioeconomic prospects can then still be modified.
4
Directions for Further Research
In most studies, groups of children and adolescents seeking treatment were assessed. It should be realized that these are selective samples of adolescents that usually are highly motivated to achieve weight loss through therapy, which may have been driven by their level of psychosocial distress. Whether these psychosocial difficulties also characterize children and adolescents with obesity who are not currently in treatment or seeking methods to lose weight remains unknown. Studies which assessed children in an unselected population setting suggest a lesser degree of impairments in the various HRQOL domains (Wake et al., 2002; Williams et al., 2005). The cross-sectional nature of all studies performed up till now precludes conclusions regarding causality, and therefore, their generalizability is limited. Prospective longitudinal studies examining the impact of overweight status on children and adolescents may help to further clarify the relationship between weight, BMI and HRQOL. Furthermore, so far only few studies have used the overweight-specific HRQOL measures, particularly the recently developed IWQOL. We would recommend future research to use this or similar condition specific measures, complementary to generic measures. While generic measures of HRQOL allow for comparison with other chronic conditions, they usually lack the sensitivity to detect differences in HRQOL in cross sectional or longitudinal analysis of an exclusively overweight sample. Using both measures, they can complement each other. Comparison of the relation of obesity with HRQOL among different countries and crosscultural validation of condition (i.e., obesity) specific HRQOL measures are also important topics for future research.
2513
2514
146
Health-Related Quality of Life in Obese Children and Adolescents
. Table 146-4 Summary of directions for future research on obesity and HRQOL Focus on longitudinal instead of cross-sectional studies Usage of condition (obesity) specific HRQOL measures Cross cultural comparison of impact of obesity on HRQOL Attention for (long term) efficacy and cost-effectiveness of interventions on HRQOL
Finally, a lot of attention is directed at various available treatment options, such as cognitive behavioral therapy (with or without peer-based support), drug treatment, and (bariatric) surgical treatment. Well-designed, prospective studies, which evaluate the efficacy, (cost)effectiveness and safety of these weight loss interventions are mandatory, particularly to assess which therapy is most appropriate for an individual obese child or adolescent (> Table 146-4).
Summary Points Obese children and adolescents report significantly impaired HRQOL. Obesity appears to have a greater impact upon physical and social functioning than on mental health domains.
A relevant association between BMI and the degree of HRQOL impairment has been described.
Race may have a modulating effect on the relation between BMI and HRQOL. Social support is especially important to HRQOL in children and adolescents. Obesity-related comorbid conditions may aggravate the negative effects of obesity on HRQOL.
References Apajasalo M, Rautonen J, Homberg C. (1996a). Qual Life Res. 5: 532–538. Apajasalo M, Sintonen H, Holmberg C, et al. (1996b). Qual Life Res. 5: 205–211. Bennett K, Torrance GW. (1996). Measuring health state preferences and utilities: rating scale, time trade-off, and standard gamble techniques. Spilker B (ed.) Quality of Life and Pharmacoeconomics in Clinical Trials. Lippincott-Raven, Philadelphia, PA, pp. 253–265. Boyle MH, Furlong W, Feeny D, et al. (1995). Qual Life Res. 4: 249–257. De Beer M, Hofsteenge GH, Koot HM, HiraSing RA, Delemarre-van de Waal HA, Gemke RJBJ. (2007). Acta Paediatr. 96: 710–714. Doyle AC, Le Grange D, Goldschmidt A, Wilfley DE. (2007). Obesity (Silver Spring). 15: 145–154.
Eiser C, Jenney MEM. (1996). BR J Cancer. 73: 1313–1316. Eiser C, Kopel SJ. (1997). Children’s perception of health ad illness. In: Petrie KJ, Weinman JA, (ed.) Perceptions of Health and Illness: Current Research and Applications. Harwood Academic Publishers, Singapore. Eiser C, Morse R. (2001). Arch Dis Child. 84: 205–211. Fallon EM, Tanofsky-Kraff M, Norman AC, McDuffie JR, Taylor ED, Cohen ML et al. (2005). J Pediatr. 147: 443–450. Fekkes M, Theunissen NC, Brugman E, Veen S, Verrips EG, Koopman HM, Vogels T, Wit JM, ad VerlooveVanhorick SP. (2000). Qual Life Res. 9: 961–972. Fontaine KR, Barofsky I. (2001). Obes Rev. 2: 173–182. Garcı´a-Morales LM, Berber A, Macias-Lara CC, LucioOrtiz C, Del-Rio-Navarro BE, Dorantes-Alva´rez LM. (2006). Clin Ther. 28: 770–782.
Health-Related Quality of Life in Obese Children and Adolescents Guyatt GH, Feeny DH, Patrick DL. (1993). Ann Int Med. 118: 622–629. Guyatt GH, Ferras CE, Halyard MY, Revicki DA, Symonds TL, Varricchio CG et al. (2007). Mayo Clin Proc. 82: 1229–1239. Hedley AA, Ogden CL, Johnson CL, Carroll MD, Curtin LR, Flegal KM. (2004). JAMA. 291: 2847–2850. Inge TH, Xanthakos SA, Zeller MH. (2007). Int J Obes. 31: 1–14. Ingerski LM, Janicke DM, Silverstein JH. (2007). J Pediatr Psychol. 32: 869–874. Janicke DM, Marciel KK, Ingerski LM, Novoa W, Lowry KW, Sallinen BJ et al. (2007). Obesity (Silver Spring). 15: 1799–1807. Jelalian E, Mehlenbeck R. (2002). J Clin Psychol Med Settings. 9: 15–23. Jelalian E, Mehlenbeck R, Lloyd-Richardson EE, Birmaher V, Wing RR. (2006). Int J Obes (Lond). 30: 31–39. Kolotkin RL, Crosby RD. (2002). Qual Life Res. 11: 157–171. Kolotkin RL, Head S, Brookhart A. (1997). Obes Res. 5: 434–441. Kolotkin RL, Zeller MH, Modi AC, Samsa GP, Quinlan NP, Yanovski JA et al. (2006). Obesity (Silver Spring).14: 448–457. Kortt MA, Clarke PM. (2005). Qual Life Res. 14: 2177–2185. Landgraf JMJM, Abetz L, Ware JE. (1996). The CHQ User’s Manual. The Health Institute, New England Medical Center, Boston, MA. Lee AJ, Morgan CL, Morrissey M, Wittrup-Jensen KU, Kenedy-Martin T, Currie CJ. (2005). Diabet Med. 22: 1482–1486. Matza LS, Swensen AR, Flood EM, Secnik K, Leidy NK. (2004). Value Health. 7: 79–92. Pinhas-Hamiel O, Singer S, Pilpel Fradkin A, Moda D, Reihman B. (2006). Int J Obes (Lond). 30: 267–272. Ravens Sieberer U, Bullinger M. (1998). Qual Life Res. 7: 399–407. Ravens-Sieberer U, Redegeld M, Bullinger M. (2001). Int J Obes Relat Metab Disor. 25 (Suppl 1): S63–S65. Schipper H, Clich JJ, Olweny CL. (1996). Quality of life studies: definitions and conceptual issues. In:
146
Spilker B (ed.) Quality of Life and Pharmacoeconomics in Clinical Trials, 2nd ed. Lippincott-Raven, Philadelphia, PA, pp. 11–23. Schwimmer JB, Burwinkle TM, Varni JW. (2003). JAMA. 289: 1813–1819. Spencer NJ, Coe C. (1996). Child Care Health Dev. 22: 367–379. Starfield B, Bergner M, Ensminger M, et al. (1993). Pediatrics. 91: 430–435. Stei RE, Jessop DJ. (1990). Med Care. 28: 1041–1055. Stern M, Mazzeo SE, Gerke CK, Porter JS, Bean MK, Laver JH. (2007). J Pediatr Psychol. 32: 90–94. Swallen KC, Reither EN, Haas SA, Meier AM. (2005). Pediatrics. 115: 340–347. Ten S, Maclaren N. (2004). J Clin Endocrinol Metab. 89: 2526–2539. Van den Hurk K, van Dommelen P, van Buuren S, Verkerk PH, Hirasing RA. (2007). Arch Dis Child. 92: 992–995. Varni JW, Seid M, Rode CA. (1999). Med Care. 37: 126–139. Vogels T, Verrips GH, Verloove-Vanhorick SP, et al. (1998). Qual Life Res. 7: 457–465. Wadden TA, Stunkard AJ, Rich L, Rubin CJ, Sweidel G, McKinney S. (1990). Pediatrics. 85: 345–352. Wake M, Salmon L, Waters E, Wright M, Hesketh K. (2002). Int J Obes Relat Metab Disord. 26: 717–724. Walker Lowry K, Sallinen BJ, Janicke DM. (2007). J Pediatr Psychol. 32: 1179–1195. Ware JEJ. (1996). The SF-36 health survey. In: Spilker B (ed.) Quality of Life ad Pharmacoeconomics in Clinical Trials. Lippincott-Raven, Philadelphia PA, pp. 337–345. Weiss R, Dziura J, Burgert TS, Tamborlane WV, Taksali SE, Yeckel CW et al. (2004). N Engl J Med. 350: 2362–2374. Williams J, Wake M, Hesketh K, Maher E, Waters E. (2005). JAMA. 293: 70–76. Zeller MH, Modi AC. (2006). Obesity (Silver Spring). 14: 122–130. Zeller MH, Roehrig HR, Modi AC, Daniels SR, Inge TH. (2006). Pediatrics. 117: 1155–1561.
2515
147 Implementing Interventions to Enhance Quality of Life in Overweight Children and Adolescents J. Lamanna . N. Kelly . M. Stern . S. E. Mazzeo 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2518
2
Quality of Life . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2518
3 3.1 3.1.1 3.1.2 3.1.3 3.2 3.2.1 3.2.2 3.2.3 3.2.4 3.3 3.3.1 3.3.2 3.3.3
Specific QOL Improvements Associated with POI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2519 Physical QOL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2519 Weight Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2519 Physiological Functioning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2520 Physical Fitness Improvements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2521 Psychological QOL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2521 Depression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2521 Self-Esteem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2521 Eating Behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2522 Body Perceptions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2523 Social QOL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2523 Peer Relationships . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2523 Teasing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2524 Family Relationships . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2524
4 4.1 4.2 4.3 4.4
Components of Programs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2524 Physical Activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2524 Nutrition Counseling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2525 Behavioral Modification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2526 Parent Involvement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2526
5 5.1 5.2
Multidisciplinary POI Programs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2527 Outpatient Intervention . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2527 Inpatient Intervention . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2527
6
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2531 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2534
#
Springer Science+Business Media LLC 2010 (USA)
2518
147
Implementing Interventions to Enhance QOL in Overweight Children and Adolescents
Abstract: The rates of pediatric overweight in the United States and in other industrialized countries have risen dramatically in recent years. Pediatric overweight is associated with poor physical, psychological, and social Quality of Life (QOL). Overweight children and adolescents are at greater risk for weight-related health problems, emotional and psychological problems, and social stigmatization than their normal-weight peers. To address the pediatric overweight crisis and to improve QOL in this population, a number of interventions have been developed. Rather than focusing on weight loss, interventions tend to encourage the development of healthy lifestyle behaviors such as regular physical activity and a nutritious diet. Interventions have been found to yield promising enhancements in QOL. These programs offer one or more treatment components, including physical activity, nutritional counseling, behavioral modification, and parent involvement. Inpatient and outpatient facilities are the primary outlets for these services. This chapter focuses primarily on the impact of pediatric overweight interventions on QOL, and also describes the modalities through which they enhance QOL. List of Abbreviations: BM, Behavior Modification; BMI, Body Mass Index; CBT, Cognitive Behavioral Therapy; NC, Nutritional Counseling; PA, Physical Activity; PI, Parent Involvement; POI, Pediatric Overweight Intervention; QOL, Quality of Life; TEENS, Teaching, Education, Exercise, Nutrition, and Support
1
Introduction
The prevalence of overweight among children and adolescents is increasing at alarming rates in both the United States (Ogden et al., 2006) and in other industrialized countries (Kiess et al., 2001; Rudolf et al., 2001). In response to this crisis, a number of pediatric overweight interventions (POIs) have been developed. These interventions are administered primarily through inpatient and outpatient treatment facilities and include components addressing physical activity, nutritional counseling, behavioral management, and parental involvement (Barlow and Expert Committee, 2007, Haddock et al., 1994). In addition to the research available on interventions, there is evidence to support the efficacy of these components individually. In addition to promoting weight management and healthy lifestyle habits, POIs have been found to enhance quality of life (QOL). In this chapter, we first highlight some specific QOL improvements related to participation in POI. Next, we discuss types of interventions and intervention program components, focusing on the mechanisms through which they improve QOL. Finally, we provide readers with special considerations for POI.
2
Quality of Life
In the literature, little consensus has been reached regarding an exact definition of QOL. Definitions vary widely. For example, QOL has been described as the degree to which human physical, spiritual, social, economic, and psychological needs are met (Dempster and Donnelly, 2000) and as a multidimensional construct subjectively based on the individual’s perception of his or her well-being or objectively based on societal standards of well-being (Felce and Perry, 1995). According to Schipper et al. (1996), physical QOL refers to strength, energy, and the ability to carry on normal daily activities in addition to normal physical functioning. Psychological QOL refers to mental well-being with regard to psychological
Implementing Interventions to Enhance QOL in Overweight Children and Adolescents
147
processes, such as anxiety and depression. Also according to Schipper and colleagues, social QOL refers to the quality of an individual’s relationships with peers, family, and the general community. For the purposes of this chapter, we use Schipper’s conceptualization of QOL in our discussions of the enhancements in QOL associated with POI. When compared to their normal weight peers, overweight children and adolescents have been found to have lower QOL in the physical (Friedlander et al., 2003; Schwimmer et al., 2003; Pinhas-Hamiel et al., 2006), psychological (Friedlander et al., 2003; Schwimmer et al., 2003), and social domains (Pinhas-Hamiel et al., 2006; Schwimmer et al., 2003). It is beyond the scope of this chapter to discuss the impact of pediatric overweight on QOL. However, it is clear that the detrimental effects of pediatric overweight on QOL provide a strong rationale for POI.
3
Specific QOL Improvements Associated with POI
Because of the negative effects of pediatric overweight on QOL, overweight children and adolescents should seek some type of intervention to improve their QOL (Barlow and Expert Committee, 2007). While unstructured attempts to reduce overweight may be successful for some children and adolescents, Barlow and Expert Committee (2007) recommend that when primary care efforts to treat pediatric overweight fail, overweight children and adolescents should enroll in a multidisciplinary POI. Multidisciplinary POIs that promote healthy eating and physical activity can decrease the degree of overweight in children and adolescents (Wilfley et al., 2007) and have the greatest capacity for long-term efficacy (Jelalian and Saelens, 1999). Many POIs encourage participants to emphasize increasing their frequency of healthy lifestyle habits and focus very little, if at all, on weight loss (Stern et al., 2006; Stern et al., 2007). POIs that promote healthy lifestyle changes, as opposed to weight-focused diets, have been suggested to have the greatest potential for success (Wilfley et al., 2007). Because healthy lifestyle habits are associated with long-term weight management, and QOL improvements have been found to be associated with successful weight management (Dreimane et al., 2007; Fullerton et al., 2007), it seems plausible to infer that positive healthy lifestyle changes are vital to improving QOL. These enhancements have been observed in the physical, psychological, and social domains of QOL (see > Figure 147-1).
3.1
Physical QOL
A number of physical QOL enhancements have been associated with POI. Some of these enhancements include weight management, improvements in anthropometric measurements, improved physiological functioning, and enhanced physical fitness.
3.1.1
Weight Management
The goal of most POIs is to promote weight management via the implementation of healthy lifestyle habits, such as eating low-fat, nutrient dense foods and engaging in regular physical
2519
2520
147
Implementing Interventions to Enhance QOL in Overweight Children and Adolescents
. Figure 147-1 Quality of Life components addressed by pediatric overweight intervention programs
activity (Barlow and Expert Committee, 2007). Weight management is necessary for reducing the risk of future weight-related health problems (Spear et al., 2007). In addition, weight loss is associated with improved well-being (Dreimane et al., 2007) and physical QOL (Fullerton et al., 2007). Changes in body composition subsequent to POI are measured in terms of a variety of anthropometric measurements. Because the degree of overweight in children and adolescents is measured using the BMI percentile rank, most outcome studies report changes in BMI percentile. Many studies have yielded significant decreases in BMI (e.g., Myers et al., 1998; DeStefano et al., 2000; Braet et al., 2004; Gately et al., 2005; Jiang et al., 2005; Kirk et al., 2005; Edwards et al., 2006); however others have not (Cameron, 1999; Yin et al., 2005; Daley et al., 2006). Nonetheless, studies often find decreases in anthropometric measurements such as percentage of body fat (Gately et al., 2005; Kirk et al., 2005; Nemet et al., 2005; Yin et al., 2005; Savoye et al., 2007), fat mass (DeStefano et al., 2000; Gately et al., 2005), waist circumference (Gately et al., 2005), and hip circumference (Gately et al., 2005) as well as increases in lean body mass (Kirk et al., 2005; Yu et al., 2008).
3.1.2
Physiological Functioning
Another primary goal of POIs is to improve physiological functioning. There is evidence to suggest that POI can improve physiological processes that have been negatively impacted by overweight (Barlow and Expert Committee, 2007). Specifically, decreases in blood pressure (Gately et al., 2005; Jiang et al., 2005; Kirk et al., 2005; Reinehr et al., 2006a; Reinehr et al., 2006b), cholesterol (Jiang et al., 2005; Kirk et al., 2005; Reinehr et al., 2006a; Reinehr et al., 2006b; Savoye et al., 2007), insulin levels (Kirk et al., 2005; Reinehr et al., 2006a; Savoye et al., 2007), triglyceride levels (Reinehr et al., 2006b) and increases in bone mineral density (Yin et al., 2005) have been observed in children and adolescents who have participated in POIs. Further, a decrease in BMI has been linked to the reduction of sleep disorders in children, specifically sleep apnea (Chan et al., 2004).
Implementing Interventions to Enhance QOL in Overweight Children and Adolescents
3.1.3
147
Physical Fitness Improvements
Physical fitness improvements are also associated with POI. Specifically, increases in physical activity (Nemet et al., 2005; Piko and Keresztes, 2006), endurance (Nemet et al., 2005), and physical fitness (Nemet et al., 2005; Yu et al., 2008) have been observed. In a 12-week exercise program for overweight boys, aerobic and resistance training led to increases in peak volume of oxygen uptake and resting energy expenditure (DeStefano et al., 2000). QOL enhancements resulting from physical fitness improvements include increased cardiovascular (Carrel et al., 2005; Yin et al., 2005) and aerobic fitness (Gately et al., 2005; Kirk et al., 2005) and psychosocial enhancement (Piko and Keresztes, 2006; Yu et al., 2008). Overall, these positive findings on physical fitness provide evidence of direct improvements in QOL as a result of POI.
3.2
Psychological QOL
There is evidence to suggest that POI can enhance psychological QOL, especially with regard to depression, self-esteem, eating behaviors, and body perceptions.
3.2.1
Depression
Overweight children and adolescents report higher rates of depression than their normal weight peers (Pinhas-Hamiel et al., 2006). Despite the potential benefit of enhanced mood in response to POI, relatively few studies report changes in depression scores. However, some have yielded promising results (Myers et al., 1998; Edwards et al., 2006). Edwards and colleagues suggest that socialization with children with similar problems along with increased self-efficacy for weight control may contribute to decreased depression. The exercise components of intervention programs, in particular, appear to have a positive effect on depression. Stella et al. (2005) suggest that aerobic exercise is associated with reduced depression because it causes biochemical changes in the brain that increase serotonin levels which elevate mood. However, not all studies manipulating exercise in overweight children and adolescents have yielded reductions in depression scores. Daley et al. (2006) found no changes in depression, positive affect, or negative affect in response to an exercise intervention.
3.2.2
Self-Esteem
Overweight children and adolescents tend to have lower self-esteem than their normal weight peers (Erermis et al., 2004). In a review of the impact of pediatric overweight programs on selfesteem, Lowry et al. (2007) reported that the majority of studies examining self-esteem change as a result of intervention reported significant increases in either global self-esteem or a related construct, such as self-worth, self-concept, perceived competence, and physical appearance. The mechanisms through which interventions affect self-esteem are not entirely clear. However, Lowry and colleagues proposed some plausible hypotheses. First, decreases in weight status may be associated with increases in self-esteem. It may be that children who lose weight experience increased self-esteem as a result of the satisfaction associated with their success. Second, Lowry and colleagues suggest that the parental involvement often encouraged by POIs
2521
2522
147
Implementing Interventions to Enhance QOL in Overweight Children and Adolescents
can be associated with increased self-esteem. Parents can provide a supportive environment that promotes a healthy lifestyle, reinforces healthy behaviors, and institutes family lifestyle changes. Third, Lowry and colleagues suggest that group interventions enhance self-esteem because they provide peer social support. Other studies have also documented changes in overall self-esteem (Braet et al., 2004; Savoye et al., 2005; Daley et al., 2006; Jelalian et al., 2006), and in a number of self-esteem domains. Improvements in perceived physical appearance (Braet et al., 2003; Braet et al., 2004; Jelalian et al., 2006) and athletic competence (Braet et al., 2003; Braet et al., 2004) perhaps develop subsequent to improvements in physical activity. Increased perceived social acceptance (Braet et al., 2003; Braet et al., 2004) may be attributed to the peer support provided by intervention programs or from decreased peer stigmatization. Despite these promising findings, some studies have not found significant improvements in self-esteem following participation in POI. For example, one study reported no change in self-esteem (Dremaine et al., 2007) and one found changes due only to weight loss (Huang et al., 2007). Further, one study found that self-esteem actually decreased after intervention, perhaps because those in the intervention felt ‘‘singled out’’ for treatment (Cameron, 1999). Although high self-esteem is generally thought to promote positive adaptation, it can also be associated with negative healthy lifestyle perceptions and behaviors. Stern and colleagues (2006) found that overweight treatment-seeking adolescents with higher self-esteem were less likely to believe that their weight was a problem, less likely to believe that their appearance would improve if they lost weight, and were less likely to exercise regularly than those with lower self-esteem. These maladaptive perceptions and behaviors about weight are potential barriers to successful POI. Despite the potential barriers high self-esteem poses to overweight intervention, many studies have documented increases in self-esteem in response to POIs (Lowry et al., 2007). Not only has self-esteem been shown to increase after participation, but self-esteem appears to affect other components of QOL. For example, Stern and colleagues (2007) found that in an adolescent treatment seeking-sample, self-esteem partially mediated the relationship between teasing and QOL, thereby serving as a protective factor against the adverse effects often associated with teasing.
3.2.3
Eating Behavior
Eating behavior modification is a healthy lifestyle change encouraged by many POIs. The consumption of food high in fat and calories and low in nutrients contributes to overweight. Therefore, most POIs promote healthy modification of eating behavior. However, relatively few studies report changes in eating-related behavior and attitudes. Results are mixed among the studies that have yielded changes in eating behaviors. One study (Edwards et al., 2006) found no differences in food preoccupation, dieting patterns, and eating attitudes among 8- to 13-year-olds who participated in a family-based POI. Studies have found that intervention has no impact on emotional eating (Braet et al., 2003; Braet et al., 2004), but may be effective in decreasing dietary restraint (Braet et al., 2004). Changes in actual eating behavior, however, are difficult to measure as they are based largely on self-report of behaviors that occur outside of the intervention setting. Disordered eating behaviors such as binge eating have been observed in POI seeking populations (Decaluwe et al., 2003). Binge eating poses a specific challenge for POI programs
Implementing Interventions to Enhance QOL in Overweight Children and Adolescents
147
because it is often associated with other underlying psychological issues (Stice et al., 2002) such as depression (Ross and Ivis, 1999) and low self-esteem (Decaluwe et al., 2003), and because it requires specific behavior modification to treat (Glasofer et al., 2007). This evidence illustrates the need for POI programs to specifically address binge eating. One possible intervention technique, cognitive behavioral therapy (CBT), has been found to be effective in treating binge eating in the pediatric overweight population (Braet et al., 2004).
3.2.4
Body Perceptions
Overweight children and adolescents are susceptible to sensitivity about their weight, poor body image, and low body satisfaction (Huang et al., 2007). Therefore, POI programs have begun to investigate changes in body perceptions as a result of intervention. Some specific changes that have been observed are decreased drive for thinness (Braet et al., 2003; Braet et al., 2004), body dissatisfaction, weight concern, and shape concern (Braet et al., 2004). However, other evidence suggests that participation in POI alone does not yield changes in body perceptions. Some studies indicate that only adolescents who participate in a POI and either lose or maintain weight experience positive changes in body perceptions such as body image (Huang et al., 2007), body satisfaction, and physical appearance esteem (Walker et al., 2003). In these studies, weight loss appeared to make the difference, as those who lost or maintained weight experienced the greatest changes in body perceptions. It is important to monitor how changes in body perceptions relate to weight loss. That is, beneficial improvements in body perceptions contribute to improvements in QOL. However, some children and adolescents may interpret improvements in body perceptions as indicators that their weight is no longer a health issue, perhaps contributing to reduced motivation to continue engaging in healthy lifestyle behaviors. Another concern is that too much emphasis on weight and associated health risks in POIs could contribute to poor body perceptions (Huang et al., 2007). To prevent the development of maladaptive body perceptions, it has been recommended that intervention programs maintain a supportive, non-critical environment (Barlow and Dietz, 1998).
3.3
Social QOL
Overweight, treatment-seeking children and adolescents have been described by their mothers as being socially withdrawn and isolated (Zeller et al., 2004b). However, POI programs have been found to have the capacity to enhance social QOL, particularly regarding peer and family relationships. These enhancements include increased social competence (Myers et al., 1998), improved social well-being (Braet et al., 2004), and decreased social problems (Epstein et al., 1998; Myers et al., 1998). Consideration of social functioning in intervention programs is vital, given the isolating and stigmatizing effects of pediatric overweight (Strauss and Pollack, 2003). Improvements in social functioning and social QOL are therefore key potential benefits of POIs.
3.3.1
Peer Relationships
Youth who have participated in POIs report increased peer support (Resnicow et al., 2000). Peer support has been linked to several positive treatment outcomes, including increased
2523
2524
147
Implementing Interventions to Enhance QOL in Overweight Children and Adolescents
participation in physical activity (Andersen and Wold, 1992) and enhanced QOL (Zeller and Modi, 2006). Myers and colleagues propose that these programs work because as children lose weight, they become less socially stigmatized and develop improved peer relationships. The improved peer relationships, in turn, increase the availability of peer-related physical activity while decreasing opportunities for overeating, thereby helping children maintain these healthy lifestyle behaviors.
3.3.2
Teasing
Overweight children and adolescents are often subjected to teasing by peers (Robinson, 2006). Although some studies have shown that POI contributes to improvements in social functioning (Epstein et al., 1998; Myers et al., 1998; Braet et al., 2004) to our knowledge, there has been no documented evidence of reduced frequency of teasing subsequent to POI. Teasing should certainly be addressed by POIs. Its reduction is an important target for future clinical research.
3.3.3
Family Relationships
Improved family functioning has been found to be a result of POI programs, especially those that provide family or parent involvement components. Prior to treatment, families of overweight children and adolescents are characterized by high maternal psychological distress, high family conflict, and negative meal-time interaction (Zeller et al., 2007). Although only a limited amount of research has examined the effects of POIs on family functioning, preliminary findings have been encouraging. For example, some families experienced increased cohesion (Kirschenbaum et al., 1984; Dreimane et al., 2007), increased mutual support (Kirschenbaum et al., 1984), reductions in parental distress (Epstein et al., 2000; Epstein et al., 2001), and reductions in maternal psychopathology (Myers et al., 1998).
4
Components of Programs
Many POIs are comprised of a combination of four components: physical activity, nutritional counseling, behavioral modification, and parent involvement. The purpose of this section is to describe the mechanisms through which POIs seem to enhance QOL. > Figure 147-2 provides a representation of these components.
4.1
Physical Activity
Exercise is a necessary and vital component of effective POI (Barlow and Expert Committee, 2007) and a wide variety of approaches have been implemented to increase physical activity (Snethen et al., 2006). While increasing physical activity is an important component of weight management, it does not appear to be sufficient if implemented alone (Spear et al., 2007). POIs encourage participants to engage in physical activity outside of the intervention facility using such techniques as self-monitoring energy expenditure (e.g., Epstein et al., 2000) and tracking physical activity using devices such as accelerometers or pedometers (Resnicow et al., 2000; Salmon et al., 2005). Some programs encourage the reduction of sedentary
Implementing Interventions to Enhance QOL in Overweight Children and Adolescents
147
. Figure 147-2 Common components of outpatient pediatric overweight intervention programs
behaviors such as television viewing (Robinson, 1999; Epstein et al., 2000; Epstein et al., 2004). Other programs take a more active approach by providing exercise facilities (Sothern et al., 2002; Savoye et al., 2007) or teaching participants athletic skills (Resnicow et al., 2000; Salmon et al., 2005; Yin et al., 2005). Studies have noted several improvements in pediatric QOL outcomes resulting from increased physical activity including increased self-esteem (Calfas and Taylor, 1994; Weiss et al., 1990), decreased depression (Shepard, 1995; Stella et al., 2005), increased self-image (Kirkcaldy et al., 2002), and improved self-efficacy (Sallis et al., 2000; Neumark-Sztainer et al., 2003).
4.2
Nutrition Counseling
Adopting nutritious eating habits is also a central component of a healthy lifestyle. It has therefore been one of the main areas of focus in multidisciplinary POI. Often, participants are given information regarding nutrition and healthy eating via individual or group nutrition counseling sessions (Valverde et al., 1998; Savoye et al., 2005; Savoye et al., 2007). POIs that include a nutrition component are more effective in producing positive outcomes when compared to those without a nutrition component (Collins et al., 2006). Although it appears that having a nutrition component is vital to successful POI, it is not clear which of the many different nutrition components is most effective (Collins et al., 2006). However, a review of the extant studies suggests that a few conclusions can be drawn. First, research suggests that the long-term maintenance of weight loss is better achieved through educational-based activities and diet modifications (Murphree, 1994), rather than through severe diet restriction (Valverde et al., 1998) or structured meal plans (Savoye et al., 2005). Second, the inclusion of a nutrition component in which participants are taught to make better food choices is
2525
2526
147
Implementing Interventions to Enhance QOL in Overweight Children and Adolescents
preferable to those that focus on prohibiting certain foods (Savoye et al., 2007). Further, although it is difficult to identify the specific effects of a nutritional component of POI on QOL, it is clear that in order to maximize long-term improvements in QOL, it is necessary to target and change maladaptive eating styles in addition to promoting physical activity.
4.3
Behavioral Modification
Specific behavioral modification techniques may be necessary to change the unhealthy lifestyle habits that contribute to overweight, as education alone is often inadequate in effective weight management (Spear et al., 2007). To facilitate behavioral change, many POIs provide some type of behavioral modification or behavioral support component. Behavioral modification modalities include self-monitoring of eating behavior and physical activity, stimulus control strategies, and contingency management (Jelalian and Saelens, 1999). In addition to behavioral modification, the efficacy of cognitive behavioral therapy (Duffy and Spence 1993; Braet et al., 2004; Herrera et al., 2004; van den Akker et al., 2007) has also been examined. With cognitive behavioral therapy (CBT) for POI, children and adolescents learn to monitor and understand how their thoughts are associated with weight-related behaviors. In a review of POI behavioral modification components, Spear and colleagues (2007) suggest that when compared to cognitive techniques, behavioral techniques are more effective in improving healthy lifestyle behaviors such as diet and physical activity. Like physical activity and nutritional components, it is difficult to evaluate the specific effects of behavioral modification within multidisciplinary POIs. However, because these comprehensive programs that include behavioral modification have yielded some successes overall, we can argue here that the behavioral component may be an important element, although further study is needed to determine how much of a relative effect this component has on outcomes.
4.4
Parent Involvement
Many POIs encourage or require parental participation and there is evidence to support the effectiveness of parental involvement (e.g., Golan, 2006). Young et al. (2007) suggest that those interventions that promote active parental involvement yield larger effect sizes than those that do not. Parents’ capacity to role model healthy eating and physical activity is the primary rationale for involving them in their children’s overweight intervention (Jelalian and Saelens, 1999; St Jeor et al., 2002). In addition, parental involvement is needed in POI, as children most often depend on parents to provide food and opportunities to engage in physical activity. Unfortunately, children do not always agree with their parents on weight and healthy lifestyle-related issues. For example, parents may perceive their overweight children as having poorer QOL than the overweight children perceive themselves as having (Hughes et al., 2007). Parents and children may also have discrepant views on weight status issues. Stern and colleagues (2006) found that although daughters are more likely than mothers to be currently trying to lose weight, daughters are less likely to understand genetic influences on weight and more likely to believe that if one is active, weight status is not important.
Implementing Interventions to Enhance QOL in Overweight Children and Adolescents
147
One reason parents may refer their children to POI is that they tend to view their offspring’s overweight as more detrimental than does the child (Zeller et al., 2004b; Hughes et al., 2007). Thus, differences in child and parent perceptions of the magnitude of QOL impairments resulting from overweight may be one barrier to child motivation to change. Despite the potential for differences in parent and child perceptions on the impact of overweight, parents and children should be encouraged to work together to achieve a healthier lifestyle. Although many parents of children who are involved in POIs have accurate perceptions of their children’s problems, some parents have distorted views of their children’s overweight. They may see their children as ‘big-boned’ or ‘solid’ rather than overweight (Young-Hyman et al., 2000; Stern et al., 2006). These distorted perceptions are another possible barrier to beneficial parental involvement in POI.
5
Multidisciplinary POI Programs
There is evidence for the effectiveness of the four major component areas of POI (i.e., physical activity, nutritional counseling, behavioral modification, and parent involvement) in both community and clinical pediatric overweight populations. Multidisciplinary POIs that provide all four of these components include primarily inpatient and outpatient clinics, but also smaller interventions conducted in school (Cole et al., 2006) or weight-management camp settings (e.g., Walker et al., 2003; Gately et al., 2005). Successes observed in POIs conducted in schools and camps suggest that healthy behavior modifications can be effectively promoted and implemented in a number of youth-friendly environments. Although findings ascertained from POIs conducted in schools and camps are important, there is a relative paucity of evidence for their effectiveness. Therefore, our focus in this section will be exclusively on formal inpatient and outpatient POIs that include physical activity, nutritional counseling, behavioral modification, and parent involvement.
5.1
Outpatient Intervention
The majority of literature on POI evaluates outpatient programs. Many of these programs are university-based and are comprised of physical activity, nutritional counseling, behavioral modification, and parental involvement components (see > Table 147-1). > Table 147-2 provides a detailed description of the Teaching, Education, Exercise, Nutrition, and Support (TEENS) program, the outpatient multidisciplinary POI program with which we are affiliated. We provide this description as a prototype for outpatient POI. > Figure 147-3 illustrates the TEENS timeline.
5.2
Inpatient Intervention
Inpatient POIs offer the most intensive intervention available for pediatric overweight. For severely overweight children and adolescents, inpatient care may be necessary, as outpatient treatment programs are not very successful in treating the severely overweight population
2527
Behavior modification: Weekly sessions to discuss accomplishments; positive reinforcement of healthy eating & physical activity; self-monitoring, commitment, setting goals, relapse prevention also discussed Exercise: Frequency and intensity based on overweight severity; increases over the course of the program Nutrition: Recommendations based on overweight severity, but includes calorie monitoring, portion control; increased protein consumption, reduced carbohydrate & fat consumption Family involvement: Group nutrition education sessions attended by patient and family; focus on identifying food groups, cooking activities, tips on reading labels and grocery shopping, tips for dining out
93, 13- to 17-year-olds 1-year medical, psychosocial, nutrition, & exercise (BMI ≥ 85th% percentile) intervention; family involvement encouraged
Committed to Kids
Intervention Behavior modification: 1x/wk during first 6 months; included self-awareness, goal setting, stimulus control, coping skills, cognitive behavior strategies, contingency management Exercise: 2x/wk during first 6 months; 50 minutes of warm-up, aerobic exercise, cool-down Nutrition: Promoted low-fat, nutritious foods and portion control Parent involvement: Parents attended nutrition counseling sessions
Method
209 overweight 8- to Treatment: Family based 16-year olds; (105 program included exercise, treatment; 69 control) nutrition, and behavioral modification Control: Weight counseling every 6 months
Sample
Bright Bodies
Program Name
. Table 147-1 Outpatient POI programs Reference
147
Participants reduced BMI Sothern from 32.3 (SD=1.3) at et al., baseline to 28.2 (SD=1.2) 2002 at 1-year.
6- and 12-month follow- Savoye up intervention vs. et al., control group: 2007 Significant decreases in weight, BMI, body fat %, kg of estimated body fat mass, total cholesterol, fasting insulin, insulin resistance
QoL Outcomes
2528 Implementing Interventions to Enhance QOL in Overweight Children and Adolescents
Behavior modification: Behavior therapy 2x/ week during first 3 months; individual family psychotherapy from months 3 to 9 Exercise: 1x/week throughout program Nutrition: 6 group sessions of a nutritional course during the first 3 months; “Optimized Mixed Diet” was taught – dietary recommendations based on German dietary guidelines Parent involvement: Parent groups met 2x/ month during the first three months, 1x/ month from months 3 to 9
Behavior modification: Instruction on healthy lifestyle habits Exercise: Included exercise intervals, exercise videos, walking, dancing, and modified sports (participants were kept continually active) Nutrition: Included instruction on food guide pyramid, reading food labels, and monitoring portion sizes; tips for dining out, reducing sugar and cholesterol consumption, and understanding food nutrients and additives Parent involvement: Parents were informed of the risks of overweight, and the importance of healthy lifestyle habits Intervention group over 2 year period: decreased BMI, systolic blood pressure, insulin level, and cholesterol concentration
8 week program: Improved parent perception of child’s general health, physical functioning, bodily pain, behavior, and mental health. 12 week program: Improved parent perception of child’s physical limitations, bodily pain, behavior, health, and family cohesion. Reinehr et al., 2006a
Dreimane et al., 2007
Listed above are the results of studies conducted on several multidisciplinary outpatient pediatric overweight intervention programs. Their sample sizes, research methodologies, intervention strategies, and QoL outcomes are provided. BMI - Body Mass Index SD - Standard Deviation
240, 6-to 14-year-olds 1 year intervention program; (203 treatment, 37 reduction in intensity every control) 3 months; follow-up 2 years after starting intervention
Obeldicks
8- or 12-week program; participants attended once per week; exercise, nutrition education, behavior modification, family involvement combined into one, 90-minute per week, session
264, 7- to 17-yearolds, (BMI ≥ 85th% percentile) (180 enrolled in an 8 wk program; 84 enrolled in a 12 wk program)
Kids N Fitness
Implementing Interventions to Enhance QOL in Overweight Children and Adolescents
147 2529
2530
147
Implementing Interventions to Enhance QOL in Overweight Children and Adolescents
. Table 147-2 TEENS program description Component
Description
Behavioral treatment
1. At enrollment, behavioral specialists (counseling psychology doctoral students) conduct a detailed intake session to evaluate each participant’s developmental history, past or current methods for controlling weight, dietary intake, level of physical activity, family functioning, and goals for participation in the program. 2. After the participant has received medical clearance to enroll in the program, the participant and behavior specialist meet biweekly for 30 minutes. The purpose of these sessions is to make and maintain goals, monitor program adherence, and to address personal or family issues that may be impacting treatment. 3. After approximately 2 to 3 sessions, the behavior specialist refers the participant to group behavioral support sessions. During the behavioral group support sessions, each participant has the opportunity to discuss with his or her same sex-peers issues related to their treatment.
Physical Activity
1. The TEENS program facility has a gym complete with cardiovascular equipment, resistance machines, and free weights. 2. The goal for each exercise session is to reach a maximum heart rate at or above 150 beats per minute, or 70-80% of the participant’s maximum heart rate. 3. Typical sessions include: a. 10 minutes warm-up b. 20-30 minutes cardiovascular activity c. 20-30 minutes strength training d. 10 minute cool down
Nutritional counseling
1. Registered dieticians provide individual nutrition lessons to the participants and their parents.
Parent Involvement
1. TEENS is a family-based program and participants’ parents are expected to be involved in their child’s treatment. 2. Parents are required to attend enrollment and intake sessions, as well as progress meetings with staff members at various points throughout the program. They are encouraged, and may be required, to attend their child’s individual behavioral support and nutrition sessions. 3. There is a 12-week, psychoeducational group for parents only. Topics such as family meal planning, environmental and genetic precursors of overweight, promoting healthy body image, and dealing with teasing are discussed.
Listed above is a description of TEENS, a multidisciplinary outpatient pediatric overweight intervention program. TEENS - Teaching, Education, Exercise, Nutrition, Support
(Braet et al., 2003). Inpatient programs are especially valuable for overweight children who come from families who are less supportive of their efforts by offering a live-in environment in which they can learn healthy eating behaviors and engage in physical activity (Braet et al., 2004). A sample of inpatient POIs, their descriptions, and outcomes are listed in > Table 147-3.
Implementing Interventions to Enhance QOL in Overweight Children and Adolescents
147
. Figure 147-3 Virginia Commonwealth University TEENS program timeline. TEENS Teaching, Education, Exercise, Nutrition, and Support; PA Physical Activity; BM Behavior Modification; NC Nutritional Counseling; PI Parent Involvement
6
Conclusion
The prevalence of pediatric overweight is rising at alarming rates. Overweight is associated with poor physical, psychological, and social QOL. In response to the high rates of pediatric overweight and its detrimental effects on QOL, POIs have been established in a number of countries worldwide. These interventions offer physical activity, nutritional counseling, behavioral modification, and parent involvement components. In this chapter, we have discussed the specific QOL improvements associated with POI and have described the ways in which these programs enhance QOL. One challenge of evaluating multidisciplinary POIs is that it is difficult to determine which components are most effective (Braet et al., 2003) in producing QOL enhancements. However, the successes that multidisciplinary POIs yield suggest that this approach may be most effective. There are a number of demographic, economic, psychological, and cultural factors that need to be more successfully integrated into POIs. For example, economic factors (Cote et al., 2004), minority status, depression, and poor self-concept (Zeller et al., 2004a) have been found to predict attrition. There is a need for intervention programs that are sensitive to African Americans, as they are at increased risk of overweight (Ogden et al., 2006). Treatment
2531
Sample
76 obese 10- to 17- year olds (38 treatment; 38 control)
122 inpatient overweight 7- to 17-year olds
Program Name
Zeepreventorium
Zeepreventorium
. Table 147-3 Inpatient POI programs
Nutrition: 100-1600 cal/day; 3 meals, 2 snacks per day Physical activity: 14+ hr/week Psychological intervention: cognitive behavioral modification; food cue resistance training; random assignment to extended coping program or to standard treatment program
Nutrition: 1500-1800 kcal/day; 3 meals, 2 snacks per day Physical activity: 10 + hr/week Psychological intervention: cognitive behavioral modification Parent involvement: Received information on healthy food preparation and physical activity
Intervention Braet et al., 2003
Reference
Pre- vs. post intervention; Braet Decreased percentage et al., overweight, body mass index, & 2004 weight; parent-reported overall behavioral & internalizing problems; external eating, eating concern, frequency of binge eating, and bulimic symptomatology; drive for thinness, body dissatisfaction, weight concern, and shape concern. Increased global sense of selfworth, and perceived athletic & physical competence; dietary restraint *All changes were maintained at a 14-month follow-up; also increased perceived academic competence and social acceptance at follow-up
Intervention group: 48% decrease in median adjusted BMI; decreased drive for thinness & external eating; increased self-perceived physical appearance, athletic competence, & social acceptance
QoL Outcomes
147
10-month inpatient treatment program
10-month inpatient treatment program; Compared treated children to wait-list controls
Method
2532 Implementing Interventions to Enhance QOL in Overweight Children and Adolescents
BMI ≥ 95th percentile with presence of severe, overweightrelated health issue
3 phases: 1.) Pre-admission phase 2.) Weight loss and maintenance establishment 3.) Cognitivebehavioral treatment for addressing obstacles Nutrition: Moderate caloric restriction Physical activity: Individualized plans based on physiological functioning Psychological intervention: cognitive behavioral psychotherapy Parent involvement: 1x/week individual training sessions with dietician and behavioral specialist; taught ways to enhance healthy habits for the whole family None reported
Fennig & Fennig, 2006
Listed above are the results of studies conducted on several multidisciplinary outpatient pediatric overweight intervention programs. Their sample sizes, research methodologies, intervention strategies, and QoL outcomes are provided. BMI - Body Mass Index SD - Standard Deviation
Overweight Intervention Program at the Medical Psychiatric Unit of Schneider Children’s Medical Center of Israel
Implementing Interventions to Enhance QOL in Overweight Children and Adolescents
147 2533
2534
147
Implementing Interventions to Enhance QOL in Overweight Children and Adolescents
programs specifically for African American girls have been somewhat successful, perhaps because they are more culturally relevant (Resnicow et al. 2000, 2005). The available literature suggests those who are financially disadvantaged, of diverse racial backgrounds, or who have psychological or emotional issues are at particularly high risk for attrition from POIs. A greater understanding of strategies that can be used to decrease attrition among these populations is clearly necessary. Despite some promising results, many POIs have failed to yield significant QOL enhancements. Some proposed reasons for these failures include the difficulty involved in adhering to intense intervention protocols (Epstein et al., 1996; Yanovski and Yanovski, 2003), the costs associated with intervention (Jelalian and Saelens, 1999), and a lack of research evidence for effective intervention (Wilson et al., 2003). In addition, what differentiates successful pediatric overweight participants from unsuccessful pediatric overweight participants is largely unknown (Reinehr et al., 2003). Clearly, more research is needed to help increase our understanding of why some POIs show success overall whereas others do not. Such information should help us make real changes to our multidisciplinary POIs and thereby improve our intervention strategies.
Summary Points The rates of pediatric overweight are high in the United States and in many other industrialized countries.
Intervention programs have been developed to address the pediatric overweight epidemic. Intervention programs have the capacity to enhance QOL in all domains. To promote long-term weight management, intervention programs encourage healthy lifestyle behaviors, not weight loss.
POIs typically include physical activity, nutritional counseling, behavioral modification, and parent involvement components.
Multidisciplinary programs, primarily housed in outpatient or inpatient clinics, have the greatest capacity for success because they address multiple issues related to pediatric overweight.
References Andersen N, Wold B. (1992). Res Q Exerc Sport. 63(4): 341–348. Barlow SE, Dietz WH. (1998). Pediatrics. 102: e29. Barlow SE. (2007). Expert Committee Pediatrics. 120: S164–S192. Braet C, Tanghe A, De Bode P, Franckx H, Van Winckel M. (2003). Eur J Pediatr. 162: 391–396. Braet C, Tanghe A, Decaluwe V, Moens E, Rosseel Y. (2004). J Pediatr Psychol. 29(7): 519–529. Calfas KJ, Taylor WC. (1994). Pediatr Exerc Sci. 6: 406–423. Cameron JW. (1999). Issues Compr Pediatr Nurs. 22: 75–85.
Chan J, Edman JC, Koltai PJ. (2004). Am Fam Physician. 69(5): 1147–1154. Cole K, D’Auria J, Garner H. (2006). J Spec Pediatr Nurs. 11(3): 166–177. Collins CE, Warren J, Neve M, McCoy P, Strokes BJ. (2006). Archives of Pediatr Adolesc Med. 160: 906–922. Cote MP, Byczkowski T, Kotagal U, Kirk S, Zeller M, Daniels S. (2004). Int J Qual Health Care. 16(2): 165–173. Daley AJ, Copeland RJ, Wright NP, Roalfe A, Wales JKH. (2006). Pediatrics. 118(5): 2126–2134.
Implementing Interventions to Enhance QOL in Overweight Children and Adolescents Decaluwe V, Braet C, Fairburn CG. (2003). Int J Eat Disord. 33: 78–84. Dempster M, Donnelly M. (2000). Soc Work Health Care. 32(1): 45–56. DeStefano RA, Caprio S, Fahey JT, Tamborlane WV, Goldberg B. (2000). Pediatr Diabetes. 1: 61–65. Dreimane D, Safani D, MacKenzie M, Halvorson M, Braun S, Conrad B, Kaufman F. (2007). Diabetes Res Clin Pract. 75: 159–168. Duffy G, Spence SH. (1993). J Child Psychol Psychiat. 34 (6): 1043–1050. Edwards C, Nicholls D, Croker H, Van Zyl S, Viner R, Wardle J. (2006). Eur J Clin Nutr. 60: 587–592. Epstein LH, Coleman KJ, Myers MD. (1996). Med Sci Sports Exerc. 28(4): 428–435. Epstein LH, Goldfield GS. (1999). Med Sci Sports Exerc. 31, (11), Suppl 1, S553. Epstein LH, Myers MD, Raynor HA, Saelens BE. (1998). Pediatrics. 101:554–570. Epstein LH, Paluch RA, Gordy CC, Dorn J. (2000). Archives of Pediatr Adolesc Med. 154: 220–226. Epstein LH, Paluch RA, Kilanowski CK, Raynor HA. (2004). Health Psychol. 23(4): 371–380. Epstein LH, Paluch RA, Saelens BE, Ernst M, Wilfley DE. (2001). J Pediatr. 139: 58–65. Erermis S, Cetin N, Tamar M, Bukusoglu N, Akdeniz F, Goksen D. (2004). Pediatr Int. 46: 296–301. Felce D, Perry J. (1995). Res Dev Disabil. 16(1): 51–74. Friedlander SL, Larkin EK, Rosen CL, Palermo TM, Redline S. (2003). Arch Pediatr Adolesc Med. 157: 1206–1211. Fullerton G, Tyler C, Johnston CA, Vincent JP, Harris GE, Foreyt JP. (2007). Obesity. 15(11): 2553–2556. Gately PJ, Cooke CB, Barth JH, Bewick BM, Radley D, Hill AJ. (2005). Pediatrics. 116(1): 73–77. Glasofer DR, Tanofsky-Kraff M, Eddy KT, Yanovski SZ, Theim KR, Mirch MC, Ghorbani S, Ranzenhofer LM, Haaga D, Yanovski JA. (2007). J Pediatr Psychol. 32(1): 95–105. Golan M. (2006). Int J Pediatr Obes. 1(2): 66–76. Haddock CK, Shadish WR, Klesges RC, Stein RJ. (1994). Ann Behav Med. 16(3): 235–244. Herrera EA, Johnston CA, Steele RG. (2004). Children’s Health Care. 33(2): 151–167. Huang JS, Norman GJ, Zabinski MF, Calfas K, Patrick K. (2007). J Adolesc Health. 40: 245–251. Hughes AR, Farewell K, Harris D, Reilly JJ. (2007). Int J Obes. 31: 39–44. Jelalian E, Mehlenbeck R, Lloyd-Richardson EE, Birmaher V, Wing RR. (2006). Int J Obes. 30: 31–39. Jelalian E, Saelens BE. (1999). J Pediatr Psychol. 24 (3): 223–248. Jiang JX, Lia XL, Greiner T, Lian GL, Rosenqvist U. (2005). Arch Dis Child. 90: 1235–1238. Kiess W, Galler A, Reich A, Muller G, Kapellen T, Deutscher J, Raile K, Kratzsch J. (2001). Obesity Rev. 2: 29–36.
147
Kirk S, Zeller M, Claytor R, Santangelo M, Khoury PR, Daniels SR. (2005). Obes Res. 13(5): 876–882. Kirkcaldy BD, Shephard RJ, Siefen RG. (2002). Social Psychiatry and Psychiatr Epidemiol. 37: 544–550. Kirschenbaum DS, Harris ES, Tomarken AJ. (1984). Behav Ther. 15: 485–500. Lowry KW, Sallinen BJ, Janicke DM. (2007). J Pediatr Psychol. 32(10): 1179–1195. Murphree D. (1994). J Fam Pract. 38: 45–48. Myers MD, Raynor HA, Epstein LH. (1998). Arch Pediatr Adolesc Med. 152: 855–861. Nemet D, Barkan S, Epstein Y, Friedland O, Kowen G, Eliakim A. (2005). Pediatrics 115: e443-–e449. Neumark-Sztainer D, Story M, Hannan PJ, Tharp T, Rex J. (2003). Arch Pediatr Adolesc Med 157(8): 803–810. Ogden CL, Carroll MD, Curtin LR, McDowell MA, Tabak CJ, Flegal KM. (2006). JAMA. 295(13): 1549–1555. Piko BF, Keresztes N. (2006). J Community Health. 31(2): 136–145. Pinhas-Hamiel O, Singer S, Pilpel N, Fradkin A, Modan D, Reichman B. (2006). Int J Obes 30: 267–272. Reinehr T, Brylak K, Alexy U, Kersting M, Andler W. (2003). Int J Obes. 27: 1087–1092. Reinehr T, de Sousa G, Toschke AM, Andler W. (2006a). Am J Clin Nutr. 84: 490–496. Reinehr T, de Sousa G, Wabitsch M. (2006b). J Pediatr Gastroenterol Nutr. 43(4): 506–511. Resnicow K, Taylor R, Baskin M, McCarty F. (2005). Obes Res 13(10): 1739–1748. Resnicow K, Yaroch AL, Davis A, Wang DT, Carter S, Slaughter L, Coleman D, Baranowski T. (2000). Health Educ Behav. 27(5): 616–631. Robinson S. (2006). J Sch Nurs. 22(4): 201–206. Robinson TN. (1999). JAMA 282(16): 1561–1567. Ross HE, Ivis F. (1999). Int J Eat Disord. 26(3): 245–260. Rudolf MCJ, Sahota P, Barth JH, Walker J. (2001). BMJ. 322:1094–1095. Sallis JF, Prochaska JJ, Taylor WC. (2000). Med Sci Sports Exerc. 32: 963–975. Salmon J, Ball K, Crawford D, Booth M, Telford A, Hume C, Jolley D, Worsley A. (2005). Health Promot Int. 20: 7–17. Savoye M, Berry D, Dziura J, Shaw M, Serrecchia JB, Barbetta G, Rose P, Lavietes S, Caprio S. (2005). J Am Diet Assoc. 105: 364–370. Savoye M, Shaw M, Dziura J, Tamborlane WV, Rose P, Guandalini C, Goldberg-Gell R, Burgert TS, Cali AMG, Weiss R, Caprio S. (2007). JAMA. 297: 2697–2704. Schipper H, Clinch JJ, Olweny CLM. (1996). In: Spilker B (ed) Quality of Life and Pharmacoeconomics in Clinical Trials 2nd ed. Lippencott-Raven, Philadelphia, pp. 11–23. Schwimmer JB, Burwinkle TM, Varni JW. (2003). JAMA. 289(14): 1813–1819.
2535
2536
147
Implementing Interventions to Enhance QOL in Overweight Children and Adolescents
Shepard RJ. (1995). Res Q Exerc Sport. 66(4): 298–302. Snethen JA, Broome ME, Cashin SE. (2006). J Pediatr Nurs. 21: 45–56. Sothern MS, Schumacher H, von Almen TK, Carlisle LK, Udall JN. (2002). J Am Diet Assoc. 102(3): S81–S85. Spear BA, Barlow SE, Ervin C, Ludwig DS, Saelens BE, Schetzina KE, Taveras EM. (2007). Pediatrics. 120 (Supp 4): S254-–S288. Stella SG, Vilar AP, Lacroix C, Fisberg M, Santos RF, Mello MT, Tufik S. (2005). Braz J Med Biol Res 38(11): 1683–1689 Stern M, Mazzeo SE, Gerke CK, Porter JS, Bean MK, Laver JH. (2007). J Pediatr Psychol. 32(1): 90–94. Stern M, Mazzeo SE, Porter J, Gerke C, Bryan D, Laver J. (2006). J Clin Psychol Med Settings. 13(3): 217–228. Stice E, Presnell K, Spangler D. (2002). Health Psychol. 21(2): 131–138. St Jeor ST, Perumean-Chaney S, Sigman-Grant M, Williams C, Foreyt J. (2002). J Am Diet Assoc. 102(5): 640–644. Strauss RS, Pollack HA. (2003). Arch Pediatr Adolesc Med. 157:746–752. Valverde MA, Patin RV, Oliveira FLC, Lopez FA, Vitolo MR. (1998). Int J Obes. 22: 513–519. van den Akker ELT, Puiman PJ, Groen M, Timman R, Jongejan MTM, Trijsburg W. (2007). J Pediatr. 151: 280–283. Walker LLM, Gately PJ, Bewick BM, Hill AJ. (2003). Int J Obes. 27: 748–754.
Weiss MR, McAuley E, Ebbeck V, Wiese DM. (1990). J Sport Exerc Psychol. 12: 21–36. Wilfley DE, Tibbs TL, Van Buren DJ, Reach KP, Walker MS, Epstein LH. (2007). Health Psychol. 26 (5): 521–532. Wilson P, O’Meara S, Summerball C, Kelly S. (2003). Qual Saf Health Care. 12: 65–74. Yanovski JA, Yanovski SZ. (2003). JAMA. 289(14): 1851–1853. Yin Z, Gutin B, Johnson MH, MooreJ. Hanes JB, Cavnar M, Thornburg J, Moore D, Barbeau P. (2005). Obes Res. 13(12): 2153–2161. Young KM, Northern JJ, Lister KM, Drummond JA, O’Brien WH. (2007). Clin Psychol Rev. 27: 240–249. Young-Hyman D, Herman LJ, Scott DL, Schlundt DG. (2000). Obes Res. 8: 241–248. Yu CC, Sung RY, Hau KT, Lam PK, Nelson EA, So RC. (2008). J Sports Med Phys Fitness. 48: 76–82. Zeller M, Kirk S, Claytor R, Khoury P, Grieme J, Santangelo M, Daniels S. (2004a). J Pediatr. 114: 466–470. Zeller MH, Modi AC. (2006). Obesity. 14(1): 122–130. Zeller MH, Reiter-Purtill J, Modi AC, Gutzwiller J, Vannatta K, Davies WH. (2007). Obesity. 15(1):126–36. Zeller MH, Saelens BE, Roehrig H, Kirk S, Daniels SR. (2004b). Obes Res. 12(10): 1576–1586.
148 Adolescent Quality of Life in Australia A. H. Lee . L. B. Meuleners . M. L. Fraser 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2538
2
Health and Well-Being of Australian Adolescents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2539
3
Importance of Measuring QOL in the Adolescent Population . . . . . . . . . . . . . . . . . 2539
4
Challenges for Measuring Adolescent QOL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2540
5
Dimensions of Adolescent QOL: Existing Evidence . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2541
6
Generic Measures for Assessing Adolescent QOL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2541
7 7.1 7.2 7.3 7.4 7.4.1 7.4.2 7.4.3 7.5 7.6
Western Australian Study of Adolescent QOL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2545 Study Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2545 Research Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2545 The Measuring Instrument . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2545 Statistical Analyses and Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2546 QOL and Its Determinants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2546 Assessing Measurement Properties Using Structural Equation Modeling . . . . . . . . 2546 Variations in QOL over a Six Month Period . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2547 Contribution to Adolescent QOL Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2549 Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2551
8
Conclusions and Future Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2552 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2552
#
Springer Science+Business Media LLC 2010 (USA)
2538
148
Adolescent Quality of Life in Australia
Abstract: Although the majority of adolescent > quality of life (QOL) research has focused on chronic diseases, the QOL measure can be highly useful in providing a broader understanding of the health and well-being of the general adolescent population. Adolescent QOL is a relatively new field so researchers face several challenges. Adolescents are a unique group and the unique dimensions that make up their QOL are only just emerging. Several generic assessments of adolescent QOL are being produced but predominantly in Europe and North America. One particular Australian based longitudinal study has contributed to the understanding of adolescent QOL. Encouragingly, it reported that Australian adolescents, both with and without a chronic disease, describe their QOL positively. It also provided initial validation for adolescent QOL in Australia to comprise of five dimensions, namely, ‘‘physical health,’’ ‘‘environment,’’ ‘‘social,’’ ‘‘psychological’’ and ‘‘opportunities for growth and development,’’ and showed that these dimensions are interdependent. The longitudinal nature of the study also revealed the dynamic nature of QOL and that potentially modifiable variables of adolescent ‘‘control’’ and ‘‘opportunities’’ could have a significant positive impact on QOL. Research would be enhanced by the development of pertinent adolescent QOL measures based on the most recent modification and validation of internationally developed instruments. Aboriginal and Torres Strait Islanders and rural or remote adolescents are at particular risk of poor QOL and these groups should be targeted for improvement. It is recommended that QOL research be used for developing policy, health intervention programs, monitoring the QOL status of the general adolescent population and identifying those at risk of low QOL. List of Abbreviations: AIHW, Australian Institute of Health and Welfare; CFA, > confirmatory factor analysis; CHQ, child health questionnaire; ComQOL-S, comprehensive quality of life scale – school version; EFA, > exploratory factor analysis; HRQOL, > health-related quality of life; QOL, quality of life; QOLPAV, quality of life profile- adolescent version; SEM, structural equation modeling; WA, Western Australia; WHO, World Health Organization
1
Introduction
Adolescent quality of life (QOL) is a relatively new field in Australia. While the majority of research has focused on adolescents with a > chronic illness, QOL can also be an extremely useful measure for the general, healthy adolescent population. It is a complex concept and there is a lack of agreement regarding precisely what is meant by this term (Wallander et al., 2001). One definition is ‘‘the degree to which the person enjoys the important possibilities of his/her life’’ (Raphael et al., 1996). This holistic approach draws attention to the determinants of health at a range of levels and dimensions. Health-related quality of life (HRQOL) is a measure developed specifically to examine the impact of illness, injury or medical treatment on an individual’s QOL. Previously it included dimensions related to illness or treatment only but is now evolving to incorporate broader factors, blurring the boundary with QOL (Ravens-Sieberer et al., 2006). Adolescents are a unique group with a unique set of dimensions and factors that make up their QOL. A particular challenge is to determine the dimensions of adolescent QOL (Wallander et al., 2001). In addition, few QOL instruments have been developed or validated specifically for use with Australian adolescents. A longitudinal study of adolescent QOL in Western Australia (WA) included healthy adolescents in their sample and has provided valuable information on their QOL, potential dimensions to measure QOL and the interrelationships between these dimensions (Meuleners et al., 2003).
Adolescent Quality of Life in Australia
148
This chapter discusses the current health and well-being of Australian adolescents, the challenges for measuring QOL and evidence for the inclusion of dimensions to measure adolescent QOL. In addition, existing measures of adolescent QOL are reviewed and the WA adolescent QOL study is discussed in detail. Finally, future directions in the field will be explored.
2
Health and Well-Being of Australian Adolescents
It is not easy to gauge or measure how adolescents in Australian society are faring (Eckersley et al., 2006). A recent report compiled by the Australian Institute of Health and Welfare (AIHW) on young Australians aged 12–24 years compared different data sources and highlighted the contradictory information on their health and well being (AIHW, 2007). Several traditional statistical measures of health and well-being have indicated a positive picture of young people in Australia. For example, life expectancy and mortality continues to improve, largely due to decreases in death due to transport injury, suicide and drug dependence disorder (AIHW, 2007). Australian youths are generally highly educated and in 2004–2005, 94% rated their own health as either ‘‘good,’’ ‘‘very good’’ or ‘‘excellent’’ (AIHW, 2007). However, other sources showed that adolescents in Australia are not faring well physically or psychologically. For example, approximately 25% were overweight or obese and less than half met recommended physical activity guidelines in 2004–2005 (AIHW, 2007). The 2004 National Drug Strategy Household Survey reported almost one third of young people drank alcohol at levels that placed them at risk of alcohol-related harm in the short term, 23% had used an illicit drug in the 12 months prior to the survey and around 17% were smokers (AIHW, 2005a). In addition, mental disorders were the leading contributor to the burden of disease and injury in young people with depression, anxiety and substance use accounting for the majority (AIHW, 2007). The rate of completed suicide for young Australians was also among the highest in the world (AIHW, 2000). These statistics thus portray a considerably blurred picture of the overall QOL of Australian adolescents. Research that can provide information over a broad range of aspects of adolescent life is lacking in Australia, meaning conclusions about QOL have to be based on the accumulation of studies examining a single issue or a few key areas (Smart and Sanson, 2005). A recent overview noted that young people commonly self-report optimism and well-being in qualitative studies but their lives appear fairly negative with respect to objective criteria (Eckersley et al., 2006). Several explanations for this contradiction have been proposed. Young people are resilient and adaptable. Moreover, what may be considered a health risk, such as drug use, may be regarded as a life enjoying experience by young people (Eckersley et al., 2006).
3
Importance of Measuring QOL in the Adolescent Population
> Adolescence
is a period of rapid emotional, physical and intellectual transition. Stressors can put this group at risk of various health and behavioral problems (Hurrelmann and Richter, 2006). There is ample evidence that adolescents’ decisions and behaviors can impact on their health and QOL. Physical health, mental health, education and employment outcomes can be
2539
2540
148
Adolescent Quality of Life in Australia
affected immediately and into the future (AIHW, 2005b; Eckersley et al., 2006). Therefore, adolescence is a key period for preventive health intervention. Past research on adolescent QOL has focused on individuals with chronic disease. Disease and treatment have shifted to areas of chronic disease over the years with Attention Deficit Hyperactivity Disorder, conduct disorders and mental disorders becoming highly prevalent in Australian adolescents (AIHW, 2007). This shift has made the development of QOL measures extremely important for planning prevention and care programs (Eiser and Morse, 2001). In terms of healthy adolescents, QOL research allows the monitoring of population health status over time, detection of sub-groups who may be at risk of poor QOL, as well as assessment of the impact of public health interventions (Ravens-Sieberer et al., 2001). While mortality, morbidity and behavioral risk measures are important in tracking health trends, they do not capture the perspective of adolescents themselves (Patrick et al., 2002). A recent review has suggested the need for a more holistic approach to health and well-being research for young Australians (Eckersley et al., 2006). QOL measures gather data over a broad range of aspects of adolescent life that are essential for the development of effective policies and interventions.
4
Challenges for Measuring Adolescent QOL
It is clear that measuring and monitoring the QOL of healthy Australian adolescents have many benefits. However, there are several challenges. Firstly, QOL research in general is a relatively new field. While it is widely agreed that QOL is multidimensional (Rajmil et al., 2004; Raphael et al., 1996; Ravens-Sieberer et al., 2006), precisely which dimensions make up this construct is still under investigation. This issue is important as measuring too few dimensions can lead to ignoring meaningful information whereas measuring too many will result in non-interpretable and unreliable measurement dimensions (Coste et al., 2005). Furthermore, the concept of HRQOL which was once considered a sub-domain of QOL (Schipper et al., 1996), is now evolving to become broader, blurring the boundary between the two concepts (Zullig et al., 2005). Secondly, the majority of research and instrument development in the field has focused on adult QOL. Adolescents operate within different frames of reference and their life experience and daily activities differ markedly from the adult population (Bullinger et al., 2006; Ravens-Sieberer et al., 2006). Consequently, there are fundamental differences in the way they understand and assess their own QOL. Additionally, research that has targeted adolescents has focused mostly on the chronically ill (Ravens-Sieberer et al., 2006). While disease-specific research is useful, the measuring instruments cannot be applied generically and do not enable comparisons across conditions and healthy populations. These instruments also place emphasis on presence of symptoms and functional abilities (Eiser and Morse, 2001) that are not of high importance to the QOL of a healthy population. Finally, information on the QOL of healthy adolescents and how it should be measured is only just emerging in Australia. While international research is informative, the adolescent experience, their priorities and perception of QOL could be influenced by their culture and values (Bullinger et al., 2006). For example, one study identified academic achievement as the most pressing concern for adolescents in Hong Kong (Hui, 2000), which may not be the case in Australia.
Adolescent Quality of Life in Australia
5
148
Dimensions of Adolescent QOL: Existing Evidence
The concept of adolescent QOL has emerged from frameworks initially developed with the adult population (Schipper, 1990). However, one qualitative study has provided evidence that the major components usually measured for adults are meaningful for adolescents but that the importance, spread and distribution of these components change throughout age groups (Rajmil et al., 2004). While dimensions of adult QOL are still controversial, measurement instruments consist at a minimum of physical, psychological and social dimensions (Bullinger, 2002; Ware, 2003). Previous research also indicates that personal characteristics such as age, gender and socioeconomic status could be associated with QOL and should be assessed (Ravens-Sieberer et al., 2007; Vingilis et al., 2002). Adolescents face experiences that are unique to their specific age group. Past research has highlighted the dimensions of family, school and friendships as being particularly important for adolescent QOL (AIHW, 2007; Matza et al., 2004; Raphael et al., 1996). These groups provide their main form of social support and their health and wellbeing has been shown to be associated with a sense of connectedness to family, school and the community (AIHW, 2003). Most Australian adolescents live with their parents, siblings and other family members who provide them with physical, emotional and economic support (AIHW, 2007) and have a direct influence on their QOL. In addition, dimensions commonly assessed for adult QOL including the performance of basic functional tasks and economic or vocational status (Fayers and Machin, 2000), are much less relevant for adolescents. A recent review of the major components of HRQOL instruments for children and adolescents cited physical well-being, psychological well-being, energy and vitality, self perception, cognitive functioning, social functioning and support including friends, sexual life and family, autonomy and independence, psychosocial relations to the material environment and general health perception/life quality as the major components (Ravens-Sieberer et al., 2006). Other possible but less commonly assessed dimensions are also emerging from Australian and international adolescent studies. These include environmental factors such as neighborhood socioeconomic disadvantage, > social capital and neighborhood safety (Drukker et al., 2006; Meyers and Miller, 2004). A role for less tangible, more difficult to measure factors such as beliefs, values and spirituality has also been suggested (Eckersley et al., 2006; WHOQOL SRPB Group, 2006). An Australian study of adolescents in rural Queensland identified loneliness and neighborhood and school belongingness as being significantly associated with adolescents’ subjective QOL (Chipuer et al., 2003). Another study of Western Australian adolescents indicated that improved control and opportunities in the adolescent’s life had a positive impact on adolescent QOL (Meuleners and Lee, 2003). Although an extensive list of possible dimensions is emerging, it is unsure whether they comprehensively encompass adolescent QOL and whether some dimensions should be weighted more heavily than others.
6
Generic Measures for Assessing Adolescent QOL
An increasing number of generic assessments of QOL suitable for the general adolescent population are being produced, predominantly in Europe and North America. The majority of
2541
2542
148
Adolescent Quality of Life in Australia
these questionnaires measure HRQOL but considering the overlap with QOL, all of these measures are included in the discussion. Reviews of QOL questionnaires for adolescents have revealed a reasonably coherent understanding of QOL, but the distribution of items among dimensions varies considerably (Rajmil et al., 2004; Ravens-Sieberer et al., 2006). The features of ten such questionnaires are listed in > Table 148-1. All of them cover physical, psychological, social and school dimensions (Cummins, 1997; Landgraf et al., 1998; Patrick et al., 2002; Raphael et al., 1996; Ravens-Sieberer and Bullinger, 1998; Ravens-Sieberer et al., 2005; Sapin et al., 2005; Starfield et al., 1993; Varni et al., 1999; Vogels et al., 1998). Some also address personal care (Landgraf et al., 1998; Raphael et al., 1996; Ravens-Sieberer et al., 2005; Varni et al., 1999; Vogels et al., 1998) while others extend to environmental dimensions (Raphael et al., 1996; Ravens-Sieberer et al., 2005; Sapin et al., 2005; Varni et al., 1999; Vogels et al., 1998 ). All instruments were self-report and the majority included parent or teacher-report questionnaires to cater for younger children. It is appropriate to collect information by self report as long as the age, maturity and cognitive development of the respondents are accounted for during instrument development (Ravens-Sieberer et al., 2006). A limitation of existing instruments is the lack of international comparability. Moreover, since adolescents vary in their age and developmental stage, a single questionnaire may not be suitable for both younger and older adolescents (Ravens-Sieberer et al., 2006). Some instruments were also validated on small or non-representative samples (e.g., Starfield et al., 1993) or target small age ranges (e.g., Patrick et al., 2002; Sapin et al., 2005; Vogels et al., 1998). Finally, not all the instruments generated items with input from the target group (e.g., Starfield et al., 1993) while others were originally created for the chronically ill (e.g., Varni et al., 1999; Vogels et al., 1998). The KIDSCREEN Quality of Life Questionnaire (Ravens-Sieberer et al., 2005), developed for 8- to 18-year-olds, is the first generic instrument to comprehensively fulfil the standards promoted by the World Health Organization (WHO) for measuring child and adolescent HRQOL (WHO, 1994). It was developed in seven European countries with consultation from the target group and was tested through random sampling of over 15,000 children and adolescents (Ravens-Sieberer et al., 2005). The questionnaire was valid and reliable and could be applicable in Australia after validation. The Comprehensive Quality of Life Scale – School Version (ComQOL–S) was the only instrument that was developed for an Australian population although it was originally designed and evaluated for adults (Cummins, 1997). Few modifications were made for the adolescent population. While it addresses psychological and social dimensions in depth, very few school or neighborhood factors are included. Both the child/adolescent and parent versions of the Child Health Questionnaire (CHQ), developed in the US (Landgraf et al., 1998), have been validated and modified in Australia. It performed well at an item and scale level. Although the physical and psychosocial dimension scores were not supported for population level analyses, the CHQ may be of value for adolescents with health problems (Waters et al., 2000). The Quality of Life Profile- Adolescent Version (QOLPAV), developed in Canada (Raphael et al., 2006), was validated for a study of Western Australian adolescents (Meuleners et al., 2001). This questionnaire was developed with direct input from adolescents and covered a broad range of dimensions including control and opportunities for improvement and change.
The Netherlands
The Netherlands Organization for Applied Scientific Research – Academic Medical Centre Child QOL Questionnaire (TACQOL)
Physical complaints Mobility Independence Cognitive function Social function Positive emotions Negative emotions
8–15 5–15
4–7d 8–12 13–16 4–16
Germany
Revidierter KINDer Lebensqualitatsfragebogan (KINDL-R)
Physiological well-being Psychological wellbeing Self-esteem Family Friends School
General health Physical functioning Limitations 10–18 in schoolwork and activities with friends 5–18 Behavior Mental health Emotional or time impact on the parent Family cohesion Change in health Bodily pain or discomfort Self esteem Limitations in family activities
14–20
US
Being Belonging Becoming
Method of response
35
54
Cummins (1997)
Raphael et al. (1996)
Self report 56 Parent report
Self report 12 24 24 Self report 24 Self report Parent report
Vogels et al. (1998)
RavensSieberer and Bullinger (1998)
Self report 87 28, 50, Landgraf Parent report 98 et al. (three (1998) versions)
Self report
Self-report
Starfield et al. (1993)
Number of itemsb Reference
188 45 12–17c Self report 6–11 Self report 45–188 6–11 Parent report
Child Health Questionnaire (CHQ)
Canada
Quality of Life Profile – Adolescent Version (QOLPAV)
Satisfaction Complaints Resilience Health conditions Attainments of social goals Risk behaviors
Material well-being Health Productivity Intimacy 11–18 Safety Community Emotional well-being
US
Child Health and Illness Profile (CHIP)
Dimensions includeda
Age (years)
Comprehensive Quality of Life Scale – Australia School Version (ComQOL – S)
Country of origin
Name
. Table 148-1 Generic QOL instruments for adolescents
Adolescent Quality of Life in Australia
148 2543
Austria Switzerland Germany Spain France The Netherlands UK
France
KIDSCREEN Quality of Life Questionnaire
Vecu de Sante Percue Adolescent (VSP-A)
This table summarizes the features of ten generic measures for assessing adolescent QOL a Dimensions refer to the broader areas addressed in each instrument. Each dimension consists of several items b Items refer to the total number questions making up the dimensions c The CHIP has two different forms for two different age groups d The KINDL-R has three different forms for three different age groups e The KIDSCREEN has a full version consisting of 52 items and an abbreviated version consisting of 27 items
Physical well being Body image Vitality Psychological well being Relationship with friends Relationship with parents Relationship with teachers Relationship with medical staff School performance Leisure activities
11–17
Physical well-being Psychological well being 8–18 Moods and emotions Self-perception Autonomy Parents relations and home life Peers and social support relations School environment Bullying Financial resources
Self Relationships Surroundings General quality 12–18 of life
US
5–18 2–18
Youth Quality of Life InstrumentResearch Version (YQOL-R)
Dimensions includeda Physical Emotional Social School
Country of origin
Age (years)
57
Self report 40 Parent report Teacher report
Self report 52/27e Parent report
Self report
Sapin et al. (2005)
RavensSieberer et al. (2005)
Patrick et al. (2002)
Varni et al. (1999)
Number of itemsb Reference
Self report 23 Parent report
Method of response
148
The Pediatric Quality of Life Inventory US Generic Scores Scale (Peds QL 4.0)
Name
. Table 148-1 (continued)
2544 Adolescent Quality of Life in Australia
Adolescent Quality of Life in Australia
7
Western Australian Study of Adolescent QOL
7.1
Study Overview
148
This study aimed to provide a better understanding of adolescent QOL in WA (Meuleners et al., 2001). Although the study included adolescents with and without a chronic illness, the questionnaire and conceptual framework emphasized a > wellness approach rather than illness approach (Raphael et al., 1996). The QOLPAV questionnaire was administered to investigate adolescent QOL (Meuleners et al., 2001). Five dimensions possibly underlying adolescent QOL, the interdependent relations between them (Meuleners et al., 2003), and the effects of covariates on QOL were also investigated (Meuleners and Lee, 2005). Finally, the dynamic changes in adolescent QOL over a 6 month period were documented and modeled (Meuleners and Lee, 2003).
7.2
Research Design
Study participants were recruited through the public and private secondary school system in the Perth metropolitan area of WA during 1999–2000. Stratified sampling with replacement was employed and a total of 30 schools, reflecting diversity in socio-economic status were approached. Of these, 20 schools agreed to participate and questionnaires were sent to the home of each student aged between 10 and 19 years after consent was granted from parents. Of the 500 consented adolescents who initially agreed to participate, 112 with a chronic condition and 251 without returned the questionnaire, giving an overall response rate of 72.6% (Meuleners et al., 2001).
7.3
The Measuring Instrument
The instrument QOLPAV (Raphael et al., 1996) was suitable for this study of healthy adolescents and those with a chronic illness. It was found to be reliable and correlated with measures of adolescent personality, self-reported health status and tobacco and alcohol use (Raphael et al., 1996). The framework underlying this instrument focused on possibilities for adolescents in three areas of life, ‘‘being,’’ ‘‘belonging’’ and ‘‘becoming.’’ As shown in > Figure 148-1, these domains were further divided into nine sub-domains. The domains encompassed a multidimensional approach and emphasized the holistic nature of adolescent QOL. The QOLPAV consisted of 54 items and was a self-administered questionnaire. Each of the 54 items was scored using a five-point > Likert scale. Adolescents rated the importance of each item and their satisfaction with the item. Scores were computed for the three domains, each of the nine sub-domains as well as the overall QOL. It also included items concerning control over the nine sub-domains and nine more items referring to opportunities for improvement and change (Meuleners and Lee, 2003). Additional demographic variables, information related to gender, age, school, grade, family living situation, socio-economic status, presence of a chronic condition, its type and duration and sick time was also collected using the structured questionnaire. Two questions regarding the adolescent’s perception of his/her overall physical health and QOL were also included in the questionnaire as a means of validating the overall QOL score (Meuleners et al., 2001). After review
2545
2546
148
Adolescent Quality of Life in Australia
. Figure 148-1 Essential dimensions and sub-dimensions making up the QOLPAV. This figure summarizes the domains and sub-domains of adolescent QOL underlying the instrument QOLPAV
by two adolescent health experts, the questionnaire incorporating QOLPAV and demographic section was pilot tested on a sample of 20 Australian adolescents, which led to minor modifications. Assessment of reliability and content validity was also undertaken (Meuleners et al., 2001).
7.4
Statistical Analyses and Results
7.4.1
QOL and Its Determinants
After administration of the baseline questionnaire, the results indicated that overall QOL scores for chronically ill and healthy adolescents were positive with both groups reporting an acceptable to very good QOL (Meuleners et al., 2001). There was no significant difference between the two groups. Stepwise regression was undertaken on the combined sample of participants to explore the latent factor determinants of QOL. Age, perceptions of health and control and opportunity were found to be significant determinants of QOL. No significant differences were observed between males and females (Meuleners et al., 2001).
7.4.2
Assessing Measurement Properties Using Structural Equation Modeling
Next, confirmatory factor analysis (CFA) was applied to determine the measurement properties of the latent factors or dimensions underlying adolescent QOL. Recursive structural equation modeling (SEM) was then undertaken to determine the direction and magnitude of the interdependent effects (Meuleners et al., 2003). Exploratory factor analysis (EFA) was conducted as a preliminary to CFA. Weak items, accounting for a low percentage of the variance, were deleted from further analysis. The EFA
Adolescent Quality of Life in Australia
148
results together with a literature review provided guidelines for selecting items to be included in the CFA. Five dimensions consisting of 18 items in total were retained. These domains were ‘‘social,’’ ‘‘environment,’’ ‘‘psychological,’’ ‘‘physical health’’ and ‘‘opportunities for growth and development’’ (Meuleners et al., 2003). A second-order CFA was conducted to address the factor structure of the five dimensions and 18 items. The completely standardized solution to the second-order CFA is shown in > Figure 148-2. Results showed that the 18 items were reliable measures of their respective dimensions with associations ranging from 0.45 to 0.80. The direct effect of the five dimensions on QOL was strong (ranging from 0.68 to 0.90), thus convergent validity was achieved (Meuleners et al., 2003). A recursive SEM was next fitted to estimate the direction and magnitude of the effects among the five identified dimensions of QOL. > Figure 148-3 illustrates the best fitting solution, with standardized estimates of the direct effects of one factor on another. The R2 values refer to the squared multiple correlation coefficients for the structural equations. For example, 51% of the variability in the ‘‘psychological’’ dimension was accounted for simultaneously by the direct effects of the ‘‘environment’’ and the ‘‘health’’ dimension scores, as well as the indirect effect of the ‘‘environment’’ score, mediated by the ‘‘health’’ score (Meuleners et al., 2003). The SEM modeling found that ‘‘environment’’ had significant direct and indirect effects on the other four factors. Significant variables for the ‘‘environment’’ dimension included ‘‘the feeling of safety,’’ ‘‘the home lived in,’’ ‘‘the school the adolescent attends’’ and ‘‘the neighborhood.’’ The ‘‘social’’ dimension had little effect on the other dimensions except ‘‘opportunities for growth and development.’’ Meanwhile, ‘‘opportunities for growth and development’’ were significantly influenced by the ‘‘social,’’ ‘‘health’’ and ‘‘psychological’’ dimensions. The ‘‘psychological’’ dimension also had a strong positive effect on ‘‘opportunities for growth and development’’ and a lesser effect on the ‘‘social’’ dimension (Meuleners et al., 2003). In the next stage, the effects of covariates or personal characteristics on adolescent QOL and its five dimensions ‘‘physical health,’’ ‘‘environment,’’ ‘‘social,’’ ‘‘psychological’’ and ‘‘opportunities for growth and development’’ were assessed using SEM (Meuleners and Lee, 2005). The variables age, control, chronic condition and perception of health were chosen based on literature review. Individual items constituting each of the five identified dimensions were not included due to the small sample size. Instead, composite regression scores from the five factor second-order CFA were used (Meuleners and Lee, 2005). > Figure 148-4 presents the fitted model. The standardized g weights, indicating the strength of the relationship, showed several moderate associations between the personal characteristic variables and the QOL construct. In particular, ‘‘control’’ had a significant positive effect on the QOL construct (g = 0.43, p < 0.05) and poorer ‘‘health’’ exerted a significant negative impact (g = 0.34, p < 0.05). ‘‘Age’’ showed an inverse relationship but was not statistically significant. Support for the five dimensions to be included in adolescent QOL was also evident. All the factor loadings on the QOL construct were high (0.60–0.76) and significant at the 1% level. This analysis confirmed that adolescent QOL is a complex interplay between different factors (Meuleners and Lee, 2005).
7.4.3
Variations in QOL over a Six Month Period
It was of interest to examine how adolescent QOL varied over a 6 month period using overall QOL scores and the five dimension scores on ‘‘physical health,’’ ‘‘environment,’’ ‘‘social,’’
2547
2548
148
Adolescent Quality of Life in Australia
. Figure 148-2 Completely standardized solution to second-order CFA showing loadings and error variances for the 18 items, five underlying latent factors and a single factor. This figure presents the fitted model structure underlying adolescent QOL based on second-order confirmatory factor analysis, where h1 to h5 represent the 5 first-order factors, and j1 represents the one second-order factor, QOL
‘‘psychological’’ and ‘‘opportunities for growth and development.’’ A second identical questionnaire was sent to participants 6 months after they returned the initial questionnaire. Three hundred participants completed both questionnaires (Meuleners and Lee, 2003). Overall QOL scores remained positive after 6 months though there was a significant decrease in the mean QOL score. > Table 148-2 shows the distribution of QOL scores at
Adolescent Quality of Life in Australia
148
. Figure 148-3 Solution to recursive SEM, showing standardized direct effects and covariances. This figure shows the fitted recursive structural equation model of the five dimensions of adolescent QOL, where h1 to h5 represent the 5 first-order factors, and R2 refers to the squared multiple correlation coefficients for the structural equations
baseline and 6 months. It is possible the questionnaire timing could account for the decrease due to the change between winter and summer months in relation to school activities, sport and social activities. Univariate and multivariate tests on the five dimensions of QOL revealed significant changes in the ‘‘social,’’ ‘‘physical health’’ and ‘‘opportunities for growth and development’’ scores for the combined data, however, the effect size scores were all minimal (Meuleners and Lee, 2003). A longitudinal multilevel model was fitted to the data to determine the stability of the significant variables ‘‘age,’’ ‘‘control,’’ ‘‘opportunities’’ and ‘‘perceptions of health.’’ It indicated that 62% of the variation in QOL was due to differences between individuals while 38% was due to within adolescent difference (time difference). Improved ‘‘control’’ and ‘‘opportunities’’ appeared to have a significant positive impact on QOL while increasing ‘‘age’’ and deteriorating ‘‘physical health’’ had the opposite effect (Meuleners and Lee, 2003).
7.5
Contribution to Adolescent QOL Research
This study provided valuable evidence about adolescent QOL in Australia. It focused on a wellness rather than an illness perspective (Raphael et al., 1996) and included
2549
2550
148
Adolescent Quality of Life in Australia
. Figure 148-4 Structural equation model of adolescent QOL. This figure shows the fitted structural equation model of the covariates and dimensions of adolescent QOL, where j1 to j4 represent the independent covariates with g the corresponding effect on the QOL construct
. Table 148-2 Distribution of QOL scores QOL scorea Very problematic (QOL
0.5)
127(35%)
100(33.3%)
Mean (SD)
1.28(0.68)
1.18(0.69)
This table shows the distribution of QOL scores of Australian adolescents at baseline and 6 months. QOL score = (Importance score/3) (Satisfaction score – 3) a QOL score is an overall score based on the adolescent’s rating of the importance of and satisfaction with each item
healthy adolescents in the sample. The study provided encouraging findings that WA adolescents, both with and without a chronic disease, reported their QOL positively (Meuleners et al., 2001). Past research has highlighted the lack of definitive conceptualization of adolescent QOL (Wallander et al., 2001). This study provided validation for adolescent QOL in Australia to
Adolescent Quality of Life in Australia
148
comprise five dimensions, namely, ‘‘physical health,’’ ‘‘environment,’’ ‘‘social,’’ ‘‘psychological’’ and ‘‘opportunities for growth and development’’ (Meuleners et al., 2003). Main results of the SEM indicated that the ‘‘environment’’ dimension has a significant direct and indirect effect on the other four dimensions (Meuleners et al., 2003). Other Australian research also reported associations between adolescent QOL and the environment including neighborhood, school and home factors (Chipuer et al., 2003; Meyers and Miller, 2004). Neighborhood and home factors are not always included in adolescent QOL or HRQOL instruments (Rajmil et al., 2004) despite their potential influence on QOL. The observed effect of the ‘‘psychological’’ dimension on the ‘‘social’’ and ‘‘opportunities for growth and development’’ dimensions highlights the importance of the role of psychological well-being in QOL and implies that public health practitioners should direct health promotion intervention towards this dimension. The authors noted that alternative models could also fit the observed data equally well, resulting in a different interpretation of the measurement of QOL. To confirm the validity of the measurement model and interrelationships among the five dimensions, as well as the effects of the four covariates, replication with large samples was recommended (Meuleners and Lee, 2005; Meuleners et al., 2003). This study was also unique due to its longitudinal design, providing insights into the determinants of adolescent QOL over a 6 month period. Very little is known about changes in QOL over time. The study highlighted the dynamic nature of QOL with 38% of observed variations being attributed to time, emphasizing the necessity of longitudinal study design to capture changes (Meuleners and Lee, 2003). The main finding from the multilevel modeling was the identification of covariates ‘‘physical health,’’ ‘‘control,’’ ‘‘opportunities,’’ and ‘‘age’’ that affected the overall QOL scores over the 6 month period (Meuleners and Lee, 2003). ‘‘Physical health’’ and ‘‘age’’ are commonly acknowledged in the literature as being associated with QOL (Eiser and Morse, 2001; Vingilis et al., 2002), however, the finding provides support for the inclusion of potentially modifiable social determinants of heath, ‘‘control’’ and ‘‘opportunities.’’ Finally, the study showed that CFA and SEM methods are useful for developing hypothetical models and assessing the magnitude and direction of effects among identified factors. It has validated a generic questionnaire, the QOLPAV (Raphael et al., 1996) for measuring adolescent QOL in Australia.
7.6
Limitations
All participants were volunteers, possibly contributing to a self-selection or ‘‘healthy volunteer’’ bias. The sample was also relatively homogenous in their socioeconomic status, which was not a significant predictor of adolescent QOL in the study. The variable was measured based on residential postcodes which might not be sensitive enough to detect any significant relationship. There was also a higher prevalence of females in the study, which might contribute to the lack of observed difference in QOL between genders. It is possible that relationships between the five identified dimensions of QOL are bi-directional but these could not be tested due to the small sample size (Meuleners et al., 2003). In addition, other emerging factors such as community activities, spiritual and coping styles (Eckersley et al., 2006; WHOQOL SRPB Group, 2006), were not included in the QOLPAV.
2551
2552
148 8
Adolescent Quality of Life in Australia
Conclusions and Future Research
Adolescent health information, including QOL, is important for health care researchers and policy makers. It captures the perspective of the individual and provides a holistic approach to the health and well-being of the general adolescent population. Recent developments in the area of adolescent QOL in Australia and worldwide are encouraging and have provided valuable information. Nevertheless, it remains a relatively new area and several gaps in research remain. Although several instruments to measure adolescent QOL have been developed, few of these meet all the standards promoted by the WHO (WHO, 1994) and only one instrument was developed in Australia (Cummins, 1997). New instrument development or the modification and validation of internationally developed instruments would enhance adolescent QOL research in Australia. Consideration should also be given to emerging evidence on factors influencing adolescent QOL and how to best assess them. The WA study demonstrated the dynamic nature of adolescent QOL, yet the two survey time points of 6 months apart might not be sufficient to capture the apparent change in QOL. Future studies should use longitudinal designs that examine the stability of QOL and its potential determinants over a longer period, say, five time points over a 2 year period. Aboriginal and Torres Strait Islanders and adolescents living in rural or remote areas were not included in the WA-based study (Meuleners et al., 2001). However, indicators of health and well-being have suggested that these adolescent groups are at risk of poorer QOL. For example, young indigenous Australians have higher rates of death, injury and some chronic diseases such as asthma and diabetes. They are also more likely to experience obesity, physical inactivity, smoking, imprisonment and lower educational attainment (AIHW, 2007). Similarly, young people living in remote areas have substantially higher rates of death and hospitalization for some health conditions, and are more likely to engage in certain risky health behaviors than their city counterparts (AIHW, 2007). The two groups are potential targets for future QOL research. Finally, it is important to increase the awareness of health care researchers, health promotion professionals and policy makers about adolescent QOL. Information obtained from regular usage of the QOLPAV or another validated measure could form the basis of establishing an adolescent QOL database. This would be invaluable for further research and refinement of the instrument. Relevant factors that positively or negatively influence adolescent QOL are useful for the development of policy and intervention programs, monitoring the QOL status of the general adolescent population and identifying those at risk of low QOL. Putting QOL research findings into practice could enhance the overall health and QOL of adolescents in Australia.
Summary Points It is difficult to understand the health and well-being of the Australian adolescent population from traditional mortality and morbidity statistics. QOL measures the perspective of the adolescents themselves and can provide a holistic view of their health status. Researchers face various challenges because adolescent QOL is a relatively new field and the majority of existing literature is based on adults or the chronically ill.
Adolescent Quality of Life in Australia
148
The unique dimensions of adolescent QOL are only just emerging in the literature.
Several generic assessments of adolescent QOL have been developed in Europe and North America. One Western Australian-based study has contributed to the understanding of adolescent QOL in Australia. It reported that WA adolescents, both with and without a chronic disease, describe their QOL positively. The study provided initial validation for adolescent QOL in Australia to comprise the five dimensions, ‘‘physical health,’’ ‘‘environment,’’ ‘‘social,’’ ‘‘psychological’’ and ‘‘opportunities for growth and development.’’ It was further demonstrated through SEM that these dimensions were interdependent. The longitudinal design of the study revealed the dynamic nature of QOL and identified the potentially modifiable variables of adolescent ‘‘control’’ and ‘‘opportunities’’ as having a significant positive impact on QOL. Australian research should be directed towards the development of adolescent QOL measures based on the most current research or the modification and validation of internationally developed instruments. Aboriginal and Torres Strait Islanders and rural or remote adolescents are at particular risk of poor QOL in Australia so that future research should target these groups. It is recommended to use the findings for developing policy and health intervention programs, monitoring QOL status of the general adolescent population and identifying those at risk of low QOL.
References Australian Institute of Health and Welfare. (2000). Australia’s Health 2000 – The Seventh Biennial Health Report of the Australian Institute of Health and Welfare. AIHW, Canberra. Australian Institute of Health and Welfare. (2003). Australia’s Young People: Their Health and Wellbeing, cat. no. PHE 50. AIHW, Canberra. Australian Institute of Health and Welfare. (2005a). 2004 National Drug Strategy Household Survey – Detailed Findings. Drug statistics series no. 16, cat. no. PHE 66. AIHW, Canberra. Australian Institute of Health and Welfare. (2005b). A Picture of Australia’s children, cat. no. PHE 58. AIHW, Canberra. Australian Institute of Health and Welfare. (2007). Young Australians: Their Health and Wellbeing 2007, cat. no. PHE 87. AIHW, Canberra. Bullinger M. (2002). Restor Neurol Neurosci. 20: 93–101. Bullinger M, Schmidt S, Petersen C, Ravens-Sieberer U. (2006). J Public Health. 14: 343–355. Chipuer HM, Bramston P, Pretty G. (2003). Soc Indic Res. 61: 79–95. Coste J, Bouee S, Ecosse E, Leplege A, Pouchot J. (2005). Qual Life Res. 14: 641–654.
Cummins RA. (1997). Comprehensive Quality of Life Scale – Student (Grades 7–12): ComQol-S5, 5th ed. School of Psychology, Deakin University, Melbourne. Drukker M, Kaplan C, Schneiders J, Feron FJ, van Os J. (2006). BMC Public Health. 6: 133. Eckersley RM, Wierenga A, Wyn J. (2006). Flashpoints & Signpoints: Pathways to Success and Wellbeing for Australia’s Young People. Australia 21, Australian Youth Research Centre, VicHealth, Melbourne. Eiser C, Morse R. (2001). Arch Dis Child. 84: 205–211. Fayers P, Machin D. (2000). Quality of Life: Assessment, Analysis and Interpretation. Wiley, West Sussex, England. Hui EK. (2000). J Adolesc. 23: 189–203. Hurrelmann K, Richter M. (2006). J Public Health. 14: 20–28. Landgraf JM, Maunsell E, Speechley KN, Bullinger M, Campbell S, Abetz L, Ware JE. (1998). Qual Life Res. 7: 433–445. Matza LS, Swensen AR, Flood EM, Secnik K, Leidy NK. (2004). Value Health. 7: 79–92. Meuleners LB, Lee AH. (2003). Pediatr Int. 45: 706–711.
2553
2554
148
Adolescent Quality of Life in Australia
Meuleners LB, Lee AH. (2005). Qual Life Res. 14: 1057–1063. Meuleners LB, Lee AH, Binns CW. (2001). Asia Pac J Public Health. 13: 40–44. Meuleners LB, Lee AH, Binns CW, Lower A. (2003). Qual Life Res. 12: 283–290. Meyers SA, Miller C. (2004). Adolescence. 39: 121–144. Patrick DL, Edwards TC, Topolski TD. (2002). J Adolesc. 25: 287–300. Rajmil L, Herdman M, Fernandez de Sanmamed MJ, Detmar S, Bruil J, Ravens-Sieberer U, Bullinger M, Simeoni MC, Auquier P. (2004). J Adolesc Health. 34: 37–45. Raphael D, Rukholm E, Brown I, Hill-Bailey P, Donato E. (1996). J Adolesc Health. 19: 366–375. Ravens-Sieberer U, Auquier P, Erhart M, Gosch A, Rajmil L, Bruil J, Power M, Duer W, Cloetta B, Czemy L, Mazur J, Czimbalmos A, Tountas Y, Hagquist C, Kilroe J. (2007). Qual Life Res. 16: 1347–1356. Ravens-Sieberer U, Bullinger M. (1998). Qual Life Res. 7: 399–407. Ravens-Sieberer U, Erhart M, Wille N, Wetzel R, Nickel J, Bullinger M. (2006). Pharmacoeconomics. 24: 1199–1220. Ravens-Sieberer U, Gosch A, Abel T, Auquier P, Bellach BM, Bruil J, Dur W, Power M, Rajmil L. (2001). Soz Praventivmed. 46: 294–302. Ravens-Sieberer U, Gosch A, Rajmil L, Erhart M, Bruil J, Duer W, Auquier P, Power M, Abel T, Czemy L, Mazur J, Czimbalmos A, Tountas Y, Hagquist C, Kilroe J. (2005). Expert Rev Pharmacoecon Outcomes Res. 5: 353–364.
Sapin C, Simeoni MC, El Khammar M, Antoniotti S, Auquier P. (2005). J Adolesc Health. 36: 327–336. Schipper H. (1990). Oncology. 4: 51–57. Schipper H, Clinch J, Olweny J. (1996). Quality of Life Studies: Definitions and Conceptual Issues. Lippincott-Raven, New York. Smart D, Sanson A. (2005). Fam Matters. 70: 46–53. Starfield B, Bergner M, Ensminger M, Riley A, Ryan S, Green B, McGauhey P, Skinner A, Kim S. (1993). Pediatrics. 91: 430–435. Varni JW, Seid M, Rode CA. (1999). Med Care. 37: 126–139. Vingilis ER, Wade TJ, Seeley JS. (2002). Can J Public Health. 93: 193–197. Vogels T, Verrips GH, Verloove-Vanhorick SP, Fekkes M, Kamphuis RP, Koopman HM, Theunissen NC, Wit JM. (1998). Qual Life Res. 7: 457–465. Wallander JL, Schmitt M, Koot HM. (2001). J Clin Psychol. 57: 571–585. Ware JE, Jr. (2003). Arch Phys Med Rehabil. 84: S43–S51. Waters E, Salmon L, Wake M. (2000). J Pediatr Psychol. 25: 381–391. World Health Organization, Division of Mental Health. (1994). Measurement of Quality of Life in Children. MNH/PSF/94.5. World Health Organization, Geneva. WHOQOL SRPB Group. (2006). Soc Sci Med. 62: 1486–1497. Zullig KJ, Valois RF, Drane JW. (2005). Health Qual Life Outcomes. 3: 64.
149 Health-Related Quality of Life Among University Students M. Vaez . M. Voss . L. Laflamme 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2557
2
Student Quality of Life in Higher Education – a Transition . . . . . . . . . . . . . . . . . . . 2558
3
Dealing with Stress as a University Student . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2558
4
Student Quality of Life and Well-Being during the Years at University . . . . . . . 2559
5
Students Concerns and Academic Performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2559
6
Theoretical Perspective on Student Retention and Persistence . . . . . . . . . . . . . . . . . 2561
7 7.1
University Student Health, Lifestyle and Quality of Life: a Cohort Study in Sweden . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2561 Student Questionnaire Surveys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2561
8 8.1
Findings from Studies within the Project . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2562 First Year Student Health and Quality of Life- Cross Sectional Studies . . . . . . . . . 2562
9 9.1 9.2
Health and Quality of Life – from First to Third Year at University . . . . . . . . . . 2564 Individual Changes in Health Status, Health Risk Behaviors and Perceived Stress . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2564 Academic Performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2564
10
Health and Quality of Life at the First and the Final Academic Year . . . . . . . . . 2565
11 11.1 11.2 11.3 11.4
Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2566 Self-Rated Psychological Health and General Health at Group Level . . . . . . . . . . . . 2566 Self-Rated Psychological Health and General Health at Individual Level . . . . . . . . 2566 Student Quality of Life in the First and the Final Year of Academic Life . . . . . . . 2567 Associations to Current Quality of Life in the Final Academic Year . . . . . . . . . . . . . 2568
12
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2570
13
Methodological Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2572 Summary Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2572
#
Springer Science+Business Media LLC 2010 (USA)
2556
149
Health-Related Quality of Life Among University Students
Appendix A: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2573 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2573 Education . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2574 Lifestyle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2574 Stress . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2575 Physical Health . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2575 Psychological Health . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2575 General Health . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2576 Quality of Life . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2576
Health-Related Quality of Life Among University Students
149
Abstract: Since the early nineties, universities in Sweden have witnessed a significant expansion in the number of students and now the university-student population constitutes a major part of the young adult population. To study at university has become a common choice for many young adults and as a consequence the student population has diversified in terms of age, gender, ethnical background and socio-economic profile. The transition to higher education has been recognized as a critical period in the formation of the life-pathways and in the establishment of adult behavior. In this context there is a growing interest in studying the impact of enrolment in higher education on the student quality of life and well being. In this prospective study of potential factors related to the student quality of life of 3000 first year full time students (in 1998/ 1999) at a large University in Sweden were followed during their education. The students have filled in three self-administrated questionnaire/covering aspects such as demographic, lifestyle, academics, health status and quality of life, at the end of 1999, 2001 and 2003. The findings emphasized that although, a high proportion of the students rated their health and quality of life as good and the persistence rate was relatively good during the years of university, there was a tendency for deterioration in self rated general and psychological health. This pattern was more pronounced among female students. The results also suggested that the students current quality of life in 2003 was strongly associated with self-rated psychological health, psychosomatic symptoms such as depression, concentration difficulties and perceived stress due to loneliness and doubts about the future. Greater attention should be paid to these factors, in particular by providing students with a work environment conducive to the reduction of stress.
1
Introduction
Conditions during higher education and on entry into working life are of great significance for differences in career opportunities, quality of life and health between men and women, and between social groups. In particular, the transition to college or university is a critical period in the formation of the life-pathways of young adults (Astin, 1984; Lu, 1994). In countries like Sweden, choosing to study at university rather than immediately enter in the labor force is common – since there are relatively favorable economic conditions, low tuition fees, and access to study aid. To receive study support over a period of years, students must pursue their studies with a certain degree of success. Yet, it is intended that a person’s social background and financial circumstances should not present a barrier. Further, access to university should not be influenced by where in the country a person lives. All people in Sweden up to the age of 50 can obtain study aid for university courses up to a maximum of 240 weeks (National Agency of Higher Education, 2000). Since the early nineties, the Swedish Government has invested in higher education in a variety of ways. These include extending the number of places at university or college, establishing new universities, extending adult education, and developing post-high-school vocational training. These ventures, in combination with high unemployment rates in the second half of the nineties, resulted in a rapid expansion in the number of students enrolling in higher education. As a result, nowadays, studying at university is one of the major activities of the young-adult population. Nonetheless, the composition of the student population is undergoing changes, with women constituting a majority and an increasing proportion of students with immigrant background or family responsibility. Even exchange students are on the increase. All the above implies that the student population cannot be regarded as a homogenous group with similar opportunities, living
2557
2558
149
Health-Related Quality of Life Among University Students
under the same conditions, and being offered challenges and career chances of equivalent value (National Agency of Higher Education, 2000).
2
Student Quality of Life in Higher Education – a Transition
There are reasons to expect that the years at college/university mark an important ‘‘rite of passage’’ for young adults (Pascarella and Terenzini, 1991). Whether people go to college/ university immediately after high school, leave a full-time job to pursue an education, or work and study at the same time, they face at least some kind of life adjustment, usually substantial (Shield, 2002). The transition from adolescent-in-family to a university student involves considerable individual and contextual change in almost all areas of life, which may have implications for adult development and health status. Attending a university is an approved move towards personal independence, in particular from parental rules and restrictions. During their years at university, students will explore new interests, analyze innovative social norms, and also search for new social networks. They may seek to clarify their values and beliefs, learn a new set of behaviors, and create new identities and group affiliations (Latham and Green, 1997). The time spent at university is likely to be a period not only of intellectual stimulation and growth, career development, but also of increased autonomy, self-exploration, discovery and social involvement. Students may go through changes in living arrangements, form a partnership, or have (maybe for the first time) responsibility for their own financial welfare (Beder, 1997). Students also have to cope with a new work environment, and the intellectual and relational demands it places on them. In sum this life span is characterized by the high pressure of academic commitments, conflicting role demands, a rapidly changing environment, shifting between different circles of companions and social networks, financial pressures, and often a lack of time-management skills. Additionally, these years demand that individuals more closely define their own career interests, demonstrate performance, and present themselves as attractive on an increasingly competitive and individualistic labor market. At the same time, they have to show adaptive capacity when searching for balance and harmony between their various social roles. There is a sense in which they transfer between different worlds – the university on the one hand and of the family, the peer group, and (for some) a new relationship and the workplace on the other.
3
Dealing with Stress as a University Student
All these factors can generate a relatively high level of stress. This, in turn, has been an important subject in higher-education research for many years, largely because of the specificity of university life in terms of matters related to transitional and developmental issues. Earlier studies have shown that perceived high stress affects both health and academic performance and that those relationships are very strong (Bovier et al., 2004; Campbell et al., 1992; Hudd et al., 2000). Students with a high level of stress tend to perceive themselves as less healthy, to possess a lower level of self-esteem, and to be more prone to adopt a number of unhealthy lifestyles such as binge drinking, high tobacco consumption, drug use, and unsafe sex. A high level of stress has also been linked to a variety of negative outcomes, such as depression, anxiety, eating disorders, and suicide ideation (Dahlin et al., 2005; Ryan and Twibell, 2000). Further, there is a variation in the behaviors and strategies that students use in order to cope with the above mentioned transitional and developmental issues. Some individuals
Health-Related Quality of Life Among University Students
149
change their situation immediately, some accept it, others look for more information, while still others restrain themselves from acting spontaneously (Ryan and Twibell, 2000). Although many stress-related outcomes have been studied e.g., student retention, concentration difficulties, formation of unhealthy behaviors, and poor academic performance (Hirsch and Ellis, 1996; Pascarella and Terenzini, 1991; Seiman, 2005), little is know about how stress impacts on students’ quality of life and well-being (Hudd et al., 2000; Disch et al.; 1999; Chow, 2005).
4
Student Quality of Life and Well-Being during the Years at University
In general, outcomes such as quality of life and well-being of young adults particularly university students have been described by a limited number of researchers. Despite the fact that several previous studies hade stated that psychological distress is elevated among university students, and may be significantly higher than in the general population (Stewart-Brown et al.; 2000), student quality of life and well-being and its determinant and changes during the period of enrolment have received only limited attention in work-environment and public health research internationally and nationally. On the other hand there is a substantial research literature addressing students’ health-related behavior such as alcohol consumption and tobacco/drug use. The scant research that has been done on the student well-being has tended to concentrate on selected segments of students such as medical students (Aktekin et al., 2001; Brimstone et al., 2007; Dahlin et al., 2005; Hojat et al., 2003; Rosal et al., 1997; Vitaliano et al., 1989). Most studies have addressed only the role of academic-related factors in explanation of health disorders and not included other type of factors related to student life outside the university, and have restrictive cross-sectional design. In sum, there is a lack of research on student well being and quality of life in a longitudinal perspective. An improved understanding is essential in design and formation of higher education with superior quality. This area is significant either to examine how optional functioning and performance can be best obtained, and enhanced or to develop opportunity to exert a positive influence on the health of future generations through the assessment and reduction of the risks that may result in poor quality of life and well-being in later working life. In recent years, there has been a public debate, even alarm, about a perceived increase in mental-health problems among university students (Adlaf et al., 2001; Rosal et al., 1997; Raj et al., 2000; Sharkin, 1997). This particularly applies to American students, but may also be of concern in other countries (Aktekin et al., 2001). Such concern has been aggravated by observations made by the staff of student health-care organizations – which have been assembled into bodies of longitudinal material, both nationally and locally. Among increasing problems are emotional and behavioral difficulties (Rosal et al., 1997), a range of psychological problems (O’Malley et al., 1990), and further problems related to developmental issues, relationships and academic skills (Benton et al., 2003). On the other hand, no significant changes have been observed in substance abuse, eating disorders, or chronic mental illness (Robbins et al., 1985).
5
Students Concerns and Academic Performance
An extensive body of literature has the common aim to identify student major concerns, needs and its relationship with academic functioning. As a result a wide range of different concern
2559
2560
149
Health-Related Quality of Life Among University Students
areas have been identified. The most commonly concerns are academic stress, financial issues, career planning and future employment, interpersonal and personal relationships, striking balance between studies, work and family, heavy workload and lack of feedback, pedagogical shortcoming, time management, housing conditions and emotional health. So far much effort has been devoted to understand the relationships between these concerns and student academic functioning sense by including a broad range of individual, social, cognitive and behavioral factors. The consequences of student concern and need in a quality of life and well-being sense are less researched. Pritchard and Wilson (2003) found that both emotional (stress levels, depressive symptomatology, mood, fatigue, self-esteem, perfectionism, optimism) and social health factors were related (having a romantic relationship, type of residence, membership in various campus organization) to students performance and retention. According to this study which consisted of 218 undergraduate students, there was a significant correlation between emotional health and performance (grade point average GPA) regardless of gender. Moreover, students’ emotional health was related to the intention to drop out of college. Conversely social health factors did not predict intention of student departure and social health components affected student’s performance in less extant compared to emotional health (Pritchard and Wilson, 2003). Disch et al. (1999) had employed a multivariate approach to examine the links between students concern and performance. The results highlighted that major sources of concern from most (1) to least (10) importance were issues related to (1) career and employment, (2) time management, (3) physical and mental health, (4) economy, (5) living, (6) sexual behavior, (7) crime and violence (8) learning style, (9) multicultural- gender related and (10) drug and alcohol consumption (Disch et al., 1999). These findings are in line with previous research conducted by Sax, Atin, Korn (1997) who found that while during the past 12 years student involvement in their academic studies has declined, students are more likely to be concerned with their career expectations and how the achievement of an academic degree will improve their prospective quality of life (Sax et al., 1997). Several previous studies had examined the association between self-esteem, social support and student personal well-being. In the comparison of two separated college students samples (1984, 1992) Staats et al. (1995) found that despite the remaining impact of self esteem and social support on students well-being, several changes had ensued over the study period. Students in the sample 1992 had lower levels of subjective well-being, expectations and optimism about the future than students in the 1984 sample (Staats et al., 1995). Harlow and Newcomb (1990) identified several factors, namely purpose in life, intimate relationship, family and peer, perceived opportunity, health and work satisfaction which positively influenced young people’s well-being and quality of life. On the other hand they showed that powerlessness and meaninglessness were two significant barrier factors of young peoples’ quality of life (Harlow and Newcomb, 1990). According to literature on student health the most frequent symptoms from which they suffer are depression, anxiety, sleeping problems, chronic fatigue, and back ache. Results from several studies indicated that the frequency of these symptoms has an influence on selfperception of health and students who have fewer symptoms tend to evaluate their own health significantly higher (Mechanic and Cleary, 1980, Piko, 2000). Further several studies demonstrated that mental and social well-being play an important roll in self-perception of global health and quality of life among young adults (Piko et al., 1997). In a study based on 691 students, Piko et al. (1997) found that the strongest predictor of self-perceived health among the students regardless of gender was psychological well-being. This is supported by several other studies.
Health-Related Quality of Life Among University Students
6
149
Theoretical Perspective on Student Retention and Persistence
The impact of college/university on students has been viewed from different theoretical perspectives i.e., economic, psychological, organizational, educational and sociological. Examples of common conceptual models employed in this research area are given below. Over time, several theories have been developed in order to investigate which factors play a key role in why some students leave and others persist to complete a degree. A major contribution to the study of students’ attrition is made by Vincent Tinto’s integration model (Seiman, 2005; Tinto, 1975; Tinto, 1993; ). The model attempts comprehensively to incorporate the characteristics and procedures that influence students’ decisions to leave college/ university, and how these processes interact to result in attrition. The model includes different types of leaving behaviors, which are identified by Tinto as academic failure, voluntary withdrawal, permanent dropout, temporary dropout, and transfer. Tinto has argued that academic and social integration are the most important predictors of college-student persistence. A match between the academic ability and motivation of the student and the social and academic qualities of the institution fosters academic and social integration into the university. According to his model, students who are not integrated into the university are more likely to develop low commitment, and this – in turn – will result in unsuccessful adjustment, thereby increasing the risk of dropout. According to Astin’s theory of involvement the behavior that students engage in while attending higher education influence their outcome including persistence (Astin, 1984). Bean and Metzner (1985) hypothesize that one or several variables including academic attainment, intent to drop out, previous performance and educational aims and environmental factors is/are the predictors of older or nontraditional students departure decision. According to this theoretical model, environmental factors such as financial issues, the extent of employment, external support, family responsibility and opportunity to transfer have a greater influence on retention than academic related factors (Bean and Metzner, 1985). Although Tinto’s paradigmatic model is an important contribution to the literature on student departure, it does not reveal a complete understanding of the problem and student departure still remains as a major concern for institutions and faculty members.
7
University Student Health, Lifestyle and Quality of Life: a Cohort Study in Sweden
The data used in this study were drawn from the ongoing cohort study ‘‘University student health and quality of life.’’ This cohort is one of the largest longitudinal studies within the university students in Sweden (Vaez et al., 2006a,b; Vaez et al., 2004; Vaez and Laflamme, 2003, 2008; Vaez, 2004). The project was initiated so as to follow up the development of university students’ health status and quality of life during their enrollment years at the university, and the years of establishment on the labor market (See > Figure 149-1).
7.1
Student Questionnaire Surveys
In early 1998, the proposed study was presented to the University’s student health organization and to the committee with responsibility for quality assurance in education
2561
2562
149
Health-Related Quality of Life Among University Students
. Figure 149-1 Study population and design. Three data collections by questionnaire, at Linko¨ping University in Sweden, in 1999 (baseline), 2001(follow up I) and 2003 (follow up II)
and other student-related matters. Students’ health status was surveyed by means of a selfadministered questionnaire. The questionnaire was developed during 1998 on the basis of a number of instruments previously employed in national investigations of issues related to student health. A preliminary version of the questionnaire was developed with the assistance of a group of higher-education/health professionals, student representatives, and members of the quality-assurance committee. After a pre-test on 70 students, the final questionnaire comprised 44 items. Most consisted of forced-response, multiple-choice questions. An overview of the variables considered in the full version of the questionnaire is presented in appendix A.
8
Findings from Studies within the Project
8.1
First Year Student Health and Quality of Life- Cross Sectional Studies
Health status. Our findings from the cross sectional studies of freshmen emphasized that the vast majority of students rated their health as good. The health concerns and problems that were identified included both physical and psychological factors and restriction on life activities, health-care seeking, use of prescription drugs and hospitalization was considerably
Health-Related Quality of Life Among University Students
149
more common for physical than for psychological problems. Gender differences were also observed in health care consumption due to either physical or psychological issues with a significant over-representation of female students. In addition a total of 46% of students reported that they had sought care, and also of those who sought care 6% had been hospitalized and 35% had consumed prescription drugs due to physical complaints during the past years. Women were significantly over-represented in all cases. Among first year students 5% reported that they sought health care due to psychological issues, and among those 4%, all of whom females, reported having been admitted to hospital during the previous year. Further, in this study population, the most common symptoms/complaints were tiredness (35%), anxiety (24%), and concentration difficulties (22%) and symptoms related to psychological problems were more frequent among female than male students. Among the study population both physical and psychological self-rated health correlated strongly with general self-rated health, but the associations between them were not very strong – which suggests that they may capture different aspects of student life. Female students more frequently reported symptoms and had lower ratings of their self-rated health compare to their male peers. Health-risk behaviors. The study populations’ health-risk behaviors were comparable with those of other populations of university students internationally. Despite differences in the measures employed, the proportions of binge drinkers (40%) and tobacco users (24%) were similar to those observed in the USA (O’Malley et al., 1998; Wechsler and Kuo, 2000, 2003; Wechsler et al., 2001, 2002a,b) and in Eastern Europe (Steptoe and Wardle, 2001). We found that frequent drinking (2–4 times/week) and high consumption (7–9 glasses and > 10 glasses) were more common among male students, whereas occasional drinking (once a month) and low alcohol consumption (1–2 glasses and 3–4 glasses/occasion) were more common among female students. Of all respondents 24% reported use of tobacco, with more smokers than expected found among male students. Perceived quality of life. When we studied first year student quality of life we found that regardless of gender the perceived current quality of life was more strongly associated with psychological self-rated health than with physical self-rated health. However, the males’ average score on perceived current quality of life rating in 1999 were lower than those of females. In a comparative study on health status and quality of life assessment of young adults aged 20–34 years, including freshmen (n = 1997 first year students, mean age 23) and those of their working counterparts (n = 947 subjects in full-time employment) we found that first-year university students had a lower score rating on their quality of life and self-rated health than their working peers. A possible confounder in relation to our result was the differences in age distribution between the groups compared. However the differences in rating of quality of life and self-rated health remained unchanged after required age adjustment. In the same study we also found that in both groups and both genders, mean ratings of current quality of life were higher than those of former quality of life and lower than those of expected. Other comparison made between the same groups revealed that first-year university students’ health behaviors differed considerably from those of their working peers. Students smoked and used snuff in smaller proportions, they were not frequent drinkers but they drank higher quantities of alcohol than their working counterparts. Stress. Our findings indicated that the most common stressors were, in descending order of importance, to ‘‘not coping academically,’’ ‘‘poor finances,’’ and the demands imposed by acceptance of ‘‘study aid’’ (which takes the form of student loans in Sweden), ‘‘extra-curricular activities,’’ and ‘‘doubts about the future.’’
2563
2564
149 9
Health-Related Quality of Life Among University Students
Health and Quality of Life – from First to Third Year at University
In the follow up studies, the study population consisted of 1,160 full-time students who had filled in each of two self-administered questionnaire (same questions) dealing with health status and quality of life at the end of 1999 and 2001. The aim of these studies were to describe changing pattern in health status and health behaviors over time and also to examined the role of perceived stress due to a range of academic and non academic factors on student academic performance. Assessments of stress, symptoms, and health were measured at the end of the first (1999) and third year of university (2001). The students’ academic achievement was measured on the basis of whether they had been awarded their degree (about 55% of students on average) by the end of 2003. The total numbers of degrees awarded were taken from the University Registry (as of the end of November 2003).
9.1
Individual Changes in Health Status, Health Risk Behaviors and Perceived Stress
Our findings of individual changes from first to third year of university revealed that regardless of gender, the vast majority of students assessed their psychological and general health as good in both 1999 and 2001. The pattern was somewhat more pronounced for general than psychological health. Further, the occurrence of less-than-good psychological and general health was higher in 2001 compared to 1999. Deterioration in health assessment over time proved to be more important for psychological than general health. We also observed that binge drinking was much more common among males. 46% of male students were classified as binge drinkers at the end of 2001 (including binge drinkers in both 1999 and 2001 and new binge drinkers). Binge drinking became less frequent over time among both male and female students. The occurrence of binge drinking was almost 2.4% higher in 1999. Further we found that the occurrence of smoking was 1.9% higher among females and 1.4% higher among males in 1999 compared to 2001. The proportion of smokers increased between 1999 and 2001 among female students (19%). Oral moist-snuff consumption is much higher among males in general. Use of snuff was higher in 2001 compared to 1999. These changes were more obvious among male students. In the analyses of changing pattern in the sources of stress we found that the most common stressors were related, in descending order of importance, to ‘‘not coping academically,’’ ‘‘poor finances’’ and the demands imposed by acceptance of ‘‘study aid,’’ ‘‘extracurricular’’ activities and ‘‘doubts about the future.’’ Although the sources of stress in question did not change significantly between 1999 and 2003 in rank, the only, and easily understandable, exception was ‘‘doubts about the future,’’ which were more a matter of concern towards the end rather than at the beginning of the university period.
9.2
Academic Performance
Few of the factors investigated had impact on the probability of being awarded a degree, but those from the first year at university that were significantly associated to degree success
Health-Related Quality of Life Among University Students
149
remain so in the third year. Furthermore we found that perceived high level of stress due to academic related factors such as ‘‘not coping academically’’ and ‘‘study support demands’’ were substantial barriers, in particular when combined with concentration difficulties. By contrast, perceived stress caused by non academic factors such as perceived inadequacy for the family and having family issues merge as incentives. Our findings emphasize that students aged 25 or older, females, and those enrolled in comparatively shorter programs (3 years) tend to have higher odds of obtaining a degree.
10
Health and Quality of Life at the First and the Final Academic Year
The data used in this study were extracted from the data set including 5 years follow up of university students. The study population consisted of those who were registered full-time in 1999 through the academic year 2003. In total 445 students filled in questionnaires in 1999, 2001 and 2003. The 26 students among the respondents who in 2003 stated that hey had left their study program or had taken a break from the original study program were excluded. Accordingly, the current study is based on the responses of a total of 419 students. > Table 149-1 presents distribution of students according to a number of socio-demographic characteristics.
. Table 149-1 Socio-demographic characteristics of students in the cohort Socio-demographic characteristics
Faculty of Arts and Sciences n (%)
Faculty of Health Sciences n (%)
Institutet of Technology n (%)
Total
18–20
19 (12.8)
4 (12.1)
56 (23.6)
79 (18.9)
21–24
72 (48.3)
20 (60.6)
158 (66.7)
250 (59.7)
25+
58 (38.9)
9 (27.3)
23 (9.7)
90 (21.5)
Age at baseline (1999)
Gender Female Male
115 (77.2)
21 (63.6)
79 (33.3)
215 (51.3)
34 (22.8)
12 (36.4)
158 (66.7)
204 (48.7)
Having parents with academic background Yes
69 (46.9)
21 (65.6)
142 (60.4)
232 (56.0)
No
78 (53.1)
11 (34.4)
93 (39.6)
182 (44.0)
Married/cohabitant
81 (55.1)
19 (57.6)
96 (40.7)
196 (47.1)
Unmarried
66 (44.9)
14 (42.4)
140 (59.3)
220 (52.9)
Living alone
56 (37.6)
11 (33.3)
97 (40.9)
164 (39.1)
Living with others
93 (62.4)
22 (66.6)
140 (59.1)
255 (60.1)
Marital status in 2003
Living conditions in 2003
The majority of the study group is 21–24 years old and living with others
2565
2566
149
Health-Related Quality of Life Among University Students
11
Results
11.1
Self-Rated Psychological Health and General Health at Group Level
> Table 149-2 demonstrates female and male students’ health assessment at the first and final year of their study program. The results indicate that the majority of both groups of students
. Table 149-2 Health assessment at the first and the final academic year by female and male university students (at the group level with 95% confidence interval (95%)) First academic year (1999)
Final academic year (2003)
Female
Male
Female
Male
Good
81.2 (75.4–85.0)
89.1 (84.0–92.7)
76.4 (70.2–81.7)
87.1 (81.7–91.0)
Moderate
13.0 (9.1–18.2)
10.0 (6.5–14.9)
15.9 (11.5–21.4)
9.0 (5.7–13.7)
5.8 (3.3–9.8)
1.0 (0.3–3.6)
7.7 (4.8–12.1)
4.0 (2.0–7.6)
Good
83.7 (78.0–88.1)
90.0 (85.1–93.5)
83.7 (78.0–88.1)
90.0 (85.1–93.5)
Moderate
13.0 (9.1–18.2)
9.5 (6.1–14.3)
10.1 (6.7–14.9)
6.5 (3.8–10.7)
3.4 (1.6–6.8)
0.5 (0.03–3.2)
6.3 (3.7–10.4)
3.5 (1.7–7.0)
Self-rated health Psychological health
Poor General Health
Poor
The proportion of students who rated their psychological and general health as poor was significantly higher in 2003 than in 1999
rated their psychological and general health as good in the baseline and at follow-up while the proportion of women who rated their health (regardless of type) as good was lower compared to men in both the first and final year of academic life. The data at hand illustrate that the proportion of students who rated their health as poor was significantly higher in 2003 than in 1999. This pattern was somewhat more pronounced for general health than psychological health as well as for men compared to women.
11.2
Self-Rated Psychological Health and General Health at Individual Level
> Table 149-3, shows the distribution of student health assessment, both psychological and general health in their first year through to final academic year at the individual level for female and male, respectively. Regardless of gender and type of health assessment considered, the vast majority of the students rated their health as good in their first year of enrolment and the good health maintained 5 years later with a higher proportion of male students. There was a tendency for deterioration in health (combining those who rated their health as good or moderate solely in 1999, but had lower ratings in 2003) over time was more pronounced for psychological health than for general health, and more pronounced among female students.
Health-Related Quality of Life Among University Students
149
. Table 149-3 Pattern in health assessment of female and male students over time of observations (1999 and 2003) Female (n = 208) Psychological health in 1999 Good Moderate Poor
Psychological health in 2003 Good
Moderate
142
19
8
12
9
6
5
2
5
Female (n = 208) General health in 1999 Good Moderate Poor
General health in 2003 Good
Moderate
153
14
7
16
6
5
1
1
5
Male (n = 201) Psychological health in 1999 Good Moderate Poor
Good Moderate Poor
Poor
Psychological health in 2003 Good
Moderate
159
13
7
16
4
0
0
1
2
Male (n = 201) General health in 1999
Poor
Poor
General health in 2003 Good
Moderate
Poor
166
11
4
14
2
3
1
0
0
> Table
149-3 shows the distribution of student health assessment, both psychological and general health in their first year through to final academic year at the individual level for female and male, respectively. Regardless of gender and type of health assessment considered, the vast majority of the students rated their health as good in their first year of enrolment and the good health maintained 5 years later with a higher proportion of male students
11.3
Student Quality of Life in the First and the Final Year of Academic Life
The mean score for all three time dimensions of quality of life was higher among men in comparison to women in the first year of education while this pattern changed somewhat in the final academic year. Female students rated their current and expected quality of life in the final year better than their male peers. The mean scores of perceived quality of life in the first and the final academic year for females and males separately are illustrated in > Figure 149-2. There was a tendency of an increased mean score of quality of life with time, i.e., both women and men had higher mean scores on current than former quality of life and higher scores on expected than current quality of life. This pattern appeared in both the first and the final year of academic studies.
2567
2568
149
Health-Related Quality of Life Among University Students
. Figure 149-2 Perceived quality of life. The mean scores of perceived quality of life (former, current and expected) measured by ladder scale in 1999 and 2003 by gender. The higher mean scores are indicative of better quality of life
11.4
Associations to Current Quality of Life in the Final Academic Year
Bivariate Correlations was computed in order to study the pairwise associations between the current quality of life rated by student in the final academic year and all variables in the questionnaire. Kendall’s tau-b statistics was used which measure the rank-order association between two scales or ordinal variables. The > Figure 149-3 illustrated the significant correlations between current quality of life in 2003 and individual factors, life style, academic related factors, stressors, health assessment and symptoms. For an overview of included variables see Appendix A. Significant correlations were notable between current quality of life in the final academic year and most of the sources of stress, health assessment and psychological symptoms. Among all significant correlations the strongest associations were found between current quality of life and self-rated psychological health followed by perceived depression, self-rated general health and perceived stress due to loneliness. No correlation coefficient was higher than 0.384.
Health-Related Quality of Life Among University Students
149
. Figure 149-3 (a). The significant correlations between current quality of life in 2003 and individual and lifestyle factors. (b). The significant correlations between current quality of life in 2003 and education-related factors. (c). The significant correlations between current quality of life in 2003 and perceived stress. (d). The significant correlations between current quality of life in 2003 and health assessment and symptoms. Correlation coefficient is a measure of linear association between two variables. The sign of the correlation coefficient indicates the direction of the relationship (positive or negative)
2569
2570
149
Health-Related Quality of Life Among University Students
. Figure 149-3 (contined)
12
Conclusion
The longitudinal study discussed in this chapter considers student concerns health and quality of life during years at university. Follow-up of 3000 first year undergraduate students attending at 52 different study programs during their education is the first step that will hopefully
Health-Related Quality of Life Among University Students
149
lead not only to an increase in knowledge of students’ health and quality of life at Linko¨ping University but also to enhance visibility of students quality of life and well-being issues in higher education nation- and worldwide. Although, in this study a high proportion of students managed to complete their degree, rated their health and quality of life as good and were still enrolled full-time on a study program during the follow up; their health in general and psychological health in particular may be under threat. A lack of previous research focusing on student quality of life and well-being in a longitudinal perspective and also the fact that comparisons between university student and their working peers are uncommon makes it difficult to compare the results found here with those of other studies. Our findings based on time comparative studies indicate that psychological health may be under threat in university-student populations, and the threat seems to be more pronounced among female students. The results also suggested that student current quality of life in 2003 were associated to lifestyle and academic factors only to a limited extent while strong correlations were found with psychological health and perceived psychosomatic symptoms such as depression and concentrations difficulty, perceived stress due to loneliness and doubt about the future. Furthermore we found that perceived stress due to not coping academically and due to study support demands are substantial barriers to students’ academic achievement. The findings from this cohort study contribute to the identification of sub-groups of students who are at particular risk of certain types of problems, and thereby can serve as a basis for initiation of interventions tailored for such groups. High alcohol consumption and binge drinking are examples of such problems, but other relevant issues include a variety of psychological disorders and sources of stress. More detailed psychological profiling of students might help to identify the individuals that might have difficulties psychologically to cope with stress. Changing students’ coping styles is necessary, and should be a first priority. Since there are indications that psychological health may be under threat in universitystudent populations, and since the students quality of life seems to be associated to health habits to only a limited extent, it might be time to switch focus in health studies from specific health-related habits – drinking, in particular – to those more generally related to life as a student. Understanding the manners in which the negative effects on health of stressors like ‘‘not coping academically,’’ ‘‘doubts about the future,’’ and ‘‘loneliness’’ can be counteracted may well help university students achieve their individual potentials, and develop into (even more) autonomous and responsible human beings. The finding that academic achievement is associated, to some extent, by perceived stress and psychological symptoms during the years at university suggests that academic advisors should pay greater attention to these factors, in particular by providing students with a work environment conducive to the reduction of stress. From this point of view the productivity of students is extremely important to higher education systems and now it is more important than ever for universities and institutions to improve the rate of students’ persistence and graduation. But productivity in higher education is not only concerned with increasing the number of places and ever facilitating access to university. The students’ graduation rate is a key component in institutional economy and it is considered as a measure of quality of education. However attention needs to be paid to enhancing achievement and persistence (as performance indictors) in order to decrease the high rate of attrition, which is a major problem for many higher-education institutions.
2571
2572
149 13
Health-Related Quality of Life Among University Students
Methodological Considerations
One set of limitations has to do with the content of the data-collection instrument itself. The questionnaire covers a wide variety of health disorders and health behaviors, but it does not investigate any single one of them in depth. The data at hand are also based on what respondents have knowledge of and