Clinical Reasoning and Evidence-Based Practice: Deliberate Decision-Making by Nurses [1st ed. 2023] 3031270681, 9783031270680

Citation preview

Clinical Reasoning and Evidence-Based Practice Deliberate Decision-Making by Nurses Jos Dobber José Harmsen Margriet van Iersel

123

Clinical Reasoning and Evidence-Based Practice

Jos Dobber • José Harmsen Margriet van Iersel

Clinical Reasoning and Evidence-Based Practice Deliberate Decision-Making by Nurses

Jos Dobber Faculty of Health Amsterdam University of Applied Sciences Amsterdam, The Netherlands

José Harmsen Faculty of Health Amsterdam University of Applied Sciences Amsterdam, The Netherlands

Margriet van Iersel Faculty of Health Amsterdam University of Applied Sciences Amsterdam, The Netherlands

ISBN 978-3-031-27068-0    ISBN 978-3-031-27069-7 (eBook) https://doi.org/10.1007/978-3-031-27069-7 The translation was done with the help of artificial intelligence (machine translation by the service DeepL.com). A subsequent human revision was done primarily in terms of content. 0th edition: © Bohn Stafleu van Loghum 2021 © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Switzerland AG The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

Foreword 1

The professional profile and the training profile of the bachelor nurse emphasise the professional role of the nurse in increasingly complex care situations. This complexity requires thinking and collaborative skills in which clinical reasoning plays a key role. By asking the right questions, the nurse is able to make optimal decisions about diagnosis, aetiology, prognosis and therapy. The complexity of modern healthcare demands from nurses to be able to carry out this reasoning both within the realm of their professional autonomy and within interprofessional problem areas. After all, the nurse is in a position to make frequent and purposeful observations, communicate with the patient and their significant others, monitor their health status, and thereby identify potential health problems at an early stage. In addition, through interventions, the nurse can support the patient’s functioning and contribute to the treatment of other disciplines. This book helps students to develop in clinical reasoning. Because the book teaches students to systematically ask themselves questions when engaging in clinical reasoning, it stimulates the exploratory and investigative aspects of clinical reasoning. After all, in complex situations there are no ready-made answers, and in reasoning a great deal of appeal is made to the student’s investigative ability. In addition, the first chapter of the book exposes the anatomy of clinical reasoning itself. Knowledge of reasoning in general improves the quality of reasoning, among other things because notorious errors in reasoning can be recognised and avoided. In the training profile of the Bachelor of Nursing, the importance of clinical reasoning is reflected in the CanMEDS role of Healthcare Professional. This book teaches the student to apply this competency in an investigative way. Assisted by student Isabella, the student not only learns clinical reasoning, but also to use the method of evidence-based practice (EBP) to involve scientific knowledge in decisions about patient care. The student learns how EBP serves clinical reasoning. The book also demonstrates the breadth of the nursing profession through examples and also through challenging films that illustrate clinical reasoning from different healthcare areas.

v

vi

Foreword 1

The decision to develop the book as much as possible in accordance with the modern educational concept 4C/ID, and to develop the ‘book’ as an interactive medium, means that it fits in with ‘today’s student’. The combination of professional nursing content and education is a success factor in learning clinical reasoning. Wilma  Scholte  op Reimer  is a nurse, epidemiologist, member of the board of governors of Hogeschool Utrecht and professor of Complex Care (Amsterdam UMC-UvA).

Foreword 2

The present book on clinical reasoning links a complex professional task at the heart of the nursing profession to good educational design. The fact that these two components have been brought together in one book is quite unique. The 4C/ID model is attracting a great deal of interest, and there are reasons for this. The evidence-­ based design model fits in well with the need in education to design and develop education that is based on scientific evidence about how people learn. Attention to the development of (professional) competences, knowledge about the integration of knowledge, skills and attitudes, and attention to transfer—i.e. applying what has been learnt in new situations—are telling examples of this. In this book, the complex professional task of clinical reasoning, a thought process of the nurse that proceeds differently with each patient and in each situation, is analysed into sub-factors. These are then worked out into four levels of complexity which serve as a framework for the design of subtask exercises and learning tasks which then guide the approach of teachers. The subtask exercises and learning tasks are presented here at levels 1 and 2, whereby the subtask exercises, entirely in accordance with the principles of the 4C/ID model, serve to support the performance of the entire reasoning task. The teacher’s environment is thus entirely based on insights from the model, and the design of the tasks for students is also consistent with the model. Teachers are also supported by practical aids, such as good examples of case studies and tests. Whether it concerns learning tasks or tests: all teaching materials can be applied flexibly, which means that they can be fitted into any curriculum. Another unique feature is the multimedia offering of patient cases: these real-life and authentic professional problems make reasoning extra interesting and challenging for the student. In short: a range of teaching aids for state-of-the-art education. Paul Kirschner  is a university professor and professor of Educational Psychology at the Open University. Together with Jeroen van Merriënboer, Paul Kirschner wrote the book Ten steps to complex learning (2013) about the 4C/ID model.

vii

Preface

Clinical reasoning is a skill that is close to our hearts. The decisions that result from clinical reasoning largely determine whether the patient receives optimal nursing care. However, clinical reasoning is a complex cognitive core skill of the nursing profession that is not easy for beginning students to learn. Because reasoning happens ‘in your head’, it is difficult to get a grip on the learning process. It is not visible and not graspable. We can think so much that we can’t express it all. Reproducing reasoning and explaining how a reasoning came about is virtually impossible. With this book, we think we have made a good step towards getting a grip on learning clinical reasoning. We have tried to make thinking processes conscious and somewhat visible. The point of clinical reasoning is that through good reasoning the nurse, in good cooperation with the patient, can make the optimal decision for the patient. In this book, decisions are classified into four types, each of which requires its own reasoning process. For each of these decisions, standard questions have been formulated that allow the student to arrive at a responsible clinical decision. By practicing a lot with these standard questions, the student can develop four cognitive schemes with which each clinical decision with accompanying clinical reasoning can be approached in professional practice. The book consists of three parts. The first part, on clinical reasoning, has been written for first and second year bachelor nursing students. The second part, on EBP, is also suitable for later years. The book offers students a basis in reasoning that they can build on in later years. The third part of the book contains various extras. The text of the book consists of several layers. The first layer is the main text; it contains what the student (in the opinion of the authors) should at least know. The second layer is the extras (see videos that are linked to various chapters). The third layer is the educational layer. Based on the 4C/ID model, complexity levels have been established, learning tasks formulated, and subtask exercises made available. The instructor environment can be found in Chap. 12. For the sake of readability, we have chosen to refer to nurses in this book as ‘she’. Of course, in these places ‘he’ can also be used. There are many terms in use to refer to the person receiving care, such as patient and client. In this book, we follow the choice of the International Council of Nurses (ICN). That is why we have chosen the word ‘patient’.

ix

x

Preface

Finally, we would like to thank everyone who contributed to this book by thinking, reading, acting in one of the films, or in any other way: Arie Berg, Dorothee Blaisse, Ype van der Brug, Margriet Dijkmans van Gunst, Sophie Dobber, Tom Dobber, Paul Doedens, Anne Eskes, Astrid Jansen, Martijn Klok, Marieke Kroon, Aster de Lange, Corine Latour, Martin Leeda, Bas van Oort, Hajar Rifi, Özlem Solmaz, Richard Steenborg, Nicky van Oostrum, Michelle Wagemaker, Elsbeth Wielinga, Christine Zwart. Amsterdam, The Netherlands Amsterdam, The Netherlands  Amsterdam, The Netherlands 

Jos Dobber José Harmsen Margriet van Iersel

Introduction

Is it safe to let someone with paraplegia sit alone on a chair? How can I best manage someone with depression? How often should I check this patient and why? Why does one patient stop smoking immediately after a heart attack, while another is having difficulty? Nurses encounter a wide range of questions in their work. They answer these questions with knowledge, which they use to make decisions. Clinical reasoning is a continuous process of critical thinking. It is at the heart of the nursing profession because it is essential for patients that nurses make the right decisions. The importance of clinical reasoning is also reflected in the training profile for all bachelor nursing courses, in which clinical reasoning has been formulated as a central core concept for the CanMEDS role of healthcare provider. This book unravels the thought process that the bachelor nurse performs, and presents this process in four clear-cut decisions that nurses must constantly make during their work: What is going on? How did it come about? What outcome can we expect? What can we do about it? Behind these four main questions are a few sub-­ questions which, when systematically worked through, address all aspects of what is relevant to informed decision-making. Part I Clinical reasoning is about contemplating, considering and designing care delivery. This is important because it leads to deliberate and better nursing care. Clinical reasoning is best learned through practice. It is not a simple skill. The book takes the reader through the experiences of Isabella, a bachelor nursing student. By following her doubts and ideas, the issues involved in learning clinical reasoning become even more apparent. Part II Evidence-based practice (EBP) is a way of making better clinical decisions because it involves current scientific knowledge in reasoning. EBP is at the service of clinical reasoning. Proper use of scientific information requires knowledge of how to find that information efficiently, how to weigh up the quality of the information found, and how to assess what the information means for the care of the patient. Here too, student Isabella helps the reader gain insight: for example, into whether you can assess whether the results of a study can be used in your own practice. Choices about how to provide care are ultimately made in consultation with the patient.

xi

xii

Introduction

Structure and Use of the Book The book offers the theory of clinical reasoning in two ‘layers’. In the first layer, the four key questions are explained in more detail and illustrated with examples, so that it becomes clear how these questions are given content in reasoning. The second layer contains more in-depth knowledge, for example, about thought processes and decision-making, but also about more nursing aspects of reasoning, such as when the nurse can rely on their intuition when making decisions or when more analysis is required. The first layer provides a general overview of the reasoning process. From this first layer the student can click on (digitally) for more content about underlying concepts or browse through the book to the second layer, where clinical reasoning is approached more theoretically. The reader can thus choose and, if desired, switch levels at any time. It is advisable to start with the general line in the first layer and from there to go deeper. The multimedia nature of the book makes this even easier for the reader, as the in-depth content is just a mouse click away from the first layer, and vice versa. The book also contains this second layer. This layer can be found in the book in Part III in depth. The checklists that are part of the in-depth content are also included separately in the Chap. 12 Checklists for quick reference. Terms preceded by an ► in the text: this term is in the glossary: click on it and you get the description. The in-depth texts can be recognised by their layout (Sect. 12.1). Texts in this layout are in the second layer and in Chap. 12. The book includes several resources as part of the chapters in Part I. In process worksheets, all questions of the reasoning process are clearly listed. During the reasoning process these can serve as a checklist. The cases provided in the form of short films ensure that the reality of healthcare provision is presented in a realistic manner. You can see how patients present their complaints and problems, how a meeting is conducted and how nurses think aloud about ‘their’ patient. The professional practice is thus realistically portrayed. In addition, there are assignments related to the dilemmas of professional practice. In addition to the main text (first layer) and the digital environment (second layer) for students, the book also contains material for teachers. This material has been developed from the 4C/ID design model. The 4C/ID model is based on recent cognitive-psychological knowledge about learning. The model offers tools for systematically designing education for a complex professional task such as clinical reasoning. An important starting point is the provision of whole tasks that are optimally aligned with the requirements of professional practice. The teacher’s environment of the book includes a brief introduction to the 4C/ID model and an explanation of how the model is used in the book. Teaching materials in the form of learning tasks and subtask exercises can be inserted into curricula as desired. There are also good examples of detailed case studies and test criteria at two levels of complexity. The material is flexible in its application, based on the fact that the structure of the study programmes differs as much as the wishes of the instructors. The teacher’s environment can be found in Chap. 13.

Contents

Part I Clinical Reasoning 1

 Background Knowledge in Clinical Reasoning ��������������������������������������   3 1.1 What Is Clinical Reasoning?��������������������������������������������������������������   3 1.2 How to Use Clinical Reasoning?��������������������������������������������������������   4 1.3 Trusting Relationship��������������������������������������������������������������������������   5 1.4 The ICF as a Thought Model in Clinical Reasoning��������������������������   6 1.4.1 What Is the ICF Diagram? What Do the ICF Terms Mean?��������������������������������������������������������������������������   6 1.4.2 How Does the ICF Support Structuring and Reasoning? ������  10 1.5 Analysing and Answering the Four Basic Questions of Clinical Reasoning��������������������������������������������������������������������������  12 1.5.1 What Are the Standard Questions in Diagnostic Reasoning? ������������������������������������������������������  12 1.5.2 What Are the Standard Questions in Etiological Reasoning? ������������������������������������������������������  13 1.5.3 What Are the Standard Questions in Prognostic Reasoning?��������������������������������������������������������  13 1.5.4 What Are the Standard Questions in Therapeutic Reasoning?������������������������������������������������������  14 1.5.5 What Is the Course of the Clinical Reasoning Process? ��������  15 1.6 The ICN Definition of Nursing as a Framework for Nursing Decisions ������������������������������������������������������������������������  16 1.6.1 What Is Meant by Autonomous Action and What by Collaborative (or Participatory) Action?��������������������������������  17 1.6.2 How Does This Division of the Nursing Professional Domain into Autonomous and Collaborative Action Affect Clinical Reasoning? ����������������������������������������������������  18 1.7 Diversity-Sensitive Reasoning������������������������������������������������������������  19 1.7.1 What Areas of Diversity Support Diversity-Sensitive Reasoning?������������������������������������������������������������������������������  19 1.7.2 Why Is Diversity-Sensitive Reasoning Important?����������������  20 1.7.3 How Does Diversity-Sensitive Reasoning Fit with the ICF?��������������������������������������������������������������������������  21 xiii

xiv

Contents

1.8 Evidence-Based Practice (EBP)����������������������������������������������������������  21 1.8.1 How Can Evidence-Based Practice Improve Clinical Reasoning and Skill Implementation?������������������������������������  21 1.8.2 What Is Evidence-Based Practice and What Components Does It Consist Of?����������������������������������������������������������������  22 1.8.3 What Is the Methods of Evidence-Based Practice? What Kind of Knowledge Does EBP Lead to?��������������������������������  23 1.9 The Use of Hypotheses in Clinical Reasoning ����������������������������������  24 1.9.1 What Is Meant by ‘Hypothesis Formulation’ and by ‘Hypothesis Testing’?��������������������������������������������������  24 1.9.2 How Can I Formulate Hypotheses and Test Criteria About Health Problems?��������������������������������������������������������������������  25 1.9.3 How Can Hypotheses About Etiology Be Tested?������������������  26 1.9.4 How Can Goals of Nursing Care Be Assessed?����������������������  26 1.9.5 What Should the Hypotheses Be About When Reasoning About Interventions?��������������������������������������������������������������  27 1.10 The Thought Structure of Clinical Reasoning������������������������������������  27 1.10.1 How Do We Think? The Dual Processing Model and Pitfalls in Decision-Making ��������������������������������������������  27 1.10.2 How Do We Think? Biases as Threats to Reasoning��������������  29 1.10.3 Intuition and Analysis ������������������������������������������������������������  30 1.10.4 Background Knowledge and Foreground Knowledge������������  32 Glossary ������������������������������������������������������������������������������������������������������  33 References����������������������������������������������������������������������������������������������������  39 2

 Diagnostic Decision: What Is the Matter with the Patient?������������������  41 2.1 What Is a Diagnosis?��������������������������������������������������������������������������  41 2.2 What Is Diagnostic Reasoning?����������������������������������������������������������  42 2.3 How Are the Health Problem, Clues and Symptoms Incorporated into the Diagnosis?������������������������������������������������������������������������������  43 2.4 When Should I Use Diagnostic Reasoning? ��������������������������������������  43 2.5 How Do I Get a Diagnosis?����������������������������������������������������������������  44 2.6 How Do I Find Out a Patient’s Symptoms?����������������������������������������  45 2.7 How Can I Assess the Severity and Importance of the Symptoms? ������������������������������������������������������������������������������  46 2.8 How Do I Get from Symptoms to an Overview of Possible Health Problems?��������������������������������������������������������������������������������  46 2.9 How Do I Get Information About the Characteristic Symptoms of the Diagnosis?��������������������������������������������������������������������������������  50 2.10 How Can I Test the Defining Characteristics?������������������������������������  50 2.11 How Do I Process All This Information into a Diagnosis?����������������  50 2.12 What About Diagnoses in the Area of Prevention?����������������������������  51 Glossary ������������������������������������������������������������������������������������������������������  52 Cases and Learning Tasks����������������������������������������������������������������������������  55 Learning Task 1: Sanne��������������������������������������������������������������������������   55 Process Worksheet for the Sanne Case��������������������������������������������������   56

Contents

xv

Process Worksheet Diagnostic Reasoning ��������������������������������������������   56 Conclusion ��������������������������������������������������������������������������������������������   58 Learning Task 2: Mr Bakker������������������������������������������������������������������   59 References����������������������������������������������������������������������������������������������������  61 3

 Etiological Decision: What May Have Caused the Problem?����������������  63 3.1 What Is Etiology? What Are Etiological and Related Factors?����������  63 3.2 What Is Etiological Reasoning?����������������������������������������������������������  64 3.3 How Are Etiological Factors Incorporated into the Diagnosis? ��������  65 3.4 How Should I Reason Etiologically in a High-Risk Diagnosis?��������  66 3.5 How Do I Find the Etiological Factors? ��������������������������������������������  66 3.6 How Can I Estimate the Importance of the Etiological Factors?��������  67 3.7 How Can I Test Whether the Factors Found Do Indeed Cause or Perpetuate the Problem? ��������������������������������������������������������������������  67 Glossary ������������������������������������������������������������������������������������������������������  68 Cases and Learning Tasks����������������������������������������������������������������������������  69 Learning Task 3: Simone Case��������������������������������������������������������������   70 Process Worksheet for the Simone case������������������������������������������������   70 Process Worksheet Diagnostic Reasoning ��������������������������������������������   70 Conclusion ��������������������������������������������������������������������������������������������   72 Process Worksheet Etiological Reasoning ��������������������������������������������   72 Conclusion ��������������������������������������������������������������������������������������������   73 Learning Task 4: Laila Case������������������������������������������������������������������   74 References����������������������������������������������������������������������������������������������������  74

4

Prognostic Decision: What Can We Achieve? ����������������������������������������  77 4.1 What Is a Prognosis? What Is Prognostic Reasoning? ����������������������  77 4.2 Why Is a Prognosis Important?����������������������������������������������������������  78 4.3 Should Prognosis Be Reasoned About First and Then Therapy, or Vice Versa?��������������������������������������������������������������������������������������  79 4.4 What Is the Difference Between Risk Factors, Etiological Factors and Prognostic Factors? ��������������������������������������������������������  79 4.5 How Do I Know What Factors Influence the Prognosis? ������������������  79 4.6 What Should I Do Once I Have Identified the Prognostic Factors? ����������������������������������������������������������������������������������������������  81 4.7 Which Types of Prognosis Are Possible? ������������������������������������������  81 4.8 How Can I Formulate Patient Goals from a Prognosis? ��������������������  81 4.9 How Do I Formulate a Patient Goal?��������������������������������������������������  82 4.10 How Can I Use the Goals in Evaluation?�������������������������������������������  83 Glossary ������������������������������������������������������������������������������������������������������  84 Cases and Learning Tasks����������������������������������������������������������������������������  85 Learning Task 5: Mrs Kimanski Case����������������������������������������������������   86 Process Worksheet for the Mrs Kimanski Case ������������������������������������   87 Process Worksheet Diagnostic Reasoning ��������������������������������������������   87 Conclusion ��������������������������������������������������������������������������������������������   89 Process Worksheet Etiological Reasoning ��������������������������������������������   89

Contents

xvi

Conclusion ��������������������������������������������������������������������������������������������   90 Process Worksheet Prognostic Reasoning ��������������������������������������������   91 Conclusion ��������������������������������������������������������������������������������������������   92 Learning Task 6: Mr Ewijk Case ����������������������������������������������������������   93 References����������������������������������������������������������������������������������������������������  94 5

 Therapeutic Decision: What Can We Do About the Problem?��������������  95 5.1 What Is the Difference Between a Therapy and an Intervention?������  95 5.2 Is Therapy Also Performed by Nurses?����������������������������������������������  96 5.3 What Is Therapeutic Reasoning?��������������������������������������������������������  96 5.4 How Do I Know What Interventions There Are?��������������������������������  97 5.5 How Do I Know What Effects I Can Expect from an Intervention? And How Do I know If There Is Strong Evidence for These Effects?���������������������������������������������������������������  98 5.6 How Do I Know If the Evidence Applies to My Patient?������������������  98 5.7 What Is Meant by the Point of Intervention?��������������������������������������  99 5.8 How Do I Find Out the Mechanism of Change of the Intervention? ����������������������������������������������������������������������������  99 5.9 How Can I Assess the Feasibility of the Intervention? ���������������������� 100 5.10 How Do I Choose from Several Alternative Interventions?���������������� 100 5.11 How Do I Get from Choosing the Intervention to Implementing It?���������������������������������������������������������������������������� 101 Glossary ������������������������������������������������������������������������������������������������������ 101 Cases and Learning Tasks���������������������������������������������������������������������������� 103 Learning Task 7: Sven Case������������������������������������������������������������������  103 Process Worksheet for the Sven Case����������������������������������������������������  104 Process Worksheet Diagnostic Reasoning ��������������������������������������������  104 Conclusion ��������������������������������������������������������������������������������������������  106 Process Worksheet Etiological Reasoning ��������������������������������������������  106 Conclusion ��������������������������������������������������������������������������������������������  107 Process Worksheet Prognostic Reasoning ��������������������������������������������  107 Conclusion ��������������������������������������������������������������������������������������������  109 Process Worksheet Therapeutic Reasoning�������������������������������������������  109 Learning Task 8: Mr Bachoni Case��������������������������������������������������������  112 References���������������������������������������������������������������������������������������������������� 112

Part II Evidence-Based Practice (EBP) 6

 Clinical Reasoning and Evidence-Based Practice (EBP) ���������������������� 115 6.1 What Is Evidence-Based Practice? How Does Evidence-­Based Practice Help Improve My Clinical Reasoning?���������������������������������������������������������������������������� 115 6.2 What Is the Method of Evidence-Based Practice?������������������������������ 116 6.3 How Can I Translate a Clinical Problem into an Answerable Question?�������������������������������������������������������������������������������������������� 117 6.4 How Do I Know What the Best Evidence Is?������������������������������������ 117

Contents

xvii

6.5 Where Do I Find the Evidence?���������������������������������������������������������� 119 6.6 How Can I Weigh the Found Evidence on Methodological Quality and Applicability in My Own Practice?�������������������������������� 119 Glossary ������������������������������������������������������������������������������������������������������ 119 References���������������������������������������������������������������������������������������������������� 120 7

What May Be the Matter with the Patient: How to Critically Appraise Articles on Diagnostic Tests?���������������������������������������������������� 123 7.1 What Is a Diagnostic Test?������������������������������������������������������������������ 123 7.2 How Can I Assess the Validity of a Diagnostic Measuring Instrument? ���������������������������������������������������������������������������������������� 124 7.3 What Is Meant by ‘Sequential’ or ‘Random Sample’? Why Is It So Important?���������������������������������������������������������������������������������� 125 7.4 What Is a Valid Reference Test? �������������������������������������������������������� 125 7.5 Why Do the Evaluators of the Test Have to Be Blinded? ������������������ 125 7.6 In What Situation Is the Requirement Dropped That Both Tests Be Performed? In That Case, How Can the Comparison Be Done Properly?������������������������������������������������������������������������������ 126 7.7 What Is ‘Selective Dropout’? What Should I Look for in Assessing Dropout?������������������������������������������������������������������������ 126 7.8 How Can I Assess the Value of the Measuring Instrument Under Study? ������������������������������������������������������������������������������������������������ 126 7.9 How Can I Assess Whether I Can Apply the Researched Diagnostic Instrument to My Patients?���������������������������������������������� 127 Glossary ������������������������������������������������������������������������������������������������������ 128 References���������������������������������������������������������������������������������������������������� 130

8

 What May Be the Cause: How to Critically Evaluate Articles on Etiological Factors?����������������������������������������������������������������������������������� 131 8.1 In What Kind of Sources Can I Find the Best Evidence About Etiological Factors?���������������������������������������������������������������������������� 131 8.2 What Does a Cohort Study of Etiological Factors Involve? �������������� 132 8.3 What Does a Case-Control Study of Etiologic Factors Involve?�������� 133 8.4 How Can I Determine the Quality of the Cohort Study?�������������������� 134 8.5 How Do I Know If the Two Groups Are Comparable?���������������������� 134 8.6 What Is Selection Bias? How Can I Estimate the Risk of Bias in Research Results?�������������������������������������������������������������� 135 8.7 What Is Information Bias? How Can I Estimate the Risk of Research Results Being Biased by Information Bias? ������������������ 135 8.8 How Do I Know If the Period of Time Until Follow-Up Is Long Enough?���������������������������������������������������������������������������������������������� 136 8.9 Why Does the Intermediate Dropout of Participants Threaten the Validity of the Study? How Can I Determine Whether Too Many Participants Have Dropped Out?�������������������������������������������������������� 136 8.10 What Is ‘Adjustment for Significant Risk Factors’? How Can I Check Whether These Factors Have Been Controlled for?�������������� 136

xviii

Contents

8.11 Do the Same Requirements Apply to Determining the Validity of a Case-Control Study?�������������������������������������������������������������������� 137 8.12 What Is Meant by New Cases? Why Is It Important That the Cases Are New?���������������������������������������������������������������������������� 137 8.13 What Is Meant by Misclassification? How Can I Find Out If There Is a High Risk of Misclassification? ������������������������������������ 138 8.14 How Can I Assess the Magnitude of the Effect of the Etiological Factor?������������������������������������������������������������������������������������������������ 138 8.15 How Can I Determine If the Point Estimate Is Really the Actual Effect? ������������������������������������������������������������������������������������������������ 139 8.16 What Does the Confidence Interval Say About Statistical Significance? �������������������������������������������������������������������������������������� 139 8.17 How Do I Know If I Can Use the Results for My Patients? �������������� 141 8.18 What Is the Difference Between Association and Causation?������������ 141 Glossary ������������������������������������������������������������������������������������������������������ 142 References���������������������������������������������������������������������������������������������������� 145 9

What Can We Achieve: How to Critically Review Articles on Prognostic Research? �������������������������������������������������������������������������� 147 9.1 What Are Prognostic Factors? Why Are They Important? ���������������� 147 9.2 In What Sources Can I Find the Best Evidence for Prognostic Factors? ���������������������������������������������������������������������������������������������� 148 9.3 How Can I Determine the Quality of a Cohort Study of Prognostic Factors? ���������������������������������������������������������������������������������������������� 148 9.4 What Is Important in the Description of the Patient Group?�������������� 149 9.5 How Can I Verify That the Follow-Up Measurement Was Conducted with a Sufficient Number of Study Participants? ������������ 150 9.6 How Do I Know If the Period to Follow-Up Is Long Enough to Measure the Effect of the Studied Factor on the Course and Outcome of the Studied Health Condition? �������������������������������� 151 9.7 How Can I Assess the Quality of the Measurements of the Outcomes?������������������������������������������������������������������������������������������ 151 9.8 How Can I Assess the Quality of the Measurements of the Potential Prognostic Factor? ���������������������������������������������������� 151 9.9 What Is the Value of Initial Research on a Prognostic Factor? How Does It Differ from Later Research?������������������������������������������ 152 9.10 How Can I Determine the Importance of the Study Results in Relation to the Prognostic Factor for the Patients?������������������������ 152 9.11 How Can I Find Out If the Study Results Can be Used for My Own Patients? ������������������������������������������������������������������������ 153 Glossary ������������������������������������������������������������������������������������������������������ 154 References���������������������������������������������������������������������������������������������������� 155

Contents

xix

10 What  Can We Do About the Problem: How to Critically Appraise Articles on Interventions? ������������������������������������������������������������������������ 157 10.1 Randomised Controlled Trial and Controlled Clinical Trial������������ 157 10.1.1 What Is a Randomised Controlled Trial (RCT)?���������������� 158 10.1.2 How Can I Critically Evaluate the Quality of a Randomised Controlled Trial (RCT)? How Can I Determine the Applicability of the Intervention to Patients in My Practice?������������������������������������������������������ 160 10.1.3 What Is Meant by Randomisation? ������������������������������������ 160 10.1.4 Why Is Randomisation Important?�������������������������������������� 161 10.1.5 What About a Controlled Clinical Trial?���������������������������� 161 10.1.6 Why Is the Comparability of the Participant Groups Important?���������������������������������������������������������������������������� 162 10.1.7 What Is Meant By Blinding?���������������������������������������������� 162 10.1.8 Why Is Treatment Comparability Important?���������������������� 163 10.1.9 How Is the Quality of the RCT Affected by Sample Size and Completeness of Follow-Up?������������������������������� 163 10.1.10 What Is Meant By Validity? How Can the Validity of an RCT Be Assessed?������������������������������������������������������ 165 10.1.11 How Can the Results of an RCT Be Interpreted?���������������� 166 10.1.12 What Is Meant by Significance?������������������������������������������ 168 10.1.13 What Does the p-Value Say About Statistical Significance?������������������������������������������������������������������������ 168 10.1.14 What Does the Confidence Interval Say About Statistical Significance?������������������������������������������������������ 169 10.1.15 What Role Do Adverse Effects Play in Assessing an Intervention?������������������������������������������������������������������������ 171 10.1.16 How Can It Be Assessed Whether the Experimental Intervention Can Be Generalised To, and Applied In, My Patient Care? ���������������������������������������������������������������� 171 10.2 Synthesis: What Is the Value of Systematic Reviews?���������������������� 173 10.2.1 What Is a Systematic Review?�������������������������������������������� 173 10.2.2 How Can the Value of a Systematic Review Be Determined?������������������������������������������������������������������������ 174 10.2.3 How Can the Research Question of the SR Be Appraised?���������������������������������������������������������������������� 175 10.2.4 What Requirements Must the Systematic Search for Relevant Studies Meet in a Systematic Review? ���������� 175 10.2.5 What Are the Requirements for the Selection Procedure of the Articles for the SR?���������������������������������� 177 10.2.6 How to Assess Whether a Systematic Review Is Based on Good or Weak RCTs?������������������������������������������ 177 10.2.7 What Is Data Extraction? What Are the Requirements for Reporting on the Data Extraction Process in a Systematic Review?������������������������������������������������������ 179

Contents

xx

10.2.8 In What Way Should the Original Studies Be Described in the SR? ���������������������������������������������������������� 179 10.2.9 What Is Heterogeneity? How Can It Be Verified That Heterogeneity Has Been Correctly Estimated and Correctly Incorporated into an SR?������������������������������������ 179 10.2.10 How Should the Results of an SR Be Reported?���������������� 182 10.2.11 How to Assess the Validity of an SR?���������������������������������� 183 10.2.12 How Can the Results of an SR Be Interpreted? How to Assess the Applicability of the Intervention in Patient Care?�������������������������������������������������������������������� 183 10.3 Evidence-Based Guidelines�������������������������������������������������������������� 183 10.3.1 What Is a Guideline? ���������������������������������������������������������� 183 10.3.2 What Value Do Guidelines Have for Clinical Reasoning?������������������������������������������������������������ 184 10.3.3 How Can the Quality of a Guideline Be Critically Assessed?���������������������������������������������������������������������������� 184 Glossary ������������������������������������������������������������������������������������������������������ 185 References���������������������������������������������������������������������������������������������������� 191 11 How  to Better Understand Patients’ Experience: Understanding Psychosocial Processes in Healthcare Through Qualitative Research���������������������������������������������������������������������������������� 193 11.1 What Is Qualitative Research? What Is the Difference Between Quantitative and Qualitative Research?�������������������������������������������� 193 11.2 What Is the Most Important Method in Qualitative Research?�������� 194 11.3 What Are the Ways of Collecting Data in Qualitative Research?������������������������������������������������������������������������������������������ 194 11.4 How Are Data Analysed in Qualitative Research? �������������������������� 195 11.5 What Are the Pitfalls of Conducting Qualitative Research?������������ 196 11.6 What Are Quality Criteria for Qualitative Research? ���������������������� 197 11.6.1 Truth Value or Credibility���������������������������������������������������� 197 11.6.2 Applicability or Transferability ������������������������������������������ 198 11.6.3 Repeatability������������������������������������������������������������������������ 199 11.6.4 Neutrality or Confirmability������������������������������������������������ 199 11.7 How Does Qualitative Research Contribute to Clinical Reasoning?���������������������������������������������������������������������������������������� 200 Glossary ������������������������������������������������������������������������������������������������������ 200 References���������������������������������������������������������������������������������������������������� 202 Part III Second Layer with Extras 12 In-Depth Texts�������������������������������������������������������������������������������������������� 205 12.1 Examples of the Four Types of Questions���������������������������������������� 205 12.2 Making a Deliberate Decision���������������������������������������������������������� 205

Contents

xxi

12.3 Conditions for Critical Thinking by the Nurse �������������������������������� 206 12.3.1 Use Good Information�������������������������������������������������������� 206 12.3.2 Assess and Weigh Information Properly ���������������������������� 207 12.3.3 Thorough Professional Knowledge ������������������������������������ 207 12.4 Examples of Standard Questions That Are Asked in Order to Eventually Answer the Basic Question�������������������������������������������� 207 12.5 Example of Body Structures and Their Body Functions (ICF)�������� 208 12.6 Deepening Knowledge of Environmental Factors���������������������������� 208 12.7 Explanation of the Standard Questions for a Diagnostic Decision�������������������������������������������������������������������������������������������� 209 12.8 Explanation of the Standard Questions for an Etiological Decision�������������������������������������������������������������������������������������������� 210 12.9 Explanation of the Standard Questions for a Prognostic Decision�������������������������������������������������������������������������������������������� 211 12.10 Explanation of the Standard Questions for a Therapeutic Decision�������������������������������������������������������������������������������������������� 212 12.11 Forms of Collaboration Between Professionals: Multiprofessional, Multidisciplinary, Interdisciplinary or Interprofessional ������������������������������������������������������������������������������ 215 12.12 Further Elaboration of the Nursing Professional Domain���������������� 216 12.13 Short Case with Autonomous and Participative Action�������������������� 217 12.13.1 Autonomous Reasoning and Acting������������������������������������ 217 12.13.2 Participatory Reasoning and Action������������������������������������ 217 12.14 Explanation of the Seven Diversity Aspects ������������������������������������ 218 12.15 Example of Blood Pressure Measurement by Isabella: First Part�������������������������������������������������������������������������������������������� 219 12.16 Blood Pressure Measurement by Isabella: Second Part�������������������� 223 12.16.1 Health Risk: The Whole Is More Than the Sum of the Parts �������������������������������������������������������������������������� 223 12.17 Considering Hypothetical Diagnoses������������������������������������������������ 225 12.18 Example of Hypothesis Formation and Test Criteria of Etiology���������������������������������������������������������������������������������������� 226 12.19 Example of Hypothesis Formation and Test Criteria for Patient Goals ������������������������������������������������������������������������������ 227 12.20 Example of Hypothesis Formation and Test Criteria for Interventions ������������������������������������������������������������������������������������ 228 12.21 Risk of Death for Smokers and Ex-Smokers Compared with People Who Never Smoked������������������������������������������������������ 229 12.22 Examples of the Use of Relevant and Irrelevant Information���������� 230 12.22.1 Use of Irrelevant Information���������������������������������������������� 230 12.22.2 Use of Relevant Information ���������������������������������������������� 231 12.22.3 Analysis ������������������������������������������������������������������������������ 231

xxii

Contents

12.23 Test Your Use of the Thinking Systems Through the Cognitive Response Test������������������������������������������������������������������������������������ 231 12.23.1 Questions ���������������������������������������������������������������������������� 231 12.24 Example of Mrs. Brown Elaborated in the Dual Processing Model������������������������������������������������������������������������������������������������ 232 12.24.1 Patient Situation������������������������������������������������������������������ 232 12.24.2 Question������������������������������������������������������������������������������ 232 12.25 More Examples of Biases ���������������������������������������������������������������� 233 12.26 Brief Overview of the Differences Between Intuition and Analysis�������������������������������������������������������������������������������������� 234 12.27 Examples of When a More Intuitive Approach Is Preferred������������ 234 12.28 Standards for BMI and Abdominal Size ������������������������������������������ 235 12.28.1 BMI�������������������������������������������������������������������������������������� 235 12.28.2 Waist Circumference: High Risk of Obesity-Related Conditions���������������������������������������������������������������������������� 235 12.29 In-Depth Text About Monitoring: Diagnosis - Monitoring Ongoing Diagnostic Reasoning�������������������������������������������������������� 235 12.30 Example of How Knowledge Leads to Further Questions��������������� 236 12.31 Brief Description of the Area Within Which Nurses Make Autonomous Diagnosis and Practice Participatory Diagnostic Reasoning������������������������������������������������������������������������������������������ 236 12.32 Deepening on Sufficient and Necessary Causes ������������������������������ 237 12.33 Example of a Nutritional Instrument: The Short Nutritional Assessment Questionnaire (SNAQ)�������������������������������������������������� 238 12.34 Explanation of the Six Types of Results ������������������������������������������ 238 12.35 Isabella’s Elaboration on Smoking Cessation Prognosis and Goals by Mrs. Jacobi������������������������������������������������������������������ 239 12.36 Autonomous and Participatory Interventions ���������������������������������� 240 12.36.1 Autonomous Nursing Interventions������������������������������������ 240 12.36.2 Participatory Interventions, Carried Out By the Nurse ������ 240 12.37 Elaboration of Isabella’s Search for Exercise Interventions ������������ 241 12.37.1 Intervention�������������������������������������������������������������������������� 242 12.37.2 Impact���������������������������������������������������������������������������������� 242 12.37.3 The Evidence Applied to Isabella’s Patient ������������������������ 242 12.37.4 Focus������������������������������������������������������������������������������������ 242 12.37.5 Mechanism of Change�������������������������������������������������������� 243 12.37.6 Feasibility of the Intervention��������������������������������������������� 243 12.37.7 Patient Acceptability������������������������������������������������������������ 243 12.38 Examples of Mechanisms of Change������������������������������������������������ 243 12.38.1 Aspirin as a Painkiller��������������������������������������������������������� 243 12.38.2 Motivational Interviewing �������������������������������������������������� 244

Contents

xxiii

12.39 Explanation of Three Questions About the Applicability of Specific Scientific Knowledge������������������������������������������������������ 244 12.39.1 Is the New Information Emerging from These Scientific Articles Likely to be True?�������������������������������������������������� 244 12.39.2 What Does This Information Tell About Patient Care?������ 244 12.39.3 Can This Information be Used for One’s Patient?�������������� 245 12.40 More Examples of Determinants������������������������������������������������������ 246 12.41 More Examples of Questions Using the PICO System�������������������� 246 12.42 The Use of PICO for Questions on Diagnosis, Etiology, or Prognosis������������������������������������������������������������������������������������������ 247 12.42.1 Diagnostic Question������������������������������������������������������������ 247 12.42.2 Etiological Question������������������������������������������������������������ 247 12.42.3 Prognostic Question������������������������������������������������������������ 248 12.43 Alternatives for the Reference Test, If It Cannot Be Carried Out on All Patients���������������������������������������������������������������������������� 248 12.44 Sensitivity/Specificity Exercise�������������������������������������������������������� 249 12.44.1 Test Questions Sensitivity/Specificity �������������������������������� 249 12.44.2 What Do the Sensitivity and Specificity Mean in This Case? Choose an Answer from A, B, C, or D �������� 249 12.45 Exercise with Positive and Negative Predictive Values�������������������� 249 12.46 Schematic Overview of the Cohort Study of Li et al������������������������ 250 12.47 Schematic Overview Case-Control Study���������������������������������������� 251 12.48 All Inclusion and Exclusion Criteria from Minneboo et al�������������� 252 12.49 Explanation of the Concept ‘Allocation Concealment’�������������������� 252 12.50 Example of a Flow Diagram ������������������������������������������������������������ 253 12.51 Explanation of the Exact Meaning of RR = 1.43 (1.14–1.78) from Minneboo et al�������������������������������������������������������������������������� 255 12.52 Elaboration of Found Relative Risk in Terms of Percentages���������� 255 12.53 Explanation of Thresholds and Significance Level�������������������������� 256 12.53.1 Some Possible Situations���������������������������������������������������� 256 12.53.2 After Statistical Testing, The Following Situations Can Arise������������������������������������������������������������������������������������ 257 12.54 Example of the Value of an SR Over Several RCTs ������������������������ 258 12.55 Example of Inclusion and Exclusion Criteria in an SR by Machado et al������������������������������������������������������������������������������ 259 12.56 Example of Response Bias���������������������������������������������������������������� 260 12.57 Overview of Factors Weakening and Strengthening the Strength of Evidence of a Systematic Review���������������������������� 260 12.57.1 Factors Weakening the Strength of Evidence���������������������� 260 12.57.2 Factors Strengthening the Strength of Evidence ���������������� 261 12.58 More Examples of Qualitative Research Questions ������������������������ 261 12.59 Different Ways of Participating and Observing in Qualitative Research�������������������������������������������������������������������������������������������� 262

xxiv

Contents

12.60 Different Forms of Triangulation������������������������������������������������������ 262 12.61 More Detailed Explanation of the Measures to Enhance Credibility ���������������������������������������������������������������������������������������� 263 12.62 Checklists������������������������������������������������������������������������������������������ 264 12.62.1 Checklist 1 Checklist for Assessing Studies on Diagnostic Accuracy������������������������������������������������������ 264 12.63 Checklist 2���������������������������������������������������������������������������������������� 265 12.63.1 Checklist for Assessment of Cohort Studies into Etiological Factors�������������������������������������������������������������� 265 12.64 Checklist 3���������������������������������������������������������������������������������������� 266 12.64.1 Checklist for Assessment of Case-Control Studies ������������ 266 12.65 Checklist 4���������������������������������������������������������������������������������������� 267 12.65.1 Checklist for Assessment of Prognostic Research�������������� 267 12.66 Checklist 5���������������������������������������������������������������������������������������� 267 12.66.1 Checklist for Assessment of RCTs�������������������������������������� 267 12.67 Checklist 6���������������������������������������������������������������������������������������� 269 12.67.1 Checklist Generalisability, Usability and Appropriateness of the Experimental Intervention ������ 269 12.68 Checklist 7���������������������������������������������������������������������������������������� 271 12.68.1 Checklist for Assessment of a Systematic Review of RCTs ������������������������������������������������������������������������������ 271 12.69 Checklist 8���������������������������������������������������������������������������������������� 273 12.69.1 Checklist for Assessment of Qualitative Research�������������� 273 12.70 Checklist 9���������������������������������������������������������������������������������������� 274 12.70.1 Checklist of Frequently Used Measures of Effect�������������� 274 12.71 Process Worksheets�������������������������������������������������������������������������� 279 12.71.1 Process Worksheet Diagnostic Reasoning�������������������������� 279 12.71.2 Process Worksheet Etiological Reasoning�������������������������� 279 12.71.3 Process Worksheet Prognostic Reasoning �������������������������� 280 12.71.4 Process Worksheet Therapeutic Reasoning ������������������������ 281 12.72 Answers to Questions ���������������������������������������������������������������������� 281 12.72.1 Cognitive Response Test About the Use of Thinking Systems�������������������������������������������������������������������������������� 281 12.72.2 Exercise Sensitivity/Specificity ������������������������������������������ 282 12.72.3 Exercise with Positive and Negative Predictive Values ������ 282 Glossary ������������������������������������������������������������������������������������������������������ 282 References���������������������������������������������������������������������������������������������������� 283 13 Education  About Clinical Reasoning and Evidence-Based Practice���� 287 13.1 Explanation of the 4C/ID Model������������������������������������������������������ 287 13.2 Practical Tools for the Organisation of Education���������������������������� 287 13.3 Process Worksheets�������������������������������������������������������������������������� 288 13.4 Test Material ������������������������������������������������������������������������������������ 288 13.4.1 Complexity Level 2 ������������������������������������������������������������ 288 13.4.2 Complexity Level 4 ������������������������������������������������������������ 288 Index�������������������������������������������������������������������������������������������������������������������� 289

Authors and Contributors

About the Authors Jos Dobber, PhD  is a mental health nurse, an educational scientist and health scientist. He is a senior lecturer in the bachelor’s and the master’s programme in Nursing at the Amsterdam University of Applied Sciences (The Netherlands), with a special focus on the content and consistency of the curriculum. He is an expert on Motivational Interviewing, and on the learning and didactics of clinical reasoning and evidence-based practice. José  Harmsen, MSc  is a general healthcare nurse and nursing scientist. She is lecturer at the Bachelor Education in Nursing (Amsterdam University of Applied Sciences). Her core task as a teacher is clinical reasoning. In addition, she provides workshops in the country on clinical reasoning to graduate nurses. Margriet van Iersel, PhD  is a mental health nurse, educational scientist and programme leader of a Master in Mental Health Nursing. She is senior lecturer at the Bachelor and Master Education in Nursing (Amsterdam University of Applied Sciences). She has a special interest for curriculum design, specifically the 4C/ID model, and she brings in her expertise into the Bachelor and Master nursing curriculum.The authors of this publication are lecturers from the bachelor’s programme in Nursing of the Amsterdam University of Applied Sciences. The contents of this publication reflect the knowledge that has been developed in this field within the Amsterdam University of Applied Sciences.

Contributors 1 W.  J.  Assendelft, MD, PhD  Department of Primary Care Medicine, Radboud University Medical Center, Nijmegen, The Netherlands A. M. Eskes, PhD  Amsterdam UMC, Amsterdam, The Netherlands

 The following authors contributed to Part II Evidence-based practice (EBP).

1

xxv

xxvi

Authors and Contributors

C.  H.  M.  Latour, PhD  School of Nursing, Amsterdam University of Applied Sciences, Integration of Psychiatric and Somatic Care, Amsterdam, The Netherlands Faculty of Health, Centre of Expertise Urban Vitality, Amsterdam University of Applied Sciences, Amsterdam, The Netherlands M. Offringa, MD, PhD  Child Health Evaluative Sciences CHES, The Hospital for Sick Children Research Institute, University of Toronto, Toronto, ON, Canada R. J. P. M. Scholten, MD, PhD  Cochrane Netherlands and Julius Centre for Health Sciences and Primary Care, University Medical Centre Utrecht, Utrecht, The Netherlands

Part I Clinical Reasoning

1

Background Knowledge in Clinical Reasoning

1.1 What Is Clinical Reasoning? When I was just starting my studies, Naomi, a fellow student doing a traineeship in the paediatric ward, told me that the mother of a baby had asked her that day whether it was safe to put her baby on his tummy in bed. Naomi did not know the answer. The nurse supervising her that day told Naomi to look up the answer. I asked Naomi if it bothered her that her supervisor did not just give the answer. But Naomi actually liked that. ‘At least this way I learn clinical reasoning’ she said, ‘all day long you are faced with questions and have to make decisions. In this way, I learn to make my own decisions about patient care’.

Nurses are constantly making decisions during their work. That decision may be because a patient asks a question, ‘Is it safe to put my baby on his tummy in bed?’ In order to answer that question, the nurse must first assess whether or not it is safe, or if it is not known whether it is safe. Another example is assessing the health condition of patients: is a particular patient’s blood pressure ‘just too high’ or is it ‘dangerously high’? The answers nurses give to these questions are therefore conclusions. The nurse uses these to make decisions: ‘The blood pressure is dangerously high, I will inform the doctor about this’. Decisions are thus made when the nurse asks herself—sometimes almost unconsciously—a question or when the patient asks the nurse a question. The questions that nurses ask themselves or receive from patients can be divided into four types of basic questions: 1. What may be the matter with the patient? (diagnostic question) 2. What may have caused this health problem? (causal question or etiological question)

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 J. Dobber et al., Clinical Reasoning and Evidence-Based Practice, https://doi.org/10.1007/978-3-031-27069-7_1

3

4

1  Background Knowledge in Clinical Reasoning

3. How is this likely to end? What do we think we can achieve? (prognostic question) 4. What do we think we can do about it? (therapeutic question) (Sects. 12.1 and 12.2) We can describe clinical reasoning as ‘the continuous process of critical thinking, data collection and analysis, focusing on the questions and problems of an individual and their loved ones, in relation to illness and health, in order to arrive at the best decision regarding the care of this (individual) patient’ (see also [1]). The word ‘clinical’ refers to the provision of care to patients, and the word ‘reasoning’ refers to critical thinking [2]. Clinical reasoning is therefore about devising and designing of nursing care. Because, in many cases, often several decisions are possible, it may not immediately be clear which decision is best for the individual patient. Therefore, clinical reasoning is an important nursing skill. Clinical reasoning ensures that, given the circumstances, the patient receives the best nursing care possible. The more skilled the nurse is at clinical reasoning, the better the care provided. However, good clinical reasoning is no guarantee of preventing errors in care delivery. After all, errors can also be made in the delivery of care. Clinical reasoning is about thinking and making decisions and therefore about dealing with information and asking the right questions. For this reason, clinical reasoning competence is about using a particular way of thinking and acting, with the nurse recognising different types of situations. If the situation includes the basic question, ‘What is going on?’, the nurse should ask herself the diagnostic questions appropriate to this basic question. In the situation of ‘What can we do about it?’, she should ask herself and answer the therapeutic questions appropriate to that situation. Clinical reasoning therefore requires the nurse to think and to come to conclusions in a particular way (Sect. 12.3).

1.2 How to Use Clinical Reasoning? It is not difficult to recognise the different basic questions. Still, it seems rather tiring to me to be constantly reasoning. Some situations are so obvious after all!

Clinical reasoning is a way of thinking through which nurses constantly reason. Each basic question is accompanied by a set of standard questions which the nurse asks herself. These standard questions are discussed in Sect. 1.5. Through practice, the nursing students learn to ask these questions automatically. It has become a mental structure (Sect. 12.4). Many of the decisions made by nurses relate to routine situations they encounter daily in their work or to straightforward and non-complex situations. Experienced nurses can assess these situations quickly and relatively easily and make appropriate decisions immediately. Clinical reasoning has limited use for these situations. The

1.3  Trusting Relationship

5

nurse can make a good decision by just answering a few simple questions in her mind. In these situations, limited analysis is sufficient to reach a good decision. Situations that are complex or new, or which the nurse experiences too little to make them routine, require significantly more of her thinking. In these situations, the risk of making a wrong or suboptimal decision is high. In non-routine situations, clinical reasoning is heavily relied upon, and therefore, a much more analytical approach is required. It is therefore important for nurses to assess the patient’s situation and timely and adequately identify when what appeared to be a routine situation becomes more complex. As the situation then places a greater demand on clinical reasoning, the nurse should switch to a higher analytical strategy.

1.3 Trusting Relationship Sometimes it seems like patients discuss very personal matters with nurses. At least, I’ve noticed that there are nurses who have such a good relationship with patients that this happens. I wonder how those nurses manage to do that. The other day I was working with Aisha, who is one of those nurses. One of the patients was a gentleman with dementia. Earlier in the week, I had noticed that some of the nurses found it difficult to care for him. Aisha had not nursed this patient before. I noticed that she took time with him, and she didn’t start acting right away. I don’t know how she did it, but she really connected with him.

Good nursing care starts with establishing a trusting relationship. The nurse makes an effort to understand the patient and his perspective and to connect with them. Listening to the patient is important, along with mutual respect, trust and compassion: a commitment to the health and well-being of the patient [3]. A trusting relationship means that the patient and the nurse both feel and know that the nurse is thinking and acting in the patient’s best interests. This relationship is particularly important in the case of extra vulnerable patients; for example, in people with intellectual disabilities and in people with dementia. In clinical reasoning, this is an essential aspect. It is not just about medical-­ biological and psychosocial problems. Not only nurses want to understand the condition, but they also need to understand the patient. Thinking about whether the relationship with the patient is good and whether his perspective is sufficiently understood by the nurse, and whether this is reflected in nursing care, is part of the essence of nursing. So, clinical reasoning always goes hand in hand with ‘seeing’ the patient; one cannot exist without the other. And with working on the relationship with the patient and adapting to his perspective, it is also about the question ‘What is going on?’ So, it is about both. The nurse enters into dialogue with the patient. Only when there is sufficient trust between the two, the patient will really tell what is on his mind.

6

1  Background Knowledge in Clinical Reasoning

1.4 The ICF as a Thought Model in Clinical Reasoning All those patient details … sometimes my head reels. I just write them down, but I can’t get an overview. How can all these data be linked together? What belongs together and why? How do I gain insight into what the consequences of the illness are for the patient? How can I bring order to this mass of data?

Mrs Fletcher Mrs Fletcher (62 years old) was hospitalised with a broken hip after a fall. Her husband quickly brought her to the hospital. There she was operated and received a hip prosthesis. After the operation, the nurse takes care of the operation wound, observes this wound and monitors Mrs Fletcher’s pain. Because of the fall and the operation, Mrs Fletcher is not mobile anymore. At this moment, she cannot take care of herself anymore. Mrs Fletcher asks several nurses if she will recover and become fully mobile again. She doesn’t want to miss the regular babysitting days with her grandchildren. But will she be able to do this when she is not fully mobile anymore? Sometimes, when she thinks about this, she starts to cry. Her husband and her daughter tell her that it won’t be as bad as all that. This does her good, but when the visitors have left, Mrs Fletcher asks the nurse ‘if it will really be alright’. ‘I will do everything I can to recover,’ she adds. She tells the nurse how glad she is to see the physiotherapist already coming to do exercises with her. In clinical reasoning, the nurse tries to gain insight into the ‘total patient’. She tries to understand how (the treatment of) the broken hip affects Mrs Fletcher’s life at the moment and possibly in the future. The nurse discusses Mrs Fletcher’s condition and treatment with the orthopaedic surgeon and the physiotherapist. When necessary or possible, the nurse supports the treatment of other disciplines by observing and participating in their treatment.

1.4.1 What Is the ICF Diagram? What Do the ICF Terms Mean? The International Classification of Functioning, Disability and Health [4] helps to understand what problems Mrs Fletcher might have physically (surgical wound, prosthesis, pain, mobility), psychologically (anxiety, grief, meaning) and socially (babysitting, relationships with children and grandchildren) and how these problems are interrelated. The ICF is a classification of human functioning in the context of health, developed by the World Health Organization (WHO). In clinical reasoning, the ICF diagram is used to systematically map out the consequences: 1. of the disease; 2. of the corresponding treatment for the patient’s functioning; 3. of the patient’s lifestyle.

7

1.4  The ICF as a Thought Model in Clinical Reasoning

diseases

body functions and body structures

activities

environmental factors

participation

personal factors

Fig. 1.1  Schematic representation of the ICF model

The ICF diagram helps to think about the problems in the context of health, the causes of those problems, their consequences, how they are expressed and their meaning for the patient. At the top of the diagram (see Fig. 1.1) is the ‘health condition (disorder or disease)’. This is not part of the ICF, but when reasoning about people’s state of health, their health condition or disease is obviously an important factor. The ICF itself comprises the other five components. All components are connected by arrows. The arrows indicate that each component can influence all the others. Although in the diagram, the component ‘personal factors’ is not linked by an arrow to ‘health condition (disorder or disease)’, age (a personal factor), for example, can have a major influence on the course and prognosis of diseases. Table 1.1 contains the definitions given by the WHO to the components of the ICF. Because the ICF is based on the healthy functioning of people, the terms on the left-hand side of the table express health. When reasoning about the health status of patients, the health disruption plays a major role (right-hand side of the table). Using the ICF for clinical reasoning, we therefore work with pairs of terms in which both the healthy and the unhealthy aspects of functioning are addressed. A joint is a body structure with a number of properties that make movement possible. Movement or mobility is the body function of the body structure of the joint. Body functions are about what these body structures do: why humans have these body structures. Often, body structures have several functions at the same time. For example, the skin is a body structure with a number of properties that have both protection and sensation (e.g. itching) as a function (Sect. 12.5). In order to perform activities, well-functioning body structures and body functions are required. This is obvious for the body structures that are necessary for movement functions. In case of impairments in the bones, joints, muscles, ligaments or connective tissues, there are limitations in movements such as lifting, walking and showering, or these activities may not be possible at all. Conversely, activities are needed to maintain or strengthen body structures. Walking develops leg muscles, while the muscles that are no longer used will weaken in a short time: muscular atrophy.

8

1  Background Knowledge in Clinical Reasoning

Table 1.1  Definitions of the ICF components Body functions Body functions are physiological functions of body systems (including psychological functions). Examples: consciousness, sleep, smell function, maintenance appropriate body weight, muscle power Body structures Body functions are anatomical parts of the body such as organs, limbs and their components. Examples: anatomical structure of the retina, of the heart, of the hip joint Activities Activities are the executions of tasks or actions by an individual. Examples: focusing attention, solving problems, moving around, washing oneself, drinking, doing housework Participation Participation means the involvement in a life situation. Examples: remunerative or non-remunerative employment, socialising, recreation and leisure

Impairments Impairments are problems in body functions such as a significant deviation or loss. Examples: impairment in pain perception, impairment in absorption of food Impairments Impairments are also problems in body structures such as a significant deviation or loss. Examples: intestinal perforation, atrophy of muscle tissue, opacity of the lens of the eye Activity limitations Activity limitations are difficulties an individual may have in executing activities. Examples: limitation in handling stress, in reading, in walking, in washing oneself Participation restrictions Participation restrictions are problems an individual may experience in involvement in life situations. Examples: restrictions in entering into informal social relationships, using money to purchase goods or services, engaging in informally or formally organised sports events

Environmental factors Environmental factors are he physical, social and attitudinal environment in which people live and conduct their lives. Examples: medicines, glasses, educational toys, money, climate, family, social norms, housing, presence or absence of health care facilities Personal factors Personal factors are a person’s individual background. Examples: age, gender, social status, life events The definitions are taken from WHO [4, 5]

Participation can be influenced by the health condition in several ways. Limitations in activities, for example, in being able to eat and drink independently, complicate taking part in dinners with friends and going out for dinner. Someone with a severe skin disorder on his face may have problems with social relationships and may want to go outside as little as possible, as many people stare at him. In this way, limitations in activities and impairments in body structures may restrict participation. Activities and participation are sometimes difficult to distinguish. Putting a photo on ‘Instagram’ and ‘moving around’ are activities, but they can also be considered as participation. The WHO solves this by stating that the categories

1.4  The ICF as a Thought Model in Clinical Reasoning

9

Communication, Mobility, Self-care and Domestic Life can be seen as both an activity and an item of participation [4, 5]. Environmental factors are factors that are external to the patient but that influence his/her functioning in the context of health. That influence can be positive (facilitators) or negative (barriers/hindrances). In the Western world, someone who suffers a myocardial infarction can be taken by ambulance to a hospital where they will receive specialist care. After proper initiation of medication and advice on a healthy lifestyle, the patient will go home. He is monitored and receives cardiac rehabilitation to improve his physical condition. With support, the patient tries to get a grip on his life again. If the same person suffers a myocardial infarction while on holiday in the steppes of Mongolia, these supportive environmental factors are absent or virtually non-existent, making the prognosis for the patient significantly worse (Sect. 12.6). The final component of the ICF is the personal factors. These are characteristics and properties of the patient himself. In addition to characteristics such as age and gender, these include beliefs (e.g. vegetarianism) and opinions on health and disease. In contrast to the other components of the ICF, the personal factors are not further elaborated in the classification itself. In clinical reasoning, the ICF diagram can be used to reason on health problems, their causes, consequences, expressions and meaning for the patient. This would look like the following for Mrs Fletcher. In Mrs Fletcher’s case, the new prosthesis and the surgical wound are impairments in body structures. Pain and the fear of no longer being able to look after her grandchildren in the future are body functions, but as long as the pain and the fear fit the situation (pain is normal after such an operation; her fear is real and does not dominate her thinking and mood), these functions are not impaired, but they do require nursing care. In addition, the body functions mobility and stability of the hip joint are currently disturbed. Temporary immobility and limitation of self-care are limitations of activities for Mrs Fletcher. For the former, Mrs Fletcher is treated by the physiotherapist. The nurse supports this treatment where possible. The care of the operation wound is also an activity. As long as the condition of this wound requires specialist knowledge and skills, these activities are taken over by a specialist (the nurse). Although unlikely, Mrs Fletcher’s recovery could be complicated and lead to permanent limitations in her mobility. In that case, she may no longer be able to look after her grandchildren. A serious restriction in participation may then arise. Normally, recovery is not a problem at Mrs Fletcher’s age. Because she understands that her own efforts play an important role in her recovery, age and views on health and illness are positive personal factors, facilitating the recovery. The support of her family, the availability of good hospital care, physiotherapy, walking aids and (after her discharge from hospital) district nursing are supportive environmental factors (facilitators).

10

1  Background Knowledge in Clinical Reasoning

1.4.2 How Does the ICF Support Structuring and Reasoning? 1.4.2.1 Diagnostic and Etiological Reasoning When asked what may be the matter with Mrs Fletcher, the nurse explains that there are five main problems. Four of them are physical. All four are caused by the impairment in the body structure of the hip joint and the subsequent operation in which a prosthesis was placed. They are (1) the surgical wound, (2) pain, (3) temporary immobility and (4) self-care limitation. These problems are part of the standard care in such surgery. There is also a psychological problem, namely (5) the fear of not being able to babysit the grandchildren anymore. The nurse discusses the problems with Mrs Fletcher, including the consequences of each problem for her functioning. During the discussion, it became clear that Mrs. Fletcher was most worried about babysitting her grandchildren. The risk that she will remain immobile and therefore unable to babysit in the future occupies Mrs Fletcher all day long. It makes her anxious and sad. The bond she has with her grandchildren is the most important thing in her life at the moment. Based on this, the nurse ranked the problems in order of importance on a scale of 1–10 (see Table 1.2). The participation problem that might arise is at the forefront for Mrs Fletcher, even though she is in the hospital for an impairment in a body structure. Further reasoning about this problem leads the nurse to a problem formulation as shown in Table 1.3. 1.4.2.2 Prognostic Reasoning Since a standard intervention to restore mobility is available (in the form of exercise; see below under therapeutic reasoning), we can already estimate its effect and include this in the reasoning about the prognosis. Practising walking will probably Table 1.2  Importance of problems and consequences for functioning Highest priority in clinical decision-making and in the occurrence of conflicting choices

10 9 8

Important, but with a lower priority

7 6 5 4

Less important in making clinical decisions or of less importance to patients

3 2 1

Based on Guyatt et al. [6]

• Fear • Wound healing • Pain • Limitation in self-care • Temporary immobility

11

1.4  The ICF as a Thought Model in Clinical Reasoning Table 1.3  Problem formulation Ms Fletcher What may be the matter with the patient? (problem) What may have caused this health problem? (etiology)

What signs and symptoms would indicate to the nurse that a fear of not being able to babysit is indeed a problem? (signs and symptoms)

• Fear of not being able to babysit • Current immobility • Uncertainty regarding recovery • Continued need to discuss her concerns • Anticipatory grief

also reduce Mrs Fletcher’s anxiety, because she will notice that her mobility is recovering. Earlier, the nurse had already determined that Mrs Fletcher is young enough to make a successful recovery. This is reinforced by the fact that Mrs Fletcher understands the importance of exercise and is fully committed to it. The support of her family, the availability of good hospital care, physiotherapy, walking aids and community nursing (after her discharge from hospital) are facilitating environmental factors. Thus, the environmental factors and personal factors promote recovery. All these factors are positive. Therefore, the nurse expects Mrs Fletcher to make a full recovery. Because Mrs Fletcher will soon notice that she is improving, the nurse thinks that Mrs Fletcher’s anxiety will soon decrease and eventually disappear.

1.4.2.3 Therapeutic Reasoning As a result of the impairment, Mrs Fletcher is temporarily unable to perform a number of activities (mobility, self-care). The (partial) taking over of these activities by the nurse is also part of the standard care. For Mrs Fletcher, most of the problems belong to the ICF component body structures and body functions. The operation wound belongs to body structures and is actually a consequence of the medical treatment. The pain, anxiety and immobility and stability of the hip joint are body functions. In the ICF component activities, the immobility problem and the self-care limitation are expressed. Mrs Fletcher can no longer walk properly and cannot perform part of her self-care. Furthermore, the activities are an important source for interventions, since exercising restores the immobility and stability of the hip joint. Self-­ care and specialist wound care are (partially) taken over by the nurse. The current immobility is the main cause of Mrs Fletcher’s anxiety. For this cause, the activities (in this case practising walking) are the focus of the intervention. The meaning of the broken hip, and of the treatment and its consequences, is most strongly expressed by Mrs Fletcher in the ICF component participation. She would be very upset if she were no longer able to look after her grandchildren. But in many other cases, the significance for the patient is also expressed in the ICF component activities. It is obvious that not being able to (properly) perform activities such as concentrating, carrying out daily routine actions, using the Internet, washing oneself or preparing meals has great significance for the person’s life.

12

1  Background Knowledge in Clinical Reasoning

1.5 Analysing and Answering the Four Basic Questions of Clinical Reasoning ‘What may be the matter with the patient?’ is a simple question. But to give a well-­ argued answer, you have to dig deeper. It requires more professional knowledge and careful observation of the patient. In short, behind that simple question is a thought structure of questions!

In complex situations, decisions regarding diagnosis, etiology, therapy/interventions and prognosis must be made thoughtfully. As nurses progress through a set series of questions, they develop a thought structure that allows these informed decisions to be made. Each of the four types of basic questions is therefore associated with a set of standard questions. These standard questions are general. The nurse can ask these questions for each patient. In addition, the nurse can supplement the questions with specific questions. These are questions that the nurse asks based on her professional knowledge of the patient’s condition. Below are the standard questions per basic question.

1.5.1 What Are the Standard Questions in Diagnostic Reasoning? The standard questions in diagnostic reasoning focus on interpreting the symptoms. They are intended to determine what health problems are present. Ultimately, the standard questions lead to an informed answer to the question: ‘What may be the matter with the patient?’. 1. What objective and subjective symptoms are present in the patient (complaints, signs, behaviours, abnormal value compared to the reference range)? 2. How can these symptoms be explained (which abnormalities in anatomy, physiology and psychosocial functioning)? 3. What else would you like to know from or about the patient? Use the answer to this question to supplement the answers to Questions 1 and 2. 4. What is the patient’s hindrance (possibly expressed in the ICF components activities, participation)? 5. How does the patient react to this (emotional behaviour, effective or ineffective self-management)? 6. Which possible health problems does this indicate (from which impairments in body structures, in body functions and limitations in activities and restrictions in participation)? Take 5 min to complete the problem list. For each problem, indicate whether it is an important or a minor problem.

1.5  Analysing and Answering the Four Basic Questions of Clinical Reasoning

13

7. What are the defining characteristics of the major health problems? 8. How can the presence or absence of these characteristics be tested? Then test whether these phenomena are actually present (Sect. 12.7). After reviewing the defining characteristics, the nurse has an overview of the health problems.

1.5.2 What Are the Standard Questions in Etiological Reasoning? In etiological reasoning, the standard questions focus on determining the causal factors and the related factors. Causal factors are the causes of the problem, while related factors may not have caused the problem, but they do (partly) maintain the problem. Ultimately, these standard questions lead to a deliberate answer to the question: ‘What may have caused the health problem?’ 1. Are there possible causes and related factors from diseases and disease processes? Which ones? 2. Are there possible causes and related factors from the (para)medical treatment? Which ones? 3. Are there possible causes and related factors from body structures? Which ones? 4. Are there possible causes and related factors from body functions? Which ones? 5. Are there possible causes and related factors from activities? Which ones? 6. Are there possible causes and related factors from participation? Which ones? 7. Are there possible causes and related factors from environmental factors? Which ones? 8. Are there possible causes and related factors from personal factors? Which ones? 9. Can these factors together cause or perpetuate this health problem? 10. Are these factors present in the patient? 11. Do these factors have this effect in this patient? (Sect. 12.8) Using the answers to these questions, the nurse can reason about the major causes and related factors in the health problem. Once these are identified, the nurse can use these causes and factors in their prognostic and therapeutic reasoning.

1.5.3 What Are the Standard Questions in Prognostic Reasoning? In prognostic reasoning, the standard questions focus on the prognosis: how will the health problem progress? Using the standard questions, the nurse considers the prognostic factors that influence the course of the health problem and what

14

1  Background Knowledge in Clinical Reasoning

maximum result can be expected. Using the standard questions, the nurse comes to an informed decision on the patient’s care goal for a specific health problem and thus answers the question, ‘What do we think we can achieve?’. 1. Which physical, psychological and personal factors (including causal and related factors) influence the course of the health problem negatively, and which positively? 2. Which environmental factors influence the course of the health problem negatively, and which positively? 3. On what mechanism of action (per factor) is that influence based? 4. All those factors taken together, is their influence positive, neutral or negative? 5. What is the maximum achievable result for this health problem? (Sect. 12.9) The answers to these questions lead to the reasoning and formulation of a care goal for the patient’s health problem.

1.5.4 What Are the Standard Questions in Therapeutic Reasoning? The standard questions in therapeutic reasoning focus on the expected effect of the interventions and on the suitability of the interventions. Answering the questions leads to an informed choice of one or more interventions and to an answer to the question, ‘What can we do about it?’. 1. Which interventions are available? 2. What effects (with respect to the goal) can be expected from these interventions? How large is that effect expected to be? 3. What evidence is there that this intervention does indeed have that effect? 4. What is the focus of intervention (is it the causes, related factors, symptomatology, prognostic factors or risk factors)? 5. What is the mechanism of change of the intervention? 6. Is the intervention feasible (resources, practicability)? 7. Is the intervention acceptable to the patient? (Sect. 12.10) Based on the answers to these questions, the nurse reasons about the most appropriate intervention(s) for this patient. Please note: To answer Questions 2 and 3 of the standard questions on therapeutic reasoning, adequate knowledge of critical appraisal of intervention research and of guidelines is required. For this, see Part II of this book.

1.5  Analysing and Answering the Four Basic Questions of Clinical Reasoning

15

DATA

Symptoms (S)

Diagnostic reasoning using the standard questions Problem (P)

Problem (P)

Etiology (E)

Risk factors (R)

Etiological reasoning using the standard questions

Prognostic reasoning using the standard questions

Prognosis

Prognosis

Intervention(s)

Intervention(s)

Therapeutic reasoning using the standard questions

Fig. 1.2  Schematic representation of the process of clinical reasoning

1.5.5 What Is the Course of the Clinical Reasoning Process? In summary, the process of clinical reasoning can be seen as follows (see Fig. 1.2). From the data about a patient, the nurse distinguishes the important, meaningful data from the data that are unimportant or of little value to the patient’s health condition. The important data are ‘a signal’ that something is wrong, while the unimportant data are a kind of ‘noise’ that sounds through the signal. Key is to distinguish signal from noise. From the important data, the nurse recognises the symptoms (represented by the abbreviation S). These symptoms are explained as much as possible by pathophysiology or deviant psychosocial processes, and on this basis, they are sorted into different health problems, represented by the abbreviation P.  The problems can be either present or current problems or problems that do not yet exist,

16

1  Background Knowledge in Clinical Reasoning

but for which the patient is at increased risk of developing this health problem. A health problem that does not yet exist, but for which the patient is at increased risk, is called a high-risk diagnosis (see Sect. 2.12). The standard questions for diagnostic reasoning help to formulate the health problems and associated symptoms. Once the P and S have been established, the nurse continues to reason about what is causing the problem. This is the etiological reasoning. Etiology (represented by the abbreviation E) involves not only causes, but also related factors: these are factors that did not cause the problem, but that contribute to maintaining the problem. In the case of high-risk diagnoses, there is not yet a problem and therefore no cause for a problem. However, the nurse has determined that the patient is at increased risk of a particular problem. This is based on the presence of risk factors (represented by the abbreviation R). For example, immobility is a risk factor for pressure ulcers (see Sect. 3.4). The standard questions on etiological reasoning can be used to help determine the correct etiological factors. With this, the nurse has identified the health problem in the form of a ‘PES’ (Problem-Etiology-Signs and Symptoms) or a ‘PR’ (Problem-Risk Factors). The next step in the process is to answer the question of how the health problem is likely to develop: the prognosis. The nurse finds out what factors influence this course, the probable magnitude of their influence and whether they influence the problem positively or negatively. Using the standard questions on prognostic reasoning, the nurse can systematically map out the plausible prognostic factors. Based on this inventory, the nurse will reason about the most likely outcome and determine the best possible outcome for the patient. The next question to answer is if and how (the course of) the health problem can be influenced by the nurse herself. Which intervention(s) can be expected to have the largest effect on this patient? The standard questions on therapeutic reasoning lead the nurse to a well-considered choice of an intervention.

1.6 The ICN Definition of Nursing as a Framework for Nursing Decisions What I find difficult is determining what I can decide for myself. Where I can act independently without consultation. And what has to go through the doctor or the psychologist or through someone else. Some colleagues say if it has to go through another discipline, ‘That’s not nursing. That’s not our business’. But when I look at what I do all day when I work, and what I reason about, I do quite a bit of work that those colleagues consider ‘non-­ nursing’. But still, my colleagues and I, we do perform those tasks, and for that reason, I actually think it is ‘nursing’. I’d like to know how that is.

The definition of nursing, as defined by the International Council of Nurses (ICN) in 2002 [7], is:

1.6  The ICN Definition of Nursing as a Framework for Nursing Decisions

17

Nursing encompasses autonomous and collaborative care of individuals of all ages, families, groups and communities, sick or well and in all settings. Nursing includes the ­promotion of health, prevention of illness, and the care of ill, disabled and dying people. Advocacy, promotion of a safe environment, research, participation in shaping health policy and in patient and health systems management, and education are also key nursing roles.

To understand this, it is important to distinguish between autonomous and collaborative (or participatory) nursing practice.

1.6.1 What Is Meant by Autonomous Action and What by Collaborative (or Participatory) Action? When the nurse acts autonomously, this means they make decisions and carry out actions for which nurses are primarily responsible. In autonomous practice, the nurse has this autonomy, since these actions fit in the nursing professional domain. In addition, the nurses’ professional knowledge and skills are sufficient in the area of these decisions and actions. Other healthcare professionals have less professional knowledge and skills in these areas. If the nurse participates in the medical treatment (by the doctor) or paramedical treatment (e.g. by the physiotherapist or dietician), another professional is primarily responsible for the treatment. For example, the doctor is responsible for the medication therapy. The nurse is delegated the administration of medication, and she participates in the medical treatment. This is referred to as collaborative or participatory care: the nurse participates in treatment for which another professional is primarily responsible. Many disciplines work together in patient care, and it is not always possible to establish clear boundaries between each person’s area of work and responsibility. There is an intensive relationship not only between nursing and medicine but also between nursing and the paramedic healthcare professionals. Whereas nursing is generalist, paramedics are more specialised in a more limited number of health problems. Nurses frequently participate in the treatment of medical and paramedic care, as complex care requires a multidisciplinary approach. In complex care, some problems can be that complex that they are ‘multidisciplinary health problems’. With these problems, various disciplines are concerned and have a share in monitoring, decision-making and/or intervening. An example of multidisciplinary health problem is medication noncompliance. Initially, the physician and the patient together decide on medication treatment, and the physician provides the patient with relevant information on that specific drug. The medication administration is delegated to the nurse. The nurse also monitors the medication compliance, the effects and the possible occurrence of side effects. Up to this point, the nurse provides collaborative (or participatory) care. However, when noncompliance appears, both disciplines are responsible for (deciding on) follow-up actions, such as finding out the patient’s motives for noncompliance and responding to these and/or strengthening the patient’s motivation for medication treatment.

18

1  Background Knowledge in Clinical Reasoning

The extent to which, and depth with which, nurses reason beyond the boundaries of their own professional nursing domain is dependent on a number of factors. Firstly, the nurse’s knowledge and reasoning skills are important. In addition, the way in which the health professionals involved work together has a significant influence. For example, in multi-professional collaboration, professionals are hardly expected to reason beyond the boundaries of their own profession, whereas in inter-­ professional collaboration, this is actually required (Sect. 12.11).

1.6.2 How Does This Division of the Nursing Professional Domain into Autonomous and Collaborative Action Affect Clinical Reasoning? The nursing professional can be derived from the ICN definition of nursing [7]: Nursing encompasses autonomous and collaborative care of individuals of all ages, families, groups and communities, sick or well and in all settings. Nursing includes the promotion of health, prevention of illness, and the care of ill, disabled and dying people. Advocacy, promotion of a safe environment, research, participation in shaping health policy and in patient and health systems management, and education are also key nursing roles.

In the context of this book, we translate this definition in terms of autonomous and collaborative action as follows: there is autonomy of action by nurses in the area of health problems, excluding diagnosis of illness and treatment, but including prevention and self-management. In addition, nurses may act in a collaborative manner by collecting data on disease manifestations and potential complications and communicating these data to the physician or paramedic. Furthermore, nurses may act in a collaborative way within medical and paramedical treatment when a part or aspect of that treatment has been delegated to them (Sect. 12.12). This means that the nurse autonomously addresses health problems in their diagnostic reasoning, excluding diagnosis of illness, but including health risks and determining the self-management of the patient and their caregivers. In addition, participatory diagnostic reasoning by nurses also extends to reasoning about illness symptoms and potential complications (Sect. 12.13). In terms of therapeutic reasoning, this reasoning is autonomously directed towards nursing care, excluding (para)medical treatment, but including prevention of health risks and the promotion of self-management of the patient and his/her carers. From a collaborative approach, the nurse reasons about the treatment by the physician and paramedics, particularly where the nurse has been delegated an intervention or aspects of the treatment. The ICN definition does not specifically address etiological and prognostic reasoning. However, in order to properly reason on nursing interventions, health promotion and prevention, it is essential to reason on risk factors, prognostic factors and (potential) causes of health problems. With respect to prognostic reasoning, nurses must set realistic goals: goals that are best achievable given the patient situation. In many cases, goals will be aimed at

1.7  Diversity-Sensitive Reasoning

19

self-management, quality of life and aspects of daily functioning. The intended results of preventive action can also be expressed in goals, for example, the prevention of malnutrition.

1.7 Diversity-Sensitive Reasoning I don’t understand. There was a patient admitted who was given many different medicines. Together with my supervisor, I explained to her how she should use them. But this woman came from an Eastern European country, and her English wasn’t good enough to understand our explanation. When, as a try out, she managed the medicines herself for a day, things went wrong. She took them at the wrong times, and some she didn’t take at all. If she does not use the medicines properly, it can go completely wrong with her health, as we had clearly explained to her, hadn’t we? That’s why we wanted to arrange community nursing for her. But the patient doesn’t want that. Really very strange. Her daughter later explained that in the region where her mother comes from, it is important not to depend on strangers. That is why her mother rejects community nursing as help and support. For her, she considers the community nurse as a stranger who comes into her house and on whom she depends. And that is why she will never accept this form of care. I hadn’t thought of it that way. People can really think very differently from what I would expect, and that means I have to provide care in a different way so it fits in with them.

Good clinical reasoning leads to patient-centred nursing care: tailored to the individual patient. This is necessary because there are major differences between patients, for example, in their experience of illness and acceptance of care, even if they have the same condition. What may be the right decision for one patient may not be right for another. After all, the other patient is not the same person, and backgrounds, needs and ways of thinking differ greatly between patients: the patient group is diverse.

1.7.1 What Areas of Diversity Support Diversity-Sensitive Reasoning? Patients can differ from each other in many areas. There are many dimensions of diversity, the areas of diversity listed below represent the most important areas. For example, Isabella’s patient has a different view on healthcare than is usual in the land in which she now lives, based on her culture and philosophy of life. The diversity areas contribute to reasoning about what is important to the patient in the area of care provision and in his or her choices about their own health [8, 9]. The areas below are derived from those formulated by the Caleidoscopia Network [10].

20

1  Background Knowledge in Clinical Reasoning

1. ethnicity, nationality, region 2. life stage and generation 3. religion and philosophy of life 4. sex and gender 5. sexual orientation 6. social class, socio-economic status (SES) 7. talents, limitations, personality (Sect. 12.14) The content of the diversity areas is not fixed facts as people change over time. Experiences influence people’s thinking and can shape people further. Another pitfall is to think that there is a fixed hierarchy in the seven areas, for example, that ‘ethnicity, nationality, region’ is the most important [11]. Instead, it depends very much on the situation which aspects exert the most influence. The seven areas influence each other and thus play a role in choices and perceptions of the health situation. These areas are not the sum of their parts (ethnicity, nationality, region + class/SES + stage of life and generation …) [12], but rather form a mosaic that can be different in each situation [8]. If the nurse has a good grasp of this mosaic, they can incorporate it into their clinical reasoning and thus make their care truly patient-centred.

1.7.2 Why Is Diversity-Sensitive Reasoning Important? Diversity-Sensitive Reasoning helps to match nursing care to the patient. By looking (preferably with the patient) at some diversity issues, the nurse gets to know better the patient’s (and family’s) background, needs, abilities and mindset. And this knowledge increases understanding of the patient’s views on how they view their body, their health and their care. It also helps to better understand the patient’s lifestyle, what care the patient does or does not want and whether the patient thinks they have any influence on their own health or illness. A community nursing team has five patients who all have a serious wound and are receiving community nursing care for it. What differences could you think of between the following five patients [8]? 1. Male, 48 years old. Physiotherapist. Lives in his motherland. Married to a lawyer. Religion: Waldorf Protestant. 2. Female, 28 years old. Refugee. Born in Syria, arrived 1 month ago. Lives alone, in a refugee shelter. Religion: Islam. 3. Male, 35  years old. Production worker. Lives in his motherland. Has a mild mental handicap, lives in a sheltered living facility. Religion: none. 4. Girl, 10 years old. Lives in her motherland. Lives alternately with one of her (divorced) parents. Religion: none. 5. Female, 81  years old. Retired administrative worker. Born in Portugal, lives here since her 25th. Widowed. Religion: Catholic.

1.8  Evidence-Based Practice (EBP)

21

1.7.3 How Does Diversity-Sensitive Reasoning Fit with the ICF? Six of the seven areas are personal factors, and therefore, they can provide deeper content to reasoning about the personal factors. The seventh area, ‘talents, limitations, personality’, refers to body functions: physiological and psychological functions of body systems, and is thus an important component of reasoning about functions [4, 5].

1.8 Evidence-Based Practice (EBP) Before I started my internship, I understood that clinical reasoning about the four basic questions would be important in practice. It’s not just about being proficient in nursing procedures, but also about being able to reason well, so that I can provide good nursing care. But I didn’t understand what science and evidence-based practice had to do with reasoning and nursing practice. How can evidence-based practice influence the way I measure blood pressure?

1.8.1 How Can Evidence-Based Practice Improve Clinical Reasoning and Skill Implementation? Through clinical reasoning, nurses arrive at the best decision regarding the care of the individual patient. An example of this type of decision is taking a patient’s blood pressure to determine if the blood pressure is out of line with what would be expected (diagnosis), and if a deviating value would warrant intervention and if so, what type of intervention. Whether an intervention is warranted depends in part on the causes of the blood pressure deviation. It also depends on the prognosis of the consequences of an untreated blood pressure versus the consequences of a treated blood pressure (Sect. 12.15). Measuring blood pressure is a nursing action that is performed everywhere in healthcare: in community nursing, in the hospital, in mental healthcare, in care for people with a mental disability and in the nursing home. The technique of this skill is not complicated. The height of the blood pressure can vary rapidly. This level, and therefore the measured value, is determined by various physiological and psychological mechanisms. This has consequences for where blood pressure should be measured (which arm), when it should be measured (before or after a meal) and how it should be interpreted (as is clear from Isabella’s example). The reason for measuring blood pressure also plays a role in this type of decision. Hence, measuring the blood pressure raises all sorts of questions. Scientific knowledge is required to develop appropriate answers, which will help the nurse to reason about the measurement of blood pressure and interpret its outcome. Using the method of evidence-based practice (EBP), the nurse will acquire the necessary knowledge and use it in their clinical reasoning.

22

1  Background Knowledge in Clinical Reasoning

1.8.2 What Is Evidence-Based Practice and What Components Does It Consist Of? Evidence-based practice can be described as ‘the integration of best research evidence with clinical expertise and patient values to facilitate clinical decision making’ [13]. Evidence-based practice consists of several components, namely: 1. the current best scientific evidence 2. the preferences, wishes and expectations of the patient 3. the materials and resources available in this healthcare facility for this patient 4. the clinical expertise of the nurse It is important not only to look at the scientific evidence. The patient himself is the starting point of healthcare, and his wishes and preferences carry a lot of weight. If a heart patient needs to change his lifestyle, he may prefer to tackle his obesity by eating healthily and taking more exercise, rather than quitting smoking, or vice versa, or he may do both at the same time. The clinical expertise of the nurse is also mentioned as a component of EBP. In fact, this means that EBP is an ingredient of clinical reasoning, because if the nurse is aware of scientific knowledge and patient preference, she uses it to reason about the best decision for this patient in this situation. In doing so, she weighs the various components: 1. What options are there? How strong is the scientific evidence for the different options? What advantages and disadvantages can be expected for each option? What is the mechanism of change of the different options? 2. What is the patient’s preferred order for his options? On what basis did the patient determine this order? 3. Can all options be carried out by the nursing team? Do we have the knowledge and skills to carry out the options properly? Are the tools we need for the options available or easily purchased? Is there enough time and budget to carry out each of the options? 4. To what extent do the patient characteristics play a role? Can the advantages and disadvantages found in scientific research also be expected for this patient? Is the mechanism of change also to be expected in this patient? In what ways does this patient resemble or differ from the patients who were examined in the scientific study? How can these similarities and differences influence the mechanism of change and the outcomes? The consideration of these questions and their answers is purely clinical reasoning, which draws heavily on the knowledge of the nurse. This knowledge is the

1.8  Evidence-Based Practice (EBP)

23

foundation of the nurse’s clinical expertise, using EBP as an ingredient in their clinical reasoning (see also Isabella’s second example in Sect. 12.16).

1.8.3 What Is the Methods of Evidence-Based Practice? What Kind of Knowledge Does EBP Lead to? The method of evidence-based practice comprises five steps. The fourth step actually consists of clinical reasoning. The five steps are based on Guyatt et al. [14]. 1. Ask: reformulate the identified clinical problem into a structured clinical question 2. Acquire: search and select the best evidence 3. Appraise: weigh the quality of the evidence, and identify the results 4. Apply: determine whether the results found in the evidence can also be expected for your patient 5. Act: decide whether and how to use the evidence (this often includes shared decision-making with the patient) In considering whether the method of EBP is warranted for a clinical question, the nurse should first consider whether it is a background question or a foreground question (see Sect. 1.10.4). The question regarding the physiological mechanisms that determine blood pressure is an example of a question for background knowledge. It is best to look up the answer in a current physiology textbook. It is important that it is a topical physiology textbook, in which recent scientific knowledge about the physiological mechanisms is included. After all, background knowledge also develops through scientific research. A question such as: ‘How can the influence of psychological mechanisms that co-determine blood pressure be limited?’ is a question for foreground knowledge. Answers to this type of question are best sought in scientific journals and in recent evidence-based guidelines. Through the method of evidence-based practice, foreground questions can be answered. This method leads, for example, to evidence about: • • • • • •

tools to make the diagnosis; knowledge of risk factors; knowledge of prognostic factors; knowledge of causes and related factors; knowledge of advantages and disadvantages of different interventions; knowledge of patient perspectives and experiences. Part II of this book covers learning to use the method of evidence-based practice.

24

1  Background Knowledge in Clinical Reasoning

1.9 The Use of Hypotheses in Clinical Reasoning It’s funny. In clinical reasoning, I learned that reasoning can be disturbed by errors, for example, that you interpret the information in such a way that it agrees with what you already thought. How often I have fallen into that trap in practice! Just the other day, I came to a supported living group, consisting of five people with intellectual disabilities, to treat a burn on the arm of one of their clients, Simone. The care organisation that hires our community nurse had recently acquired this house. Simone and the other group members moved in just a few weeks ago. I came for the third time to see Simone. The previous times, Simone was very sociable and she chatted all the time during the wound care. Really nice! But that day, she hung listlessly in a chair and did not even say hello. Oh yes, stress caused by moving, I diagnosed immediately. When I asked her to grab the wound care materials, Simone whispered, “Can’t.” It seemed to me she could start crying any moment now. There was a support worker I didn’t know yet, and he got the wound care materials. I asked Simone to lift her sleeve so that I could reach the wound. She did it agonisingly slowly. Everything seemed too much for her. Afterwards, I remarked to the support worker that moving to this new living group can cause Simone so much stress. But then he said that the psychologist is currently examining Simone to see if there is anything else going on, like depression. That scared me. When I was outside again and cycled to the next patient, I realised that I had made my diagnosis far too quickly. I have to be careful to keep on reasoning critically and not to take such a diagnosis for granted straight away, but to test it first!

1.9.1 What Is Meant by ‘Hypothesis Formulation’ and by ‘Hypothesis Testing’? Clinical reasoning is about the conscious thought process, where the nurse uses thinking skills such as analysis, hypothesis formulation, hypothesis testing and judgement. In the process of reasoning about diagnosis, etiology, interventions and prognosis, there is the risk of reaching a decision based on insufficient or incorrect information. Furthermore, the nurse may have cared for so many patients by now, that she is making decisions routinely. In both cases, this increases the risk of incorrect decisions, resulting in the patient receiving less good care than they could have. Incorrect decisions can, in some cases, lead to errors in patient care. By learning to work with hypothesis formulation and hypothesis testing in clinical reasoning, incorrect decisions due to sloppiness of thought can largely be prevented. The word ‘hypothesis’ refers to uncertainty. When one is unsure of something, an assumption is formulated in a hypothesis statement, after which the hypothesis is critically tested for correctness. To formulate the hypothesis, the nurse needs professional knowledge. To critically evaluate the hypothesis, the nurse needs systematically collected (patient) data. Hypothesis formulation and testing force the nurse to stay on alert, and it facilitates communication between the nurse, patient and his relatives.

1.9  The Use of Hypotheses in Clinical Reasoning

25

A hypothesis is a statement that expresses a possible viewpoint that has not yet been tested for correctness. The possible positions are diagnostic statements (‘This is what may be the matter with the patient’), etiological statements (‘This problem is caused by this’), prognostic statements (‘This is what we can achieve’) or therapeutic statements (‘If we carry out this intervention, the problem will become smaller’). To test the correctness of the hypothesis, test criteria are formulated. Only when the nurse understands what is going on with the patient (pathophysiological, psychological, social) is she able to formulate the correct test criteria. The formulation and testing of the hypothesis force the nurse step by step to understand the patient’s condition, while the testing through the test criteria allows the nurse to be accountable for the decision and decision-making [15].

1.9.2 How Can I Formulate Hypotheses and Test Criteria About Health Problems? The example below illustrates how hypothesis formulation and how test criteria about a health problem can be generated. Nursing student Emma is doing an internship in community nursing. In the afternoon of a hot summer day, she arrives at the apartment of 79-year-old Mrs Ruzicka. Because Mrs Ruzicka was not feeling well yesterday and was a bit nauseous Emma pays an extra visit today, since Mrs Ruzicka usually only receives care every other day. Inside the apartment on the top floor of the building, it is very warm. Emma immediately notices that Mrs Ruzicka is somewhat confused. Confusion, nausea and heat: Emma puts these together and thinks that Mrs Ruzicka may not have had enough to drink. She hypothesises that Mrs Ruzicka is suffering from dehydration. Emma looks around the living room and the kitchen to see if there is a cup or glass. This is not the case. Emma thinks that if Mrs Ruzicka is indeed dehydrated, so she can expect Mrs Ruzicka will (1) have dry, less elastic skin, (2) have a dry mouth, (3) perspire less and (4) produce less urine, but also (5) that the urine she does produce will be more concentrated. Furthermore, she expects (6) that Mrs Ruzicka is confused because her brain cells have also received less moisture and are dehydrated. Emma formulates these test criteria based on her knowledge of physiology. She now can test these criteria by observation and inspection. As the example of Mrs Ruzicka shows, the test criteria focus on the phenomena that characterise the health problem (dehydration). These phenomena are (in the case of dehydration) derived from pathophysiology. The nurse needs to know what the defining characteristics of the health problem are, in order to be able to use this

26

1  Background Knowledge in Clinical Reasoning

knowledge to generate adequate test criteria. For non-somatic problems, such as loneliness, the knowledge comes from psychological sources. For many health problems, the defining characteristics are well known, due to scientific research. For this reason, foreground literature can also be searched for defining characteristics. Finally, there exist manuals with standardised nursing diagnoses, in which common symptoms are described. The disadvantage of these manuals, however, is that a large part of the diagnoses described are not evidence-based. In addition, the risk of errors in diagnostic reasoning is greater, because nurses rely too much on the standard diagnosis they look up in the manual and therefore reason less well. As a result, there is a risk of quasi-professionalism on the part of the nurse [16]. Another risk when looking for the right health problem is that certain problems are excluded at too early a stage or that the nurse approaches the problem too unilaterally and only looks in a certain direction. To avoid this kind of narrow vision, it is important to formulate more than just one hypothesis about the problem (Sect. 12.17). Expected problems (high-risk diagnoses) can only be argued on the basis of theory. After all, a high-risk diagnosis indicates that the patient has an increased risk of that problem arising in the future. This risk can be estimated on the basis of theory and experience, but its substantiation can only be based on theory and evidence. The test criteria for hypothetical high-risk diagnoses focus on the presence of risk factors, which, in turn, are based on the best available evidence.

1.9.3 How Can Hypotheses About Etiology Be Tested? The second element of the health problem or nursing diagnosis about which hypotheses need to be formulated is the causal and related factors. Test criteria focus on whether (1) the factor in question can indeed cause or maintain the problem, (2) this factor is indeed present here and (3) whether this factor is indeed having this effect here. Because many health problems are caused or maintained by several factors simultaneously, most nursing diagnoses mention several causal and related factors (Sect. 12.18).

1.9.4 How Can Goals of Nursing Care Be Assessed? Goals of nursing care should preferably be formulated in terms of the projected patient’s behaviour over a specified period of time. The goal should clearly show what the nursing care delivers to the patient. The goals are based on the expectation that the interventions will have an effect, for example, reducing or completely resolving the health problem. This means that the goals are themselves hypotheses: they describe what is expected to be achieved. Test criteria in this case are the intermediate goals. They describe the intermediate progress that can be expected if the interventions actually work (Sect. 12.19).

1.10  The Thought Structure of Clinical Reasoning

27

1.9.5 What Should the Hypotheses Be About When Reasoning About Interventions? If the intervention acts on the cause of the problem, it is likely that the problem will be reduced. The hypothesis is therefore that a particular intervention is the best intervention for this patient. The testing criteria for the intervention are focused on the mechanism of change of the intervention that addresses the etiology or on the proven effect of the intervention. In addition, there are test criteria for the acceptability of the intervention to the patient. Again, theory and evidence are needed to formulate the correct criteria (Sect. 12.20).

1.10 The Thought Structure of Clinical Reasoning 1.10.1 How Do We Think? The Dual Processing Model and Pitfalls in Decision-Making During my 1-week internship in a hospital, I met Mr Gilbert. Mr Gilbert is a 41-year-­ old man with a broken leg. He had had surgery on his leg that day, and shortly after the doctor had visited him, I walked over to ask how he feels. With amazement in his voice, the patient told me that when the doctor was there, he had a cough, and that the doctor then told him that he should stop smoking, or else he would probably die sooner. Mr Gilbert told me that he smokes about 20 cigarettes a day. He said he had tried to quit many times before, but had never succeeded. He knew smoking was bad for him, but he just couldn’t quit. He asked me if he couldn’t just cut down, say by half to ten cigarettes a day. Would that help? I was at a loss for words. What would my answer be to this patient’s question? (Sect. 12.21)

The intuitive answer is that it is better to cut down than to continue smoking. However, this is a suboptimal response. Indeed, cutting smoking in half lowers the risk of dying from the effects of smoking. But limiting smoking to less than ten cigarettes per day reduces the risk of, for example, lung cancer by only about 35%, whereas quitting smoking altogether ultimately reduces this risk by as much as 80–90% [17]. To answer this question properly, nurses need to know if and how much reducing cigarette smoking reduces the risk of death within 2  years. This requires good-­ quality information. In many cases, nurses have this information, but it doesn’t immediately come to their minds. They need to think about it. And that ‘taking time to think’ is crucial. Our brain tends to simplify situations and difficult questions. Because this is an unconscious process, we do not notice that this process is taking place. If we do not suppress this tendency, we will make an intuitive decision. Intuitive decisions do get made quickly, but they are more often wrong or suboptimal decisions. In non-routine situations, nurses should adopt a rational-­analytical thinking strategy: they should take time to consciously think about the question or situation, analyse the situation and available information and only then come to a decision (see Fig. 1.3).

1  Background Knowledge in Clinical Reasoning

28

paent condion

- causes - diagnosis - therapy - prognosis

a queson or an uncertainty & bias - primacy - ignoring base rates - overconfidence - confirmaon bias

unconscious decision strategies (system 1)

- availability - representaveness - substuon

raonal-analyc decision strategies (system 2) - what exactly is the queson? - what other informaon do I need? informaon

intuive decision

*is there sufficient informaon? * weighing informaon - scienfic evidence - clinical experse - paent preference - costs and available resources

opmal decision

Fig. 1.3  Highly simplified dual processing model. (Based on Croskerry [18] and Kahnemann [19])

In the Dual Process Theory (DPT), these two systems are distinguished: an intuitive or heuristic system (also called ‘System 1’) and an analytical system (‘System 2’). Without we being aware of it, sensory stimuli activate all kinds of information in our brain. This causes information from the memory to enter our thoughts, but mainly it is a matter of chance which information from our memory is activated and which information is not. This may be partly relevant and partly irrelevant information. In addition, we usually have more relevant information in our long-term memory than is activated by the stimulus. So, the relevant information is available in our memory, but we don’t take the time to retrieve it, and therefore, we do not use it to assess the situation. If we remain in the left-hand half of Fig. 1.3 (System 1), an (intuitive) decision is made unconsciously on the basis of only the information that is accidentally activated by the stimulus. Such a decision is then based partly on relevant information and partly on irrelevant information. And, therefore, this thinking method quickly leads to suboptimal or wrong decisions (Sect. 12.22) [18, 19]. Before we make the decision, we may become aware that we have not yet thought hard enough about the situation or question. Via the dotted line (Fig. 1.3), we then cross over to the right-hand side of the diagram (System 2). It is also possible that we have been aware of the complexity of the situation or the question from the start.

1.10  The Thought Structure of Clinical Reasoning

29

In these cases, we think about the situation even more. But because we then think consciously, we can analyse the situation or question: what is the essence of the question or situation? What aspects are there to it? What information do I need? How do I weigh the available information? This way of analytical thinking takes more time, but significantly improves the quality of the decision (Sect. 12.23).

1.10.2 How Do We Think? Biases as Threats to Reasoning

When I learned about bias, I repeatedly noticed that biases influenced my thinking. And I saw this even more with colleagues. Now, when I have to make really important decisions, I force myself to take my time and check my reasoning. But especially when I’m tired, in a hurry or stressed, I notice that the bias sometimes weakens my reasoning.

Mrs Brown Mrs Brown (77) has diabetes and cancer with metastases. Because of problems with her balance and wounds on her leg, the community nurse comes by every day. The nurse helps Mrs Brown to get into the shower, after which she washes herself. After she has showered, her wounds are cared for. Afterwards, she manages by herself. Last week, Mrs Brown was admitted to hospital. Yesterday, she returned home. After the nurse had helped Mrs Brown into the shower, she asked if the nurse would wash her. After she had showered, she asked the nurse to help her get dressed so she would still have energy left to read the paper and take a short walk. Back at the community centre, the nurse updates Mrs Brown’s file. She is interrupted by her colleague. ‘How is Mrs Brown?’ ‘She wanted me to help her shower and dress this morning.’ ‘Yes, that’s what happens after hospitalisation. People lose their self-care. I had it last week with Mr Keitz, remember. Same story.’

1.10.2.1 What Are Biases and How Do They Affect Reasoning? Because the colleague suggested that the patient has a diagnosis of ‘self-care deficiency in washing and dressing, caused by hospitalisation’, the nurse is put on the track of this diagnosis. This is called a frame. The risk of a frame is that it suggests a particular way of looking at the patient’s condition, which automatically (i.e. unconsciously) leads the nurse to look at the patient’s condition in this way. The nurse is no longer open-minded. When this happens, the nurse’s reasoning is no longer unprejudiced, but skewed by the frame. These types of prejudices are called biases. They occur frequently. In the example, reasoning about the diagnosis is threatened by a framing bias (Sect. 12.24).

30

1  Background Knowledge in Clinical Reasoning

The nurse hears her colleague’s suggestion and immediately thinks of last week’s patient who had the same symptoms and the suggested diagnosis. With this image in mind, she is convinced that the suggested diagnosis is indeed the correct one for this patient. The problem is that the nurse has made a decision based on insufficient information and from a particular perspective. When people think back to previously experienced situations that are similar to the current situation, recent examples in particular are activated by thinking system 1. In this example, the nurse is thinking of a patient from last week (patient A). However, it is quite possible that the situation of a patient from 4 months ago (patient B) is much more similar to the situation of the current patient (patient C) than the situation of patient A from last week. Perhaps patient B had the same diagnosis as patient C, perhaps he had a different diagnosis. However, the nurse did not think of patient B at all when determining the diagnosis of patient C. Recent memories, in fact, emerge first and most easily [19, 20]. This bias is called recency. As the nurse considers whether the diagnosis can indeed be made on the basis of the data, she checks whether the diagnosis is a ‘true story for this patient’ based on the data available in her working memory. As she does this solely on the basis of the data available in her working memory at that moment, she no longer notices that information may be missing. Information that she should collect and use to rule out another diagnosis is no longer noticed. For example, Mrs Brown’s explanation that she wanted to conserve energy in order to perform other activities important to her is no longer activated by working memory and therefore is not included in working memory when making the most accurate nursing diagnosis. The nurse is incorrectly confident in her decision that this diagnosis is the correct one. This is called overconfidence bias. There are many different biases, and the human mind is very sensitive to them. Because biases work through the unconscious mind, they are difficult to influence. By being aware that they can occur in certain situations, their influence can be reduced. Most people can usually recognise biases in reasoning more easily in others than in themselves (Sect. 12.25).

1.10.3 Intuition and Analysis Many nurses feel that nursing is an intuitive profession. There are quite a few decisions that are made based on intuition. I was therefore shocked when I heard about Hodgetts’ research. He describes how in an English hospital about half of the patients who had a cardiac arrest in the hospital had symptoms in the previous 24 h that should have led to an alarm. In about one in five of those patients, the nurse had not responded at all or had responded too late to those signs. If the nurse had responded to the signs in these patients, the cardiac arrest could potentially have been prevented, according to an expert panel. This research convinced me that relying too much on intuition when making decisions can be detrimental, and that there are many situations in nursing where intuitive action is not appropriate, but analysis is necessary [21].

1.10  The Thought Structure of Clinical Reasoning

31

1.10.3.1 What Is the Difference Between Intuition and Analysis? When nurses base decisions on clinical intuition, they are making quick decisions, they make decisions quickly, unconsciously processing clinical information and using the outcome of this unconscious proces to inform their decision without further analysis. Intuitive decisions are moderately accurate. If nurses base a decision on clinical analysis, they take more time to make the decision, consciously considering exactly what decisions need to be made and how best to make them. Furthermore, clinical analysis allows nurses to use scientific information about the effect of action options and therefore their probable effect in the patient. Analytical decisions are usually fairly accurate, but if analytical methods are performed incorrectly, this can still lead to serious errors. In addition to errors arising from incorrect application of analytical methods, especially minor sloppiness in taking notes and in interpreting information can cause serious errors (Sect. 12.26) [22, 23]. Hence, reasoning can take place at different (cognitive) levels. In the case of intuitive decisions, very little conscious reasoning takes place. The more analytical the reasoning, the greater the demand on the nurse’s knowledge and the greater the nurse’s reliance on additional sources of knowledge. As it is written here, it seems that decisions should be made as analytically as possible. However, this is not the case. There are several factors that determine whether a more analytical decision is desirable or a more intuitive decision. For example, making a decision based on analysis takes more time than an intuitive decision. If a nurse is walking down the street and someone falls to the ground right before her feet from a cardiac arrest, there is no time for analysis, but an intuitive decision is necessary. If, in a routine situation, a nurse has to make a decision which she has already made many times in the past, there is no need to analyse the situation and the various courses of action. The nurse can then rely on her own experience and make a decision based on this (Sect. 12.27). From the work of Hammond et al. [24], a number of clear rules of thumb emerge about decision-making (and therefore about reasoning). The amount of time available strongly determines the ability to use an analytical approach. 1. If there is little time to make a decision, choose a more intuitive one. 2. If more time is available, use that time to analyse and obtain additional information. Opt for a more analytical decision. 3. Only use the analytical approach that you have mastered sufficiently. Otherwise, there is a risk of serious errors (e.g. calculating with a formula that you have not mastered or of which you are not sure whether it is the right formula for this question or using a scientific study that you do not understand sufficiently). Which analytical strategy is best depends on how much time is available. The list below goes from low-analytical strategies to increasingly high-analytical strategies. In a situation where a complex decision needs to be made: • How much time does it take to consult a fellow nurse? • How much time does it take to consult the protocol?

32

1  Background Knowledge in Clinical Reasoning

• How much time does it take to find, read and correctly process background literature? • Is there a specialist available? How much time will it take to consult them? • Is there an expert available? How much time will it take to consult them? • How much time does it take to search for, select, read and correctly process evidence-­based foreground literature? • Is information from evidence-based guidelines available? How much time does it take to search for it, read it and process it correctly? • Are there correct calculation rules or formulas for this decision? Are they available? How much time will it take to apply them correctly?

1.10.4 Background Knowledge and Foreground Knowledge When I first heard that a patient might have a swallowing disorder, I thought: what is a swallowing disorder, and how do you know if someone has a swallowing disorder? My internship supervisor was quick to point out that I would have to answer the first question first before I could get to the second. She told me that the background knowledge of the first question is needed to understand the foreground knowledge of the second question. So, then I thought, what is background knowledge, and what is foreground knowledge?

1.10.4.1 What Is Background Knowledge? Background knowledge is basic knowledge: knowledge about anatomy, physiology, psychology and medical and psychiatric conditions in order to know and understand the condition and problems of patients. This type of basic knowledge can be found, for example, in medical and psychology textbooks. Basic knowledge is necessary to be able to reason with sufficient depth about patients and their condition. Sufficient background knowledge on a particular subject is also a prerequisite for understanding the foreground knowledge on that subject. In addition to basic knowledge, the nurse needs to deepen specific nursing background knowledge about the patient group with whom the nurse is working. This depth of knowledge greatly enhances the quality of clinical reasoning. It allows nurses to add depth to their clinical reasoning. 1.10.4.2 What Is Foreground Knowledge? In order to answer the second question (‘How do you know if someone has a swallowing disorder?’), foreground knowledge is needed. Foreground knowledge is knowledge about ways of establishing a specific diagnosis (diagnostic test) or knowledge about interventions: What is the best intervention for this patient in this situation? Scientific research constantly produces new knowledge that must be included in the answers to the foreground questions. Foreground knowledge therefore quickly becomes outdated. It cannot simply be assumed that foreground

1.10  The Thought Structure of Clinical Reasoning

33

knowledge from 5 years ago is still correct today. Foreground knowledge can be found in scientific journals and can be accessed through a systematic search of scientific databases on the Internet.

Glossary Activities  Execution of a task or action by an individual. Examples: focusing attention, solving problems, moving around, washing oneself, drinking, doing housework [5]. Autonomous Practice  Autonomous practice means that the nurse makes decisions and carries out actions for which she is primarily responsible. In autonomous practice, the nurse has this autonomy, since these actions fit in the nursing professional domain. Background knowledge, background literature  Background knowledge is basic knowledge that nurses in training acquire through their textbooks. In contrast to foreground knowledge, background knowledge involves relatively unchanging knowledge, such as anatomy, physiology and developmental psychology. Basic questions  The four types of questions that almost all questions nurses ask in clinical reasoning fit into. The four basic questions are:   1. What may be the matter with the patient (diagnostic question).   2. What causes this? (causal or etiological question)   3. How is this problem likely to end? What do we think we can achieve? (prognostic question)   4. What do we think we can do about it? (therapeutic question) Best decision  The optimal decision that can be taken for this patient under the given circumstances. Bias (in reasoning)  In reasoning, biases are ‘predictable deviations from rational thought’ (Croskerry et  al. [25], p. ii60). There are sloppiness in thinking and errors in thinking, which cause the reasoning to be distorted rather than pure. As a result, the reasoning is to a greater or lesser degree false. Body functions  Physiological functions of body systems (including psychological functions). Examples: consciousness, sleep, smell function, maintenance appropriate body weight, muscle power [5]. Body structures  Anatomical parts of the body such as organs, limbs and their components. Examples: anatomical structure of the retina, of the heart, of the hip joint [5]. Clinical analysis  Purposeful, rule-based thinking to assess a patient situation. This involves consciously and deliberately obtaining information about a situation. This information is then consciously weighed and assessed. Decisions based on adequate clinical analysis are usually accurate. Because clinical analysis takes thought time (and sometimes time to obtain the information), it is a relatively slow mode of thinking.

34

1  Background Knowledge in Clinical Reasoning

Clinical intuition  Unconscious, associative thinking to assess a patient situation. Clinical intuition is based on a reflexive and automated thinking pattern, which is focused on finding causes and finding patterns. Decisions based on clinical intuition are prone to error. Because of the automated and reflexive way of thinking, making an intuitive decision takes little time and is thus a relatively fast mode of thinking. Collaborative practice  Participation in treatment or decision-making for which a professional other than the nurse is primarily responsible (also: participatory practice). Critical thinking  The use of higher thinking skills (such as analysis and judgement) in processing information in such a way that it increases the likelihood that it will lead to decisions and actions that are logical, correct, adequate and appropriate (compare: Huang et al. [2], pp. 95–96) Defining characteristics (signs and symptoms) Symptoms that are always or almost always present for a specific health problem. Because these symptoms are always present, they characterise the health problem. Signs and symptoms are the most important test criteria when checking whether a specific health problem is present. For example: the defining characteristic of obesity is: a BMI of 30 or more. Diagnostic question Question that aims to make a better assessment of the patient’s state of health. The question can focus on the nursing, medical and/or psychological state of the patient. Diagnostic questions concern the presence or absence of (defining characteristic) signs, symptoms, complaints and behaviour of the patient. Diagnostic reasoning Clinical reasoning to establish the correct diagnosis or health problem. Diagnostic test (diagnostic measurement instrument or tool)  Test that, or measurement instrument that, helps to make a diagnosis. The test or measurement instrument increases certainty about the patient’s condition. This is preferably done with the test that, or the instrument that, has the fewest errors. Environmental factors  The physical, social and attitudinal environment in which people live and conduct their lives. Examples: medicines, glasses, educational toys, money, climate, family, social norms, housing, presence or absence of healthcare facilities [5]. Etiological question Question aimed at discovering the cause or causes of the patient’s health problem. In addition, etiological questions can be aimed at discovering related factors. Related factors are either (1) factors that did not cause the patient’s health problem but contributed to it or (2) factors that (if the patient already has the health problem) contribute to the persistence of the health problem. Evidence-based  Based on current best evidence. Evidence-based guideline  Recommendations based on scientific evidence to support clinical decision-making and linked to a specific patient problem or condition. The scientific evidence on which the recommendations are based is clearly visible in the guideline. Guidelines are written by independent experts in the field

1.10  The Thought Structure of Clinical Reasoning

35

of the content of the guideline. Guidelines reflect the state of the art, but also address topics that are still being debated in practice and, in many cases, also practice variation. Evidence-based guidelines translate research results into practice. The recommendations are therefore not non-binding, but departures from them should be justified (see also: guideline). Evidence-based practice  ‘The integration of best research evidence with clinical expertise and patient values to facilitate clinical decision making’ [13]. Expert  Professional who has expert knowledge and experience in a defined area and who in situations within his own area of expertise can quickly access the relevant information stored in his memory via minimal information cues (compare: Kahneman [19], p. 11). Focus of the intervention  The element in the diagnosis and/or prognostic factor that the intervention addresses and that is changed by the intervention. Intervention targets can be (a) causes, (b) related factors and (c) symptoms. In addition, (d) prognostic factors and risk factors can also be a focus of the intervention. If the intervention targets only one or more symptoms, it only addresses the manifestation of the health problem (e.g. burdening symptoms). Foreground knowledge, foreground literature Foreground knowledge is relatively rapidly changing knowledge. This knowledge changes as new knowledge and insights are acquired through scientific research. Examples of foreground knowledge are knowledge about the most effective intervention for a particular health condition or about the most important prognostic factors for a particular health problem. Frame  The suggestion of a particular way of looking at a subject. Persons exposed to a frame often unconsciously look at that subject from the angle of the suggested frame. Different ways of presenting information tend to evoke different mindsets and different emotions. Although it is often obvious that there are more angles from which to look at information, the person only sees one angle (that of the frame, i.e. the way in which the information is presented) and does not think about the fact that the information can also be viewed and interpreted in other ways. Framing bias  Bias in reasoning, because the reasoning takes place from a mode of looking or thinking (whether or not suggested by another person), as a result of which other perspectives are disregarded. High-risk diagnosis  Diagnosis used to describe an increased risk of the patient. An example is ‘increased risk of pressure ulcers’. The patient does not yet have pressure ulcers, but risk factors are present, making him or her at significantly higher risk compared to other patients. A high-risk diagnosis is represented in a PR structure, where P means ‘Problem’ and R means ‘Risk factors’. Hypothesis  Statement expressing a viewpoint that has not yet been tested for accuracy. Possible positions are diagnostic, causal, prognostic or therapeutic statements. Hypothesis formulation  The process by which the healthcare professional forms an idea about what the best decision for the patient might be. The more the healthcare professional believes he knows what the best decision is, the more

36

1  Background Knowledge in Clinical Reasoning

concrete the hypothesis becomes. Ultimately, the hypothesis about the best decision is formulated as a testable statement: ‘The diagnosis is probably … The health problem is probably caused by … The patient can probably perform activity Y independently within X days … The best intervention is probably …’. Hypothesis testing  Check whether the hypothesis is correct by looking for both information that confirms the hypothesis and information that rejects the hypothesis. The search for this information is guided by test criteria that the healthcare professional has established in advance. Example: in the case of a diagnosis of increasing severity of pressure ulcers, characterised by the transition from non-­ erasable redness to a superficial skin defect, an assessment criterion for the interventions may be: ‘Does at least one of the interventions reduce the pressure, friction and/or shear, exerted on the ulcer area?’. Impairments  Problems in body functions or in body structures such as a significant deviation or loss. Examples: impairment in pain perception, impairment in absorption of food, intestinal perforation, atrophy of muscle tissue, opacity of the lens of the eye [5]. Intuitive decision-making Rapid and unconscious use of the information that automatically and immediately comes to mind in a situation or following a question, without analysing the situation and the information. An increased risk of errors and suboptimal decisions is accepted. Limitations  Difficulties an individual may have in executing activities. Examples: limitation in handling stress, in reading, in walking, in washing oneself [5]. Mechanism of action  The way in which the etiological or prognostic factor brings about the health change in the patient. The mechanism of action explains which and how physiological or psycho-social processes deviate from their normal course and how they affect the health status. Mechanism of change  The way in which the intervention (or the etiological or prognostic factor) brings about the desired change in the patient. For example, many drugs affect chemical processes in the body, and many psychological interventions affect psychological processes. The mechanism of change explains through which physiological, psychological or other process the intervention intervenes and makes changes and how those physiological or psychological changes contribute to the desired effect of the intervention. Overconfidence  A reasoning bias in which one believes too strongly in a conclusion drawn on the basis of information that happens to be available at that moment. The conclusion is not based on the information required for that conclusion, but on the information that happens to be present. With the information that is present by chance, thinking system 1 constructs a coherent story. Because the story is coherent, one believes in its correctness and subsequently trusts in its correctness and does not seek for further information. Overconfidence means that the confidence in the conclusion is determined by the coherence of the story, not by the quality of the data with which the story is constructed. Participation  Involvement in a life situation. Examples: remunerative or non-­ remunerative employment, socialising, recreation and leisure [5].

1.10  The Thought Structure of Clinical Reasoning

37

Participatory practice  Participation in treatment or decision-making for which a professional other than the nurse is primarily responsible (also: collaborative practice). Personal factors  A person’s individual background. Examples: age, gender, social status, life events [4, 5]. PES  Abbreviation for: Problem—Etiological and related factors—Signs and symptoms. This is the structure to formulate an actual diagnosis. PR  Abbreviation for: Problem—Risk Factors. This is the structure to formulate a high-risk diagnosis. Prognostic factor  Factor that, if the health problem is already present, positively or negatively influences its course and outcome and in this way also influences the health outcome(s). Prognostic question  Question aimed at discovering factors that influence the further course of the patient’s health problem (prognostic factors). In addition, prognostic questions can be aimed at determining what a realistic goal is for the care that will be provided to this patient. Prognostic reasoning  Clinical reasoning to find the prognostic factors in a health problem, weigh them and determine the prognosis. Prognostic reasoning also includes the goal setting for patient care. Rational-analytical thinking strategy Thinking strategy which involves consciously thinking about the decision to be taken, analysing the problem or decision to be taken, consciously gathering and weighing the information, various alternatives for action and consciously choosing from the alternatives. Recency  Bias in reasoning which arises because thinking system 1 is more likely to activate recent events in the memory than events from longer ago. Recency bias occurs despite the fact that events from longer ago may be much more relevant to a current decision than recent events. Related factors  Factors that do not cause the health problem, but which, once the health problem is present, help maintain the presence of this problem, or factors that did not cause the health problem, but which contributed to the arising of the problem. Restriction (participation restriction)  Problems an individual may experience in involvement in life situations. Examples: restrictions in entering into informal social relationships, using money to purchase goods or services, engaging in informally or formally organised sports events [5]. Risk factor  Factor that, if a specific health problem is not yet present, increases the risk of developing that health problem. Routine situations  Situations that the nurse often experiences and has experienced and which always call for the same decisions. Self-management  The level of independence (from care professionals) that patients have achieved. This independence relates to making one’s own decisions, being able to plan and carry out activities oneself. The patient does not have to carry out everything himself: even if he can decide when which action is carried out for him and in which way are these important aspects of self-management.

38

1  Background Knowledge in Clinical Reasoning

Specialist  Professional who has extensive knowledge in a particular area or subject based on his or her training and experience. Note: in everyday language ‘specialist’ is sometimes used as a synonym for ‘expert’. In this book, however, these terms are distinguished, with an expert having significantly more knowledge and skills within his or her area of expertise than a specialist. Suboptimal decisions  Decisions that do not lead to errors (nothing goes wrong) but are not the best decisions in the given situation. A better decision would have been possible in that situation, for example, a more effective intervention or a clearer diagnosis. System 1  Thinking system 1 is the intuitive thinking system. It is automatically and unconsciously triggered by stimuli. It is characterised by the speed of the unconscious thought process and because it works through simplification. Thinking system 1 constantly offers impressions, intuitions, intentions and feelings. Because of the way in which thinking system 1 works (quickly, but also sloppily and error-prone), it is suitable for everyday and routine matters, in which it usually offers the right suggestions. System 2  Thinking system 2 is the rational-analytical thinking system. It can consciously retrieve and process information. Thinking system 2 operates via the working memory and in most situations is only active to a limited extent. When thinking system 2 is in full operation, it uses (almost) the entire working memory. We experience this as deep thinking. In complex and non-routine situations, the use of thinking system 2 is necessary, because it significantly reduces the risk of thinking errors and (therefore) of wrong decisions. However, thinking system 2 works significantly slower than thinking system 1: it has to ‘think about it for a while’. Thinking system 2 constantly receives impressions, intuitions, intentions and feelings from thinking system 1 and in most cases accepts these suggestions without adjustment. Test criteria  Criteria by which the hypothesis about diagnostic, causal, prognostic or therapeutic viewpoint can be tested. Test criteria should be chosen and worded so as to seriously test the correctness of the viewpoint. Test criteria are usually worded in an imperative way. An example is a test criterion for fever: measured rectal body temperature is at least 38 °C (100.4 °F). Therapeutic questions  Questions aimed at determining the best interventions to be carried out on this patient, experiencing this specific health problem. Therapeutic reasoning  Clinical reasoning about the expected effect of interventions and about the appropriateness of interventions, in order to select the most appropriate interventions for the patient and to choose them together with the patient. Many interventions have both desired and undesired effects. In therapeutic reasoning, both types of effect are included in the reasoning. Wrong decision  A wrong decision is either (1) a decision that leads to harm or some other undesirable effect on the patient, while this effect could have been prevented if another reasonable decision had been made, or (2) a decision as a result of which a desired effect did not occur, while this effect would have occurred if another reasonable decision had been made. Examples: (1) An inter-

References

39

vention was chosen that would have been good for most patients, but due to this patient’s vulnerability, it led to a worsening of his condition. (2) Due to a wrong diagnosis, an intervention was carried out that belonged to the diagnosis but was useless for this patient.

References 1. Schuurmans M, Lambregts J, Project Group V&VN 2020, Grotendorst A. Beroepsprofiel verpleegkundige. Utrecht: V&VN; 2012. 2. Huang GC, Newman LR, Schwarzstein RM. Critical thinking in health professions education: summary and concensus statements of the 2011 Millennium Conference. Teach Learn Med. 2014;26:95–102. https://doi.org/10.1080/10401334.2013.857335. 3. Miller WR, Rollnick S. Motivational interviewing. Helping people change. 3rd ed. New York, NY: The Guilford Press; 2013. 4. WHO. ICF: Nederlandse vertaling van de international classification of functioning, disability and health. 3rd ed. Houten: Bohn Stafleu Van Loghum; 2018. 5. World Health Organization (WHO). n.d.. https://www.who.int/standards/classifications/ international-­classification-­of-­functioning-­disability-­and-­health. Accessed 2 Aug 2022. 6. Guyatt GH, Oxman AD, Vist GE, Kunz R, Falck-Ytter Y, Schünemann HJ. GRADE: what is ‘quality of evidence’ and why is it important to clinicians? BMJ. 2008;336:995–8. https://doi. org/10.1136/bmj.39490.551019.BE. 7. International Council of Nurses (ICN). n.d.. https://www.icn.ch/nursing-­policy/nursing-­ definitions. Accessed 3 Aug 2022. 8. Robinson MA, Cross-Denny B, Lee KK, Werkmeister Rozas LM, Yamada AM. Teaching note Teaching intersectionality: transforming cultural competence content in social work education. J Soc Work Educ. 2016;52:509–17. https://doi.org/10.1080/10437797.2016.1198297. 9. Muntinga ME, Krajenbrink VQE, Peerdeman SM, Croiset G, Verdonk P.  Toward diversity-­ responsive medical education: taking an intersectionality-based approach to a curriculum evaluation. Adv Health Sci Educ. 2016;21:541–59. https://doi.org/10.1007/s10459-­015-­9650-­9. 10. Network Caleidoscopia. n.d.. https://caleidoscopia.nl. Accessed 18 Apr 2020. 11. Hankivsky O. Women’s health, men’s health, and gender and health: implications of intersectionality. Soc Sci Med. 2012;74:1712–20. https://doi.org/10.1016/j.socscimed.2011.11.029. 12. Blackie M, Wear D, Zarconi J. Narrative intersectionality in caring for marginalized disadvantaged patients: thinking beyond categories in medical education and care. Acad Med. 2019;94:59–63. https://doi.org/10.1097/ACM.0000000000002425. 13. DiCenso A, Guyatt G, Ciliska D. Evidence-based nursing. St. Louis, MO: Mosby; 2005. 14. Guyatt G, Rennie D, Meade MO, Cook DJ. Users’ guides to the medical literature. A manual for evidence-based clinical practice. 3rd ed. New York, NY: McGraw Hill Education; 2015. 15. Rothstein JM, Echternach JL, Riddle DL.  The hypothesis-oriented algorithm for clinicians II (HOAC II): a guide for patient management. Phys Ther. 2003;83:455–70. https://doi. org/10.1093/ptj/83.5.455. 16. Van Meijel BKG.  Kleurrijke perspectieven: Notities over de GGZ-verpleegkunde. Oratie. Amsterdam: VU/VUmc; 2015. 17. Thun MJ, Carter BD, Feskanich D, Freedman ND, Prentice R, Lopez AD, et al. 50-Year trends in smoking-related mortality in United States. N Engl J Med. 2013;368:351–64. https://doi. org/10.1056/NEJMsa1211127. 18. Croskerry P.  Context is everything, or how could I have been that stupid? Healthc Q. 2009;12:e171–7. https://doi.org/10.12927/hcq.2009.20945. 19. Kahneman D. Thinking, fast and slow. London: Allen Lane/Penguin Books; 2011. 20. Lewis M. The undoing project. London: Allen Lane/Penguin Books; 2017.

40

1  Background Knowledge in Clinical Reasoning

21. Hodgetts TJ, Kenward G, Vlackonikolis I, Payne S, Castle N, Crouch R, et  al. Incidence, location and reasons for avoidable in-hospital cardiac arrest in a district general hospital. Resuscitation. 2002;54:115–23. https://doi.org/10.1016/s0300-­9572(02)00098-­9. 22. Thompson C, Dowding D.  Clinical decision making and judgement in nursing. London: Churchill Livingstone; 2002. 23. Thompson C, Yang H.  Nurses’ decisions, irreducible uncertainty and maximizing nurses’ contribution to patient safety. Healthc Q. 2009;12(Special Issue):e178–85. https://doi. org/10.12927/hcq.2009.20946. 24. Hammond KR, Hamm RM, Grassia J, Pearson T.  Direct comparison of the efficacy of intuitive and analytical cognition in expert judgment. IEEE Trans Syst Man Cybernet. 1987;SMC-17:753–70. https://doi.org/10.1109/tsmc.1987.6499282. 25. Croskerry P, Sighal G, Mamede S. Cognitive debiasing 1: origins of bias and theory of debiasing. BMJ Qual Saf. 2013;22:ii58–64. https://doi.org/10.1156/bmjqs-­2013-­002378.

2

Diagnostic Decision: What Is the Matter with the Patient?

I can still remember my first day as an intern very well. I was allowed to shadow a nurse for a whole day. Together we looked at a wound on a lady who had just been admitted. I was shocked by the wound, but my supervisor immediately said: ‘Oh, I see: stage 2 pressure ulcer’. Later, I asked her how she saw that so quickly. She then told me that I would probably learn this quickly too, because these kinds of ulcers were common in this ward, as well as in other places in the healthcare system. She said I should pay attention to what distinguishes this wound from other wounds. If I learned that, I would be able to make such a diagnosis at the end of my internship as well.

2.1 What Is a Diagnosis? A diagnosis is a description of the patient’s condition in a few words. A diagnosis is an answer to the basic question, ‘What is the matter with the patient?’ The nurse can determine the patient’s condition by gathering information about the patient and interpreting it. The interpretation leads to a name that describes the condition, in the same way that a doctor makes medical diagnoses, such as depression, pneumonia or heart attack. Nurses make nursing diagnoses, for example, obesity or pain. The name of the diagnosis is officially called ‘the label’ of the diagnosis or ‘the health problem’.

Supplementary Information The online version contains supplementary material available at https://doi.org/10.1007/978-­3-­031-­27069-­7_2. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 J. Dobber et al., Clinical Reasoning and Evidence-Based Practice, https://doi.org/10.1007/978-3-031-27069-7_2

41

42

2  Diagnostic Decision: What Is the Matter with the Patient?

2.2 What Is Diagnostic Reasoning? Diagnostic reasoning is a form of clinical reasoning. It aims to assess the patient’s condition and to answer the question: what may be the matter with the patient. In doing so, the nurse seeks, gathers and processes information from and about the patient. The nurse questions, examines, measures and observes the patient in a targeted manner. In this way they obtain information about the patient’s complaints, behaviour and symptoms. The nurse links this information to her professional knowledge, using a diagnostic mental scheme of standard questions (see below). In each of these steps, the nurse reasons: What information do I need? How do I get this information? What does this information mean for the patient’s health status? Do I have enough data or do I need more information? What health conditions are suggested by the data I have collected? For diagnostic reasoning, the nurse uses information from and about the patient, such as symptoms, behaviours and signs. She makes decisions about the patient’s condition based on these signs and symptoms. The signs and symptoms collected by the nurse, and their processing and interpretation, form a crucial part of the diagnostic reasoning process. If the nurse works with insufficient or incorrect information, this leads to errors in diagnostic reasoning and, consequently, an incorrect diagnosis. Similarly, if the nurse is working with accurate information, but does not process it correctly, or doesn’t link it to her professional knowledge or interpret it incorrectly, errors in diagnostic reasoning will occur, and a correct diagnosis will probably not be made. The standard questions in diagnostic reasoning focus on interpreting the symptoms. They are intended to determine what health problems are present. Ultimately, the standard questions lead to an informed answer to the question: ‘What may be the matter with the patient?’. 1. What objective and subjective symptoms are present in the patient (complaints, signs, behaviours, abnormal value compared to the reference range)? 2. How can these symptoms be explained (which abnormalities in anatomy, physiology and psychosocial functioning)? 3. What else would you like to know from or about the patient? Use the answer to this question to supplement the answers to Questions 1 and 2. 4. What is the patient’s hindrance (possibly expressed in the ICF components activities, participation)? 5. How does the patient react to this (emotional behaviour, effective or ineffective self-management)? 6. Which possible health problems does this indicate (from which impairments in body structures, in body functions and limitations in activities and restrictions in participation)? Take 5 min to complete the problem list. For each problem, indicate whether it is an important or a minor problem. 7. What are the defining characteristics of the major health problems? 8. How can the presence or absence of these characteristics be tested? Then test whether these phenomena are actually present (Sect. 12.7).

2.4  When Should I Use Diagnostic Reasoning?

43

After reviewing the defining characteristics, the nurse has an overview of the health problems (Sect. 12.71.1).

2.3 How Are the Health Problem, Clues and Symptoms Incorporated into the Diagnosis? Most diagnoses consist of three elements: 1. the health problem 2. the defining characteristics (and possibly other signs and symptoms) of that health problem 3. the causes and related factors for that health problem The third element is discussed in the chapter on etiological reasoning. In the current chapter, the first two elements are concerned. Thus, the health problem represents the patient’s condition in one or a few words; the defining characteristics are those complaints, behaviours and symptoms that are almost always present in a specific diagnosis. As a result, these symptoms characterise that diagnosis. For example, a BMI of 25–29.9 is the characteristic symptom for the health problem ‘overweight’ [1]. An enlarged abdomen is often used as an additional characteristic [2]. Thus, the characteristic phenomena are used as a kind of criteria for making the diagnosis. The health problem and the (characteristic) signs and symptoms can be represented as follows: ‘Overweight, characterised by a BMI of 29.1 and an abdominal girth of 102 cm’. Hence, the structure is: ‘Health problem A, characterised by characteristic signs and symptoms 1, 2, 3 and 4’ (Sect. 12.28).

2.4 When Should I Use Diagnostic Reasoning? The simple answer to this question is: always. But to answer this question properly, it is necessary to distinguish between a narrow and a broad interpretation of the term ‘diagnosis’. A narrow view of the term ‘diagnosis’ means that ‘diagnosis’ is restricted to formal medical or nursing diagnoses, such as those already mentioned above (depression, pneumonia, heart attack, obesity, pain). Medical diagnoses are made exclusively by the physician. These medical diagnoses have been scientifically researched and collected by the World Health Organization in an international medical classification system, the International Classification of Diseases and related Health Problems (ICD) [3]. Nursing does not have such an international classification system, although several manuals of nursing diagnoses do exist. A major drawback of these manuals is that most of the nursing diagnoses they contain have not been scientifically researched, so it is not certain

44

2  Diagnostic Decision: What Is the Matter with the Patient?

whether they are correct. However, there are good thought models when reasoning about nursing diagnoses, such as the ICF model (see Sect. 1.4). The term ‘diagnosis’ can also be understood more broadly, as an expression of the patient’s current or expected health condition. In their work, nurses must constantly assess the condition of patients, and therefore, they must constantly engage in diagnostic reasoning. The importance of diagnostic reasoning varies with the occurrence of changes in the patient’s condition. If a patient is in a stable condition, his condition does not change suddenly, and no diagnostic assessment is required other than that the patient’s condition is unchanged. It also happens that a patient’s condition suddenly changes unexpectedly. Because the nurse has previously monitored this patient closely and determined that his ‘condition was unchanged’, she quickly notices when his condition changes. A new diagnostic reasoning is then required to make a correct decision about the current condition. In some situations, the patient’s condition changes rapidly and continuously, and continuous diagnostic reasoning is necessary. This keeping an eye on the patient is called monitoring. The nurse constantly monitors all patients. By actively monitoring and constantly making assessments of the patient’s condition, the nurse is repeatedly carrying out mini-diagnostic reasoning. If the patient’s condition changes, the nurse can quickly switch to a more thorough diagnostic reasoning. If an acute situation arises (e.g. the patient suffers a cardiac arrest), the nurse must make a rapid diagnostic assessment, followed immediately by appropriate intervention. In many cases, the diagnostic reasoning used for monitoring is medical reasoning and, in that case, falls within the medical professional domain. This type of reasoning is very common and is part of the collaborative practice of the nurse (see Sect. 1.6). By using this diagnostic reasoning, the nurse is participating in a course of treatment for which another professional (e.g. physician) is primarily responsible (Sect. 12.29).

2.5 How Do I Get a Diagnosis? To be able to make good diagnostic decisions, nurses need to have accurate professional knowledge of their subject. This includes a thorough background knowledge of basic subjects (anatomy, physiology, pathology, psychology, psychopathology), the ability to differentiate between normal and abnormal functioning, early recognition of (warning) signs of complications, recognition of adverse reactions to medication and the ability to explain the underlying mechanism of action (see also Sect. 1.10.4). Using this knowledge as a basis, the nurse will progress through a thought process of standard questions when performing diagnostic reasoning. These standard questions are described in Sect. 1.5.1.

2.6  How Do I Find Out a Patient’s Symptoms?

45

2.6 How Do I Find Out a Patient’s Symptoms? The nurse continually collects data about the patient. She uses a combination of methods to do this. Data collection usually begins with the taking of a case history. The case history is a structured conversation between the patient and the nurse, in which the nurse purposefully seeks information from the patient. A case history is taken at the start of the care process and contributes to the basic information required for the start of care. The nurse collects both the patient’s subjective experiences (also known as ‘symptoms’) and more objectively observable clues to what may be the matter with the patient (also known as ‘signs’). If a patient tells the nurse that they often suffer from headaches, this is the patient’s subjective experience. If the nurse sees that the patient constantly rubs his left eye, this is a possible objective indication that something may be wrong with the eye. In taking the case history, complaints, health problems and the patient’s perception of these are systematically charted and analysed by the nurse. The electronic patient record contains a standard format for the taking of the case history. The patient can tell by himself what his problems are and what his complaints are. The patient can also tell about the consequences of these symptoms (‘Because I have such difficulty walking now, I can hardly get out of the house’) and how he responds to them (‘But I don’t let myself get discouraged. I invite a lot of people to my house, have the groceries delivered and exercise my muscles as much as possible indoors’). In many cases, the nurse continues to ask questions. Most patients do not tell you everything themselves. Also, the patient often does not know what important information may be. Particularly if the nurse suspects health problems in a subject while taking the history, it is important to ask follow-up questions (and possibly also observe and measure later on). During the interview, the nurse attempts to assess the extent to which the health situation is normal or abnormal. To do this, she not only uses the information provided by the patient (symptoms, consequences for daily functioning, the way in which the patient deals with the situation) but also observations of symptoms (e.g. a red complexion) and behaviour (e.g. frequent yawning). This information may be supplemented with measurements (e.g. of height, weight, abdominal size and blood pressure). The nurse then orders the information from and about the patient by comparing it with her professional knowledge. Using their knowledge of normal values (e.g. a normal BMI is between 18.5 and 25), the nurse makes an initial assessment of whether there are problems in a particular area or whether problems may be expected at a later date. If the data are normal, then the nurse can leave that topic for now and move on to another topic. If, however, the data suggest a deviation from normality, the nurse will continue to ask questions, observe further and possibly take additional measurements (Sect. 12.30). Other sources, such as family members or the patient’s file, can sometimes be better sources of information than the patient himself, then part of the history taking

46

2  Diagnostic Decision: What Is the Matter with the Patient?

may take place with a significant other of the patient. Further, there might not always be enough time or necessity for a full history to be taken. In that case, taking an emergency history or a problem-oriented history can be a good alternative.

2.7 How Can I Assess the Severity and Importance of the Symptoms? The severity of the symptoms can best be assessed on the basis of professional knowledge about normal and abnormal human functioning. In the case of symptoms, it is important to reason back to the underlying anatomical or physiological structure or to normal psychosocial functioning. 1. What is abnormal in these structures that can cause these symptoms? 2. What should this structure normally be like? 3. What is the function of this structure? (For example, the function of the colon is breakdown, digestion, and absorption of nutrients and water.) 4. When are abnormalities in these structures so severe as to constitute a health problem? By understanding the symptoms, the nurse can reason about the severity and importance of the symptoms. This gives the diagnostic reasoning greater depth and quality. Whenever a patient’s vital functions are threatened, emergency intervention is necessary. To promote rapid recognition of these situations in hospitals, hospitals work with an early warning score (EWS) or a modified early warning score (MEWS). The MEWS is a combination of measurements and observations of (changes in) blood pressure, heart rate, respiratory rate, saturation, consciousness and temperature. A certain score on the MEWS requires immediate alarm. In addition, if the patient himself indicates that he is suddenly doing much worse or that he feels different, this is a signal to call in help immediately. In diagnostic reasoning about nursing diagnoses, the nurse should at least consider the patient’s hindrance of symptoms. How do they affect his daily activities and participation? How the patient reacts to his symptoms and health problems can also lead to a nursing diagnosis, for example, if the patient experiences strong negative emotions or if there is ineffective self-management.

2.8 How Do I Get from Symptoms to an Overview of Possible Health Problems? During data collection, the types of data (such as complaints, symptoms, behaviours) are not obtained separately, but simultaneously and mixed together. The nurse must bring order to this data. As nurses compare the data with their previous knowledge and experience, they are able to create order by recognising patterns. For

2.8  How Do I Get from Symptoms to an Overview of Possible Health Problems?

47

example, a patient admitted with pancreatic disorders and a clean history may be thirsty and urge to urinate frequently. The nurse will recognise this as a sign of a high blood glucose level. She will check this by taking a blood glucose sample. Based on their background nursing knowledge, the nurse will recognise a range of associated symptoms, behaviours and complaints. Together, these symptoms form a cue, which leads the nurse to recognise the pattern: high blood glucose level. She then checks this by means of additional clinical examination: the measurement of blood glucose. Figure 2.1 shows the process of recognising a health problem through pattern recognition. Fig. 2.1  Flow of the recognition of a health problem through pattern recognition

signs symptoms behaviour complaints

CUE

pattern

specific health condition

check / verification of correctness

possibly a verification check asking the patient

48

2  Diagnostic Decision: What Is the Matter with the Patient?

Pattern recognition requires background knowledge and (preferably a great deal of) experience with the symptoms and associated health problems [4]. A risk in pattern recognition is the so-called jumping to conclusions, in which a conclusion is drawn based on too little information and a superficial assessment of the situation. This imperfect way of clinical reasoning increases the risk of errors considerably. There is another way in which nurses can reason from symptoms to possible health problems. For health problems about which the nurse has little knowledge and experience, this is a more appropriate way than pattern recognition. This approach uses a method of organising the information gathered. There are several possible ordering methods. In this book, we use the ICF diagram as an ordering method. Still, background knowledge is always necessary to come to a good classification. The first question in this way of ordering is whether and how the complaints, symptoms and so on can be explained from impairments in body structures or in normal physiology. The next question is how these impairments from normal anatomy or physiology can lead to impairments in physical and/or psychological body functions. To what limitations in activities can all this lead? And to what restrictions in participation? Which phenomenon (behavioural expression, complaint, indication) belongs to which impairment in body structure or body function and to which limitation in activity or to which restriction in participation? Based on the answers to these questions, a list of possible (or hypothetical) health problems emerges. This list is then subdivided into major and minor hypothetical health problems. The nurse then asks herself whether she needs additional information to rule out or confirm the presence of the hypothetical health problems. If additional information is needed or desired, it is collected. The nurse then arrives at a short list of hypothetical health problems. For each hypothetical health problem, the nurse will identify the defining characteristics of that health problem or diagnosis. In the final step, these defining characteristics are tested for presence in the patient. If they are indeed present, the final diagnosis can be made. Figure 2.2 shows a schematic representation of this second way of reasoning from symptoms to health problems.

2.9  How Do I Get Information About the Characteristic Symptoms of the Diagnosis?

49

signs symptoms behaviour complaints

impairments in body structures

impairments in body funtions list of hypothetical health problems ranked form important, through less important, to unimportant

activity limitations

participation restrictions

need for extra information?

narrow down the list of hypothetical health problems

defining characteristics of all health problems?

verification of the presence in the patient of the defining characteristics

establish the diagnosis (or diagnoses)

Fig. 2.2  Reasoning from symptoms to health problems

50

2  Diagnostic Decision: What Is the Matter with the Patient?

2.9 How Do I Get Information About the Characteristic Symptoms of the Diagnosis? A diagnosis can only be made if the specific characteristics that go with that diagnosis are present in the patient. Just as a chair can only be called ‘chair’ if the object in question (1) is a piece of furniture that is intended to offer a seat, (2) this seat is intended for one person and (3) the object stands on one or more legs. These three characteristics must be present for an object to be called a ‘chair’ and not, for example, a ‘sofa’. They are therefore the defining characteristics of chairs. With diagnoses, it works the same way. What distinguishes flu from a cold or sore throat? In the first instance, these are typical clinical symptoms of flu: fever, headache, muscle aches, stuffy nose, cough, general feeling of malaise, weakness and loss of appetite. The diagnosis can be confirmed by a measurement: by means of a smear of the mucous membrane of the nose or throat, the flu virus can be found in the laboratory. This measurement confirms the definitive diagnosis. Flu is a medical diagnosis. The medical classification (ICD) describes the medical diagnoses with their defining characteristics. Similarly, medical psychiatric diagnoses are described in the Diagnostic and Statistical Manual of Mental Disorders (DSM-5). As nursing does not have an internationally recognised manual of validated nursing diagnoses, nursing diagnoses with associated defining characteristics are not easy to look up. The nature of the diagnoses made by nurses (see Sect. 1.6.2) also lends itself poorly to standard diagnosis in manuals. This is because there are so many different nursing diagnoses that each manual can only include a limited number. In addition, customisation (tailoring the nursing diagnosis to the individual patient) is important for a good nursing diagnosis. As a result, the defining characteristics must be based on the professional literature. This may be nursing, medical, psychological or other professional literature. Experienced nurses know the characteristics of common nursing diagnoses by heart within the specialty in which they work (e.g. cardiology or mood disorders). They are therefore able to quickly arrive at correct diagnoses through pattern recognition (Sect. 12.31).

2.10 How Can I Test the Defining Characteristics? The nurse can test the presence of the defining characteristics by going to the patient again and questioning, observing, examining and measuring them. Sometimes, it is preferable for a colleague to perform this review, as it is a more objective check. In Sect. 1.9, the testing of hypotheses about health problems is elaborated and illustrated with examples.

2.11 How Do I Process All This Information into a Diagnosis? Upon completion of the hypothesis testing, the nurse identified two issues:

2.12  What About Diagnoses in the Area of Prevention?

51

1. What’s the problem? 2. How can I tell that this is indeed a problem? It is now possible to substantiate the claim that the patient suffers from problem A, characterised by signs and symptoms 1, 2, 3 and 4. In the next chapter, this sentence will be expanded to include ‘… caused by X, Y and Z’. In the example of the pressure ulcer, stage 2, the diagnostic statement becomes: ‘Pressure ulcer stage 2, on the coccyx, characterised by red peeling skin and a blister, caused by prolonged pressure on the coccyx during bed rest, poor nutritional status and incontinence’. It is customary to prepare the diagnosis in PES format. PES stands for: • Problem • Etiology • Signs and symptoms The diagnostic reasoning has led to identification of the Problem and Signs and symptoms. Etiology is discussed in the next chapter. The diagnosis of pressure ulcers is in PES format: P: E:

S:

Pressure ulcer grade 2, on the coccyx Prolonged pressure on the coccyx during bed rest Poor nutritional condition Urinary incontinence Red, peeled skin Blister

2.12 What About Diagnoses in the Area of Prevention? The diagnostic reasoning of nurses includes the prevention of health problems. The goal of prevention is to prevent a particular health problem from occurring. The problem is not there yet, but there are certain risk factors present that greatly increase the likelihood that the problem will arise. For example, some patients run an increased risk of becoming acutely confused (high risk of delirium). These types of risks also count as nursing diagnoses, at least as long as the risk falls within the nursing professional domain. Because the health problem has not yet occurred, no defining characteristics can be observed. A so-called high-risk diagnosis does not include characteristic symptoms. However, it is possible to determine that there is an increased risk of the health problem. In a patient, older than 70 who has had delirium before and who is being treated with various medicines at the same time, three risk factors are present for the development of delirium. The presence of these risk factors justifies the diagnosis of ‘high risk of delirium’. The risk factors therefore function, as it were, as defining characteristics.

52

2  Diagnostic Decision: What Is the Matter with the Patient?

The risk diagnosis is made in PR format. PR stands for: • Problem (in a high-risk diagnosis, this always starts with ‘high risk of …’) • Risk factors In the example of delirium, the complete diagnosis reads: P: R:

High risk of delirium 73 years old History of delirium Use of five different medications at the same time

For all high-risk diagnoses, the necessary follow-up diagnostic activity is to monitor whether the problem actually occurs or is likely to occur. This makes it possible to identify the health problem at an early stage, as soon as it occurs.

Glossary Activities  Execution of a task or action by an individual. Examples: focusing attention, solving problems, moving around, washing oneself, drinking, doing housework [5]. Autonomous practice  Autonomous practice means that the nurse makes decisions and carries out actions for which she is primarily responsible. In autonomous practice, the nurse has this autonomy, since these actions fit in the nursing professional domain. Background knowledge, background literature  Background knowledge is basic knowledge that nurses in training acquire through their textbooks. In contrast to foreground knowledge, background knowledge involves relatively unchanging knowledge, such as anatomy, physiology and developmental psychology. Body functions  Physiological functions of body systems (including psychological functions). Examples: consciousness, sleep, smell function, maintenance appropriate body weight, muscle power [5]. Body structures  Anatomical parts of the body such as organs, limbs and their components. Examples: anatomical structure of the retina, of the heart, of the hip joint [5]. Case history  Interview between a healthcare professional and the patient, with the aim of mapping out the patient’s history and the relevant symptoms, experiences, behaviour and complaints with regard to the patient’s health problems as well and as completely as possible. Collaborative practice  Participation in treatment or decision-making for which a professional other than the nurse is primarily responsible (also: participatory practice).

2.12  What About Diagnoses in the Area of Prevention?

53

Cue  A prompt that gives a professional direct access to specific information in the memory. This specific information leads to pattern recognition (compare: Kahnemann [6], p. 11). The professional recognises cues through her ­background knowledge and through extensive experience with the patterns to which the cues give access. Defining characteristics (signs and symptoms) Symptoms that are always or almost always present for a specific health problem. Because these symptoms are always present, they characterise the health problem. Signs and symptoms are the most important test criteria when checking whether a specific health problem is present. For example: the defining characteristic of obesity is: a BMI of 30 or more. Diagnosis  A decision that the health professional makes about the patient’s condition. Through that decision, the health professional converts a number of loose data (signs and symptoms) into a conclusion: the health problem is given a name (label). By assigning this name, other possible health problems on the basis of the same signs and symptoms are excluded. The health professional decides that according to her, this diagnosis is the right one and no other. Thus, there is a subjective element in diagnosing: what clinical information has been collected? How has this information been assessed by this health professional? Diagnostic reasoning Clinical reasoning to establish the correct diagnosis or health problem. Emergency history taking Abbreviated form of taking the case history in an emergency situation, in which the most important information for that moment is quickly obtained. Health problem  An illness, disability, complication of illness or disability, reactions to and consequences of illness and disability and the hindrance the patient may experience from (para)medical treatment or reactions to and consequences of (para)medical treatment. Nurses focus on health problems which are within the nursing domain, on collaborative health problems and on multidisciplinary health problems where the nurse has a part in monitoring, deciding, guiding and/ or intervening. High-risk diagnosis  Diagnosis used to describe an increased risk of the patient. An example is ‘High risk of pressure ulcers’. The patient does not yet have pressure ulcers, but risk factors are present, making him or her at significantly higher risk compared to other patients. A high-risk diagnosis is represented in a PR structure, where P means “Problem” and R means “Risk factors”. Impairments  Problems in body functions or in body structures such as a significant deviation or loss. Examples: impairment in pain perception, impairment in absorption of food, intestinal perforation, atrophy of muscle tissue, opacity of the lens of the eye [5]. Limitations  Difficulties an individual may have in executing activities. Examples: limitation in handling stress, in reading, in walking, in washing oneself [5].

54

2  Diagnostic Decision: What Is the Matter with the Patient?

Mechanism of action  The way in which the etiological or prognostic factor brings about the health change in the patient. The mechanism of action explains which and how physiological or psycho-social processes deviate from their normal course and how they affect the health status. Mechanism of change  The way in which the intervention (or the etiological or prognostic factor) brings about the desired change in the patient. For example, many drugs affect chemical processes in the body, and many psychological interventions affect psychological processes. The mechanism of change explains through which physiological, psychological or other process the intervention intervenes and makes changes and how those physiological or psychological changes contribute to the desired effect of the intervention. Monitoring  Repeatedly checking whether the interventions are indeed effective, following the course of the disease, condition and treatment. The purpose of monitoring is to be able to make quick and timely adjustments if complications arise or if the patient’s condition does not improve as expected. Monitoring is a form of diagnostic reasoning. Participation  Involvement in a life situation. Examples: remunerative or non-­ remunerative employment, socialising, recreation and leisure [5]. Participatory practice  Participation in treatment or decision-making for which a professional other than the nurse is primarily responsible (also: collaborative practice). PES  Abbreviation for: Problem—Etiological and related factors—Signs and symptoms. This is the structure to formulate an actual diagnosis. PR  Abbreviation for: Problem—Risk Factors. This is the structure to formulate a high-risk diagnosis. Problem-focused history taking  Form of history taking that focuses on gathering information within a particular problem or problem area. For example, in the case of short patient contacts (day case, outpatient clinic) or because the nurse needs more information on a particular subject. Restriction (participation restriction)  Problems an individual may experience in involvement in life situations. Examples: restrictions in entering into informal social relationships, using money to purchase goods or services, engaging in informally or formally organised sports events [5]. Risk factor  Factor that, if the health problem is not already present, increases the risk of developing that health problem. Self-management  The level of independence (from care professionals) that patients have achieved. This independence relates to making one’s own decisions, being able to plan and carry out activities oneself. The patient does not have to carry out everything himself: even if he can decide when which action is carried out for him and in which way, are these important aspects of self-management.

Cases and Learning Tasks

55

Cases and Learning Tasks Sanne (30 years old) is a single mother with a 5-year-old daughter. She is currently in a long-term relationship. Sanne is a fiscal employee at the tax office, but on sick leave at the moment. Since the age of 16, she has suffered from painful joints due to frequent inflammation: rheumatoid arthritis. Over the years, she has used various painkillers to relieve the pain in her joints. Two weeks ago, Sanne started a new medication to suppress her pain. Mobilising is still very difficult because of the inflammations in her ankles. Since a few weeks, her mother is at home with Sanne for a large part of the day to support her in taking care of Sanne and her daughter. Sanne’s mother has taken care of a large part of the household chores. Sanne and her mother try to do the shopping and to pick up her daughter from school, together. Together they visit the rheumatologist and the physiotherapist every week, as Sanne is unable to drive a car because of the persistent pain. Travelling by public transport is not an option either. Sanne’s friend Berend stays over in the weekends, so he can take care of Sanne. This way Sanne’s mother can go back to her own home for a while to put her affairs in order and to rest herself. When Sanne is alone, she often cries and is angry with herself: what are she and her life worth if she can’t even take care of herself and her little daughter! And all this while she is only 30 and actually still has a whole life ahead of her. She can’t work; her mother still takes care of her, and her boyfriend will probably quit too when he realises that he always has to be there for Sanne, and that it can never be the other way around. Sanne admits to herself that she doesn’t know what to do next, that she has lost control over her life, her illness and her care.

Learning Task 1: Sanne Watch Video 2.1. The case of Sanne contains two main problems, namely ‘loss of control’ and ‘pain’. The worked example on the process worksheet deals with the problem of ‘loss of control’. The learning task is about ‘pain’. 1. Examine the problem of ‘loss of control’ worked out in keywords on the diagnostic reasoning process worksheet. 2. Now elaborate on the health problem ‘pain’ (use a blank process worksheet). Do this in running text so that your diagnostic reasoning is easy for the reader to follow. Use all the information from the written case and film clip that you consider relevant.

56

2  Diagnostic Decision: What Is the Matter with the Patient?

Process Worksheet for the Sanne Case The worked example in the diagnostic reasoning process worksheet deals with the problem of ‘loss of control’.

Process Worksheet Diagnostic Reasoning 1. What objective and subjective symptoms (complaints, signs, behaviours, abnormal value compared to the reference range) are present in the patient? (a) What data can be distilled from observation? • Powerlessness appears from the fact that Sanne does not know what to do on several areas of life. • Frustration appears from the fact that Sanne wants to do activities (including with her daughter), but cannot do them. • Anger is shown by the fact that she lunges a little at her mother. • Negative self-image is evidenced by the fact that she wonders why her boyfriend would want a relationship with someone like her who can’t do anything. (b) What information about symptoms was obtained from the patient? • Sanne indicates to her mother that she doesn’t know what to do next on several levels. Both with her relationship and with her daughter. (c) Are data from already known information available (file, handover, etc.)? • No. 2. How can these symptoms be explained (which abnormalities in anatomy, physiology and psychosocial functioning)? (a) How does this happen in a healthy person? • A 30-year-old is in control of his/her own life and, because of a more positive self-image, is less dependent on others (mother, friend). (b) What is different in this situation? • Sanne has been chronically ill since the age of 16, which has led to an abnormal psychosocial development. • Sanne has always remained in a dependent position. (c) What explanation can be given for this in this situation? • Chronically ill at a young age. • Little progress. • Aberrant development of independence from mother, of self-esteem and of problem-solving abilities. 3. What else would you like to know from or about the patient? Use the answer to this question to supplement the answers to Questions 1 and 2. (a) Are there any missing data that might be less prominent? • How has the disease progressed? • Co-parenting with the father? • Relationship with daughter?

Cases and Learning Tasks

57

4. What is the patient’s hindrance (possibly expressed in the ICF components activities, participation)? (a) How do anatomical characteristics and body functions affect daily functioning? • Not very mobile. • Limited activities with her daughter. • No care for own household. • Can’t work. • Doesn’t come out. • Has few social contacts. 5. How does the patient respond to this (emotional behaviour, [in]effective self-management)? (a) Does the reaction lead to a problem (these are potential Problems in the PES, or potential Etiological factors, or potential signs and symptoms, or potential risk factors, or potential prognostic factors)? • Her reaction to the problem is powerlessness and frustration and passivity. Furthermore, she reacts sadly. Her self-management is limited. These are potential problems or potential signs and symptoms. (b) Does the reaction affect the prognosis (positively or negatively)? • Negative impact on prognosis, as she feels things will only get worse. 6. Which possible health problems does this indicate (from which impairments in body structures, in body functions and limitations in activities and restrictions in participation)? (a) How many and what potential problems are there? • Loss of control. • Need help caring for her child. • Few social contacts. • Limited ability to make sense of things. • Powerlessness, frustration and sadness. • Limited self-management and passivity. (b) Are there problems that are interrelated? • Certainly. The above-mentioned problems are strongly connected. For example: the moment Sanne can take control again and does so, she will be better able to organise the care for her child herself, and she will also see possibilities to make more social contacts. (c) Which problems are important, which less so, and what is the reason for this? • Important: loss of control is the central problem. The other potential problems seem to be etiological factors, signs and symptoms or prognostic factors for loss of control. 7. What are the defining characteristics of the major health problems? (a) What is described about this in the literature? • From the ICF, the defining characteristics are shown in the table below. 8. How can the presence or absence of these defining characteristics be tested? Then test whether these phenomena are actually present. (a) Is what is described in the literature also the case with the patient? • Assessment is possible by further observing Sanne and talking to her.

58

2  Diagnostic Decision: What Is the Matter with the Patient?

Characteristics of  Limitation of  Self-Management/Loss of  Control (on  the Basis of ICF)

Mental functions Impaired mental stability Impaired optimism/ powerlessness Impaired mood: tension, anger, sadness Impaired cognitive flexibility Impaired insight into himself/self-image Impaired judgement

Activities/participation Limitation in focusing attention Limitation in solving problems Limitation in taking decisions

External factors Promoting factors or hindering factors Relationship with her mother (hindering factor) Relationship with her boyfriend Berend Personal attitudes of mother and boyfriend towards Sanne and her condition

Limitation in undertaking and carrying out multiple tasks Limitation in carrying out daily routine actions Limitation in dealing with stress Limitation in bearing responsibilities Restriction in establishing and maintaining parent-child relationships Restriction in entering into and maintaining intimate relationships Restriction in dependency relations Restriction in economic independence

Conclusion P: S:

Loss of control over relationships, health, day-to-day activities Expressions of powerlessness, uncertainty about the future and negative self-image Anger, frustration, sadness, stress Limitation in problem-solving and decision-making Restrictions in dependency relationships and bearing responsibilities

Cases and Learning Tasks

59

Mr Bakker is a 60-year-old man with a history of alcohol abuse. He was divorced 4 years ago, now lives alone and has no children. In the past, he has received help from Jellinek clinic for alcohol addiction (with varying success) for his alcohol dependency, but since the divorce he refuses all help. Regularly a friend visits with whom he ‘drinks a bottle’. In the past year, Mr Bakker has begun to live increasingly in the evening and at night, in bed and sleeping at daytime, although he sleeps just a few hours. Sometimes he doesn’t go to bed at all, but sleeps on the couch in his living room. He often eats only bread or fast food. He hardly ever cooks for himself, and neither does he wash the dishes, as shown by the unwashed plates, cutlery, glasses and pans in his kitchen. Occasionally, his GP visits him because Mr Bakker is experiencing increasing abdominal pain and fatigue, consistent with his cirrhosis of the liver. After the last visit of the GP, the community nurse is called in. The community nurse is asked to assess to what extent Mr Bakker can still take care of himself (with regard to physical care, food, housekeeping etcetera). The next morning, the community nurse visits Mr Bakker and finds him slumped in his chair. He says that he has been feeling nauseous for several hours and has vomited several times. With the vomit he has seen a small amount of (bright red) blood. He says that he has no appetite ‘and by the way, also a stomach ache, so you don’t even want to eat’, and that his trousers are sagging more and more. He indicates that he often feels tired. He smokes a packet of cigarettes a day and drinks 10–12 glasses of wine a day plus sometimes a few glasses of beer. The community nurse sees a skinny man with a somewhat fat belly, who is pale and sweating. He is unshaven, has dirty clothes on and smells unpleasantly. The house is a mess. Mr Bakker is constantly scratching his leg (‘Yes, I fell and have an annoying wound there’) and occasionally his belly. The community nurse notices that the whites of his eyes look slightly yellow.

Learning Task 2: Mr Bakker Watch Video 2.2. The case of Mr Bakker contains several problems, which are at random presented by the community nurse and without any clear connection. 1. What problems do you see here? 2. Choose one of these and work it out in more detail using the basic questions for diagnostic reasoning (using a blank process worksheet). Use all the information in the written case study and film clip that you consider relevant.

60

2  Diagnostic Decision: What Is the Matter with the Patient?

Mr Zach (age 59) was admitted to the nursing home 2 months ago. He has Korsakov’s syndrome. Up until 5  years ago, Mr Zach was doing well. However, in these days he sometimes drank a lot of alcohol, and the family doctor had a serious conversation with him about this. He and his wife drank a few wines or beers every day, and they had a good time together. He occasionally had problems at work (he was an administrative assistant at a large school), for instance, because he was sometimes late for work. But that was always solved. Until 5 years ago, when Mrs Zach passed away. Mr Zach felt sad and abandoned by his wife. They had always remained unwillingly childless, and Mr Zach experienced tremendous loneliness. He told his sparse contacts that he had had a lot of bad luck in his life, and that he missed his wife so much. During this time, he began to drink more and more alcohol and missed many work days. In the first 6 months after the death of his wife, he lost all his contacts because he often drank too much and told the same story over and over again: he had a lot of bad luck in his life and now his wife has also left him. Mr Zach neglected himself more and more, he took bad care of himself, ate unhealthy, smoked a lot, drank more and more. His memory became worse. Eventually he lost his job because of his drinking. After that he stayed at home as much as possible and already drank a beer in the morning. Lately, the neighbours complained about the smell, and because nothing changed, they called the municipality, which called in the social district team. Although Mr Zach usually wouldn’t let them in, it became clear to them that he would soon need more care: neglect and pollution, severe alcohol abuse, falling, trouble with walking, memory problems. A geriatrician diagnosed Mr Zach with Korsakov’s syndrome, and not long after, an admission to the nursing home was completed. Mr Zach has been living in a nursing home for 2 months now. He lives in a ward where more people with Korsakov’s syndrome live. Because of the good care, the healthy meals and medication, Mr Zach improved in the beginning. He still tells everyone that he has had such bad luck in his life and that he misses his wife so much. Occasionally, he still falls, probably in part due to orthostatic hypotension. However, it does not help to tell him that he should not just get up and walk away, because he immediately forgets that. Since a couple of weeks, his behaviour changes. He refuses help and no longer wants his medication. He makes this known loud and clear. He does not take others into account, he curses and is sometimes verbally aggressive. He also looks more and more unkempt. He always wears the same clothes with stains from spilled food, and he argues about everything with the nurses. Last week, after a discussion about his clothing, he took off his shirt, T-shirt, shoes and socks and walked through the nursing home bare-chested and barefoot.

2  Diagnostic Decision: What Is the Matter with the Patient?

61

It’s like he’s unreachable. Sometimes it is possible to make contact, but most of the time it isn’t. Then he sits in a corner by himself, staring into space. His behaviour is so variable and unpredictable that the nurses find it difficult to deal with it. The nurses now want to get a good overview of all the problems. What is going on and what are the priorities?

References 3. WHO. Body mass index - BMI. n.d.. www.euro.who.int. Accessed 7 Jun 2020. 4. Nederlands Huisartsen Genootschap. NHG standaard Obesitas. n.d.. https://www.nhg.org. Accessed 7 Jun 2020. 5. WHO. Classification of diseases (ICD). Geneva: WHO; n.d.. https://www.who.int/classifications/icd/en/. Accessed 7 Jun 2020. 6. Sanders L. Every patient tells a story. New York, NY: Broadway Books; 2009. 7. World Health Organization (WHO). n.d.. https://www.who.int/standards/classifications/ international-­classification-­of-­functioning-­disability-­and-­health. Accessed 2 Aug 2022. 8. Kahneman D. Thinking, fast and slow. London: Allen Lane/Penguin Books; 2011.

3

Etiological Decision: What May Have Caused the Problem?

When I fell hard on my elbow while skating last winter, I became a patient in hospital myself. I had to have surgery, and that happened under complete anaesthesia. In my operating shirt, I was wheeled into the operating room, and immediately I felt very cold. Because the anaesthesiologist had not arrived yet, we had to wait a while. The surgeon had a chat with the OR nurse, and in the meantime I was getting colder and colder. Finally, the anaesthesiologist came in and looking at me he said: ‘Oh, what a stress, you are shivering all over. Are you so scared of the operation or of the narcosis?’ It occurred to me later that I should have said that his clinical reasoning was too superficial; that my problem was not ‘stress, caused by fear of surgery or narcosis, characterized by shivering’, but ‘freezing cold, caused by cool ambient temperature, airy clothing, low body fat, characterized by shivering’; that the preconception that patients are afraid of surgery and narcosis caused him to see the wrong causes and the wrong problems; and that, fortunately, it is not he who deals with these problems, but the nurse. But when I thought of this, the surgery was done, I was lying in the recovery room, with an extra blanket that the nurse had given me. Nice and warm.

3.1 What Is Etiology? What Are Etiological and Related Factors? Etiology is the science of causes. Etiological factors are factors that caused the health problem. Many health problems are not caused by a single cause, but by a number of causes that are present simultaneously: a collection of causes. If, for a

Supplementary Information The online version contains supplementary material available at https://doi.org/10.1007/978-­3-­031-­27069-­7_3. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 J. Dobber et al., Clinical Reasoning and Evidence-Based Practice, https://doi.org/10.1007/978-3-031-27069-7_3

63

64

3  Etiological Decision: What May Have Caused the Problem?

specific health problem, there is a cause among them that is always present, because without that single cause the health problem would not occur, then this cause is called a ‘necessary cause’. Not every problem has a necessary cause. For example, ‘stress’ can not only be caused by fear of surgery or narcosis but also by many other (combinations of) causes. If several causes or combinations of causes can cause the health problem, such a combination is called a ‘sufficient cause’ (Sect. 12.32). Related factors are factors that do not cause the health problem, but that help to maintain a health problem once it is present. Suppose that a 58-year-old woman has been walking with her friend for an hour three times a week for many years. This means that she meets the physical activity standard for people over 55, that is, at least 150 min of moderate intensive exercise per week, spread over various days of the week [1]. One day, this woman becomes seriously ill and is no longer able to go for a walk with her friend. After her recovery, it takes another 6 months before she is physically able to go for a walk again. During her illness, her hiking friend never came to visit her. This annoys the woman so much that she decides to end the friendship and to stop walking altogether. The result is that she has a lifestyle with too little exercise. This problem was originally caused by her illness (the etiological factor), but now that she has recovered from her illness, the problem is perpetuated by the lack of a walking buddy (this is a related factor).

3.2 What Is Etiological Reasoning? Etiological or causal reasoning is reasoning about the causes and related factors of a specific health problem. In this process, the nurse uses a fixed set of questions to systematically identify the causes and the related factors of health problems. Thus, in etiological reasoning, the standard questions focus on determining the causal factors and the related factors. Causal factors are the causes of the problem, while related factors may not have caused the problem, but they do (partly) maintain the problem. Ultimately, these standard questions lead to a deliberate answer to the question: ‘What may have caused the problem?’ 1. Are there possible causes and related factors from diseases and disease processes? Which ones? 2. Are there possible causes and related factors from the (para)medical treatment? Which ones? 3. Are there possible causes and related factors from body structures? Which ones? 4. Are there possible causes and related factors from body functions? Which ones? 5. Are there possible causes and related factors from activities? Which ones? 6. Are there possible causes and related factors from participation? Which ones?

3.3  How Are Etiological Factors Incorporated into the Diagnosis?

65

7. Are there possible causes and related factors from environmental factors? Which ones? 8. Are there possible causes and related factors from personal factors? Which ones? 9. Can these factors together cause or perpetuate this health problem? 10. Are these factors present in the patient? 11. Do these factors have this effect in this patient? (Sect. 12.8) Using the answers to these questions, the nurse can reason about the major causes and related factors in the health problem. Once these are identified, the nurse can use these causes and factors in their prognostic and therapeutic reasoning (Sect. 12.71.2).

3.3 How Are Etiological Factors Incorporated into the Diagnosis? The three elements that make up most diagnoses are: 1. the health problem (Problem) 2. the defining characteristics of that health problem (Signs and symptoms) 3. the causes and related factors for that health problem (Etiology) The first two elements have been discussed in Chap. 2; in this chapter, the third element is concerned. The actual causes and related factors are added to the health problem and the defining characteristics. Basically, it comes down to something happening (Etiology: a fall with the bike or a disagreement with a walking buddy) that leads to a negative change (Problem: severe pain when moving right arm, insufficient physical exercise), and that change is visible in complaints, signs or behaviour (Signs and symptoms: pain and a distorted face when moving right arm, no more walking). In the example on insufficient physical activity, the complete diagnosis is written as follows: Insufficient physical activity, related to the lack of a walking buddy and not wanting to walk alone and to the lack of initiative to find another walking buddy or to engage in substitute physical activity and characterised by insufficient moderate-intensity exercise (60 instead of 150 min/week). In PES format: P: E:

S:

Insufficient physical exercise No walking buddy and do not want to walk alone No initiative to find another hiking buddy No initiative for other physical activity Insufficient moderately intensive exercise (60 instead of 150 min/week)

66

3  Etiological Decision: What May Have Caused the Problem?

3.4 How Should I Reason Etiologically in a High-Risk Diagnosis? With high-risk diagnoses, a specific health problem is not yet present, but the patient is at increased risk of developing the health problem in the future. The patient is at increased risk because certain risk factors are present, which greatly increases the likelihood that the problem will occur. A high-risk diagnosis consists of the health problem for which the increased risk is present and a list of the risk factors that are present in this person. An example is the risk of premature discontinuation of medication use by patients with schizophrenia [2, 3]. About 70% of patients with schizophrenia stop taking medication within 18  months of starting their treatment, although in many cases medication is an important and effective intervention to prevent psychosis. Take, for example, the patient who is stabilised with medication after his third psychosis during an admission to a psychiatric clinic. He is discharged, but does not think he has schizophrenia. He thinks he was confused. In his opinion, this was caused by the circumstances, not by an illness. Now he thinks he’s cured. He doesn’t think that medication contributes to his recovery. ‘You have to do it yourself’, he says. The factors ‘no awareness of illness’, ‘no insight into illness’ and ‘doubt about the effect of medication’ are present in this patient. As a result, he runs a higher risk of stopping his medication (and of becoming psychotic again) than patients who do not have these factors. Because these three factors are known to increase the risk of discontinuing medication prematurely, they are therefore risk factors. The diagnosis in PR format is: P: R:

High risk of discontinuing medication prematurely No sense of disease No insight into illness Doubt about the effect of medication

Instead of etiological factors, the diagnosis therefore includes risk factors. The etiological reasoning in risk diagnoses is therefore about risk factors rather than etiological factors.

3.5 How Do I Find the Etiological Factors? To determine the etiological factors and risk factors for a particular health problem, background knowledge of the health problem is needed. Through the first steps of diagnostic reasoning, the nurse will explore the background of the problem. In many cases, this provides also insight into possible etiological and related factors. However, this is not always sufficient. For example, in the case of high-risk diagnoses, it is not the symptoms but the risk factors that are the primary focus of clinical reasoning. After all, the presence of risk factors leads to the diagnosis that there is

3.7 How Can I Test Whether the Factors Found Do Indeed Cause or Perpetuate…

67

an increased risk of a particular health problem. So, high-risk diagnoses involve background knowledge of risk factors. Nurses learn through experience to know and recognise the health problems and high-risk diagnoses common to their speciality. As with diagnostic reasoning, experienced nurses, in addition to using background knowledge, can already identify the etiological factors, related factors and risk factors of common problems through pattern recognition. Novice and student nurses do not yet have this experience. They will have to search the background literature and the foreground literature (professional journals) for the causes, related factors and risk factors. In many cases, the disease or the disease processes are one of the etiological factors, for example, in the diagnosis of pain. So, it is best to start with the question of whether the disease or disease processes are indeed an etiological factor of the health problem. Furthermore, some of the health problems result from the medical or paramedical treatment of the patient. This may be the case, for example, with the diagnosis of nausea. These are therefore the first two areas in which a search for etiological factors can be made. Using the ICF diagram as a thought model for etiological reasoning, this search can be carried out in a systematic way. For each component of the ICF diagram, the nurse can consider whether there is a possible etiological or related factor. The process worksheet Etiological Reasoning can be found in the in-depth chapter (Chap. 12).

3.6 How Can I Estimate the Importance of the Etiological Factors? It is difficult to determine how much a factor contributes to the occurrence or maintenance of a health problem, although data are available for a number of factors. For example, malnutrition in geriatric patients between the ages of 70 and 86 admitted to hospital with an acute condition (SNAQ score of 2 or higher during their hospitalisation) leads to an additional risk of death within a year of approximately 20% [4]. Thus, for these geriatric patients, malnutrition is an important risk factor. The influence of specific risk factors for a particular health problem can be found in scientific articles (see Part II). However, the magnitude of many factors is not known, and it can only be reasoned whether the etiological, related or risk factor is plausible (Sect. 12.33).

3.7 How Can I Test Whether the Factors Found Do Indeed Cause or Perpetuate the Problem? Testing is done via three questions, which can be answered as a check. This assessment is best done by a colleague, as this is usually a more objective check. The test questions for etiological factors are:

68

3  Etiological Decision: What May Have Caused the Problem?

1. Could this factor indeed be causing the problem? 2. Is this factor indeed present in this patient? 3. Does this factor in this patient indeed have the mechanism of action that could cause the problem?

Glossary Background knowledge, background literature  Background knowledge is basic knowledge that nurses in training acquire through their textbooks. In contrast to foreground knowledge, background knowledge involves relatively unchanging knowledge, such as anatomy, physiology and developmental psychology. Diagnosis  A decision that the health professional makes about the patient’s condition. Through that decision, the health professional converts a number of loose data (signs and symptoms) into a conclusion: the health problem is given a name (label). By assigning this name, other possible health problems on the basis of the same signs and symptoms are excluded. The health professional decides that according to her, this diagnosis is the right one and no other. Thus, there is a subjective element in diagnosing: what clinical information has been collected? How has this information been assessed by this health professional? Etiological factor  Factor that (partly) causes the health problem. Etiological reasoning  Clinical reasoning to establish the correct causes and related factors for the diagnosis or health problem. In case of a high-risk diagnosis, the etiological reasoning concerns the determination of risk factors for that diagnosis. Foreground knowledge, foreground literature Foreground knowledge is relatively rapidly changing knowledge. This knowledge changes as new knowledge and insights are acquired through scientific research. Examples of foreground knowledge are knowledge about the most effective intervention for a particular health condition or about the most important prognostic factors for a particular health problem. High-risk diagnosis  Diagnosis used to describe an increased risk of the patient. An example is ‘increased risk of pressure ulcers’. The patient does not yet have pressure ulcers, but risk factors are present, making him or her at significantly higher risk compared to other patients. A risk diagnosis is represented in a PR structure, where P means ‘Problem’ and R means ‘Risk factors’. Mechanism of action  The way in which the etiological or prognostic factor brings about the health change in the patient. The mechanism of action explains which and how physiological or psycho-social processes deviate from their normal course and how they affect the health status. PES  Abbreviation for: Problem—Etiological and related factors—Signs and symptoms. This is the structure to formulate an actual diagnosis. PR  Abbreviation for: Problem—Risk Factors. This is the structure to formulate a high-risk diagnosis.

Cases and Learning Tasks

69

Related factors  Factors that do not cause the health problem, but which, once the health problem is present, help to maintain the presence of this problem, or factors that did not cause the health problem, but which helped to cause the problem. Risk factor  Factor that, if the health problem is not yet present, increases the risk of developing that health problem.

Cases and Learning Tasks Within the residents group, everyone has to laugh about Simone. When she gets excited, she makes the funniest faces, jokes, dances and gestures and makes everyone laugh. Only after a support worker has told her for the third time to stop, she falls back panting on her regular spot on the couch. Simone is 40 years old and has Down’s syndrome, based on a translocation. She is a small woman, with slanted almond-shaped eyes. She has dry skin, a large tongue, a small lower jaw and a weak muscle tone, so that her tongue often hangs out of her open mouth a little. The transverse hand fold and the toe space between the big and adjacent toe, which are characteristic of the syndrome, are also present in Simone. Since her tenth birthday, she has lived in a community with other mentally handicapped people. Simone can be very funny, but it also happens often that she is not feeling well. In those periods, Simone often won’t eat, won’t wash herself, won’t brush her teeth, won’t go to the day activities, won’t go to bed and so on. Simone has moderate intellectual disability and an IQ of around 45. She does not live independently and needs constant supervision and help with care. During the day, Simone goes to the day care centre, where she does creative activities in a group of six others, and she helps with simple chores like making coffee. A second influence of Down’s syndrome is that Simone has an open ductus Botalli (or: a persistent ductus arteriosus). This means that Simone cannot do much and is tired easily. If she takes a walk, you can see her lips turning a little blue, and you can hear her breathing becoming laboured. Nowadays, children with this heart defect undergo surgery at a young age, but in the past, this was not a matter of course. Simone did not have this surgery. Simone’s living group is part of a large organisation that provides care to people with an intellectual disability. Three weeks ago, the institution moved into newly constructed housing. This was accompanied by a large-scale move and a rearrangement of resident groups and supervisory teams. Simone now lives in an apartment together with two people she already knew from her previous living group or the activity group and with two people she does not know very well. The supervising team is completely new to Simone.

70

3  Etiological Decision: What May Have Caused the Problem?

The supervisory team has to get used to Simone. The report states: ‘It was another difficult day for Simone. Yesterday she was almost unstoppable in her jokes and pranks, today she did not want to do anything and could hardly be tempted to get dressed and eat. Finally, at 11 o’clock she was sitting on the couch, dressed, and it took her half an hour to eat a slice of bread and drink a mug of milk. The senior support worker decides to take plenty of time in the next team meeting to talk about Simone, her background, her needs, and the care and support’.

Learning Task 3: Simone Case Watch Video 3.1. The case of Simone contains two main problems: ‘instability of mood and behaviour’ and ‘limited exercise tolerance’. The worked example deals with ‘mood and behavioural instability’. The learning task deals with the problem of ‘limited exercise tolerance’. 1. Examine the problem ‘instability of mood and behaviour’ worked out in keywords on the Etiological reasoning process worksheet. The process worksheet also includes diagnostic reasoning. 2. Now elaborate on the health problem ‘limited exercise tolerance’ using the standard questions for diagnostic reasoning and the standard questions for etiological reasoning. Do this in running text, so your reasoning (diagnostic and etiological) is easy for the reader to follow. Make sure that you at least investigate the open ductus botalli/persistent ductus arteriosus in the case of dysfunction of anatomical structures and include the consequences in your diagnostic and etiological reasoning. Use all the information from the written case and the film clip that you consider relevant.

Process Worksheet for the Simone case The worked example on the process sheet deals with the problem of ‘instability of mood and behaviour’.

Process Worksheet Diagnostic Reasoning 1. What objective and subjective symptoms (complaints, signs, behaviours, abnormal value compared to the reference range) are present in the patient? (a) What data can be distilled from observation?

Cases and Learning Tasks

2.



3. 4.

5.

6.

71

• Simone is alternately very cheerful and depressed, and this is difficult to influence. • Simone often feels unwell; she finds it hard to get out of bed and struggles to eat her breakfast. (b) What information about symptoms was obtained from the patient? • None. (c) Are data from already known information available (file, handover, etc.)? • Behaviour has been present for a long time. How can these symptoms be explained (which abnormalities in anatomy, physiology and psychosocial functioning)? (a) How does this happen in a healthy person? • Steady mood. (b) What is different in this situation? • Simone has Down’s syndrome. Her mood is also unstable, alternating between being much more cheerful than usual and much more sombre than usual. (c) What explanation can be given for this in this situation? • Simone just moved into a new group. • Limited coping abilities. • Possible onset of Alzheimer’s disease and/or previously undiagnosed mood disorder. What else would you like to know from or about the patient? Use the answer to this question to supplement the answers to Questions 1 and 2. (a) Is there any missing data that might be less prominent? • Diagnostic information on Alzheimer’s disease and/or mood disorder. • Information from her previous residence. What is the patient’s hindrance (possibly expressed in the ICF components activities, participation)? (a) How do anatomical characteristics and body functions affect daily functioning? • Present mood regulation strongly determines her functioning in all areas of life (such as daily activities, relationships with fellow residents). How does the patient respond to this (emotional behaviour, [in]effective self-management)? (a) Does the reaction lead to a problem (these are potential Problems in the PES, or potential Etiological factors, or potential Signs and symptoms, or potential risk factors, or potential prognostic factors)? • Simone’s behaviour is strongly determined by her mood swings. Especially when she is gloomy, this is a problem for Simone. (b) Does the reaction affect the prognosis (positively or negatively)? • Her reaction (she has no control over it herself) has a negative impact on the prognosis. Which possible health problems does this indicate (from which impairments in body structures, in body functions and limitations in activities and restrictions in participation)?

72

3  Etiological Decision: What May Have Caused the Problem?



(a) How many and what problems are there? • Instability in mood and behaviour. • Limited coping abilities. (b) Are there problems that are interrelated? • Strongly connected. (c) Which problems are important, which less so and what is the reason for this? • Instability in mood and behaviour is the main problem for Simone. Limited coping may be a prognostic factor for the course of the problem. 7. What are the defining characteristics of the major health problems? (a) What is described about this in the literature? • Here is a small representation from the literature (ICF): impairment in regulating moods; impairment in adequacy of mood. 8. How can the presence or absence of these defining characteristics be tested? Then test whether these phenomena are actually present. (a) Is what has been described in the literature also the case with the patient? • Assessment is possible by further observing Simone and talking to her. She can also be tested psychologically.

Conclusion P: S:

Instability of mood and behaviour Mood swings

Process Worksheet Etiological Reasoning 1. Are there possible causes/related factors from diseases/disease processes? If so, which ones? (a) Increased risk of developing Alzheimer’s disease from age 40. (b) Increased risk of developing depression. (c) Intellectual disability due to Down’s syndrome. 2. Are there possible causes/related factors from the (para)medical treatment? If so, which ones? (a) No information in the case. 3. Are there possible causes/related factors from body structures? If so, which ones? (a) Unknown, partly dependent on any mood disorder (abnormal neurotransmitter balance) and Alzheimer’s disease (plaque formation in the brain). 4. Are there possible causes/related factors from body functions? If so, which ones? (a) Impaired intellectual functions (moderate intellectual disability).

Cases and Learning Tasks

73

(b) Impaired range of behavioural possibilities (limited behavioural repertoire). 5. Are there possible causes/related factors from activities? If so, which ones? (a) No information in the case. 6. Are there possible causes/related factors from participation? If so, which ones? (a) New co-residents and new support workers after the move. 7. Are there possible causes/related factors from environmental factors? If so, which ones? (a) Recent removal to new housing. Receiving care from a specialised health organisation. 8. Are there possible causes/related factors from personal factors? If so, which ones? (a) No information in the case. 9. Can these factors together indeed cause and/or maintain this health problem? (a) What causes and related factors play an important role? • All of the above factors are strongly correlated. (b) What causes can cause the health problem? • Relocation, with new co-residents and supervisors. • Down’s syndrome, intellectual disability. • Limited behavioural repertoire, unable to structure her own behaviour. (c) What causes and what related factors can perpetuate the health problem? • All of the above causes. 10. Are these factors indeed present in the patient? (a) Yes. 11. Do these factors indeed have this effect in this patient? (a) More information needed from psychological research (diagnostic). Conclusion E:

Relocation, with new co-residents and supervisors Down’s syndrome, moderate intellectual disability, restriction in behavioural repertoire, not able to structure her own behaviour (possible: early onset Alzheimer’s disease) (possible: mood disorder)

74

3  Etiological Decision: What May Have Caused the Problem?

Laila is 24 years old, studies nursing and has a busy social life besides her studies. She is engaged and busy with the preparations for her wedding. This brings along with her studies a lot of stress. For 8 days, Laila has been having very frequent diarrhoea, accompanied by loss of blood. She looks pale and feels lightheaded since a few days. Laila also has painful swollen ankles. Laila has hardly eaten anything lately, and if she eats anything, she has a lot of abdominal pain. Furthermore, she has lost a lot of weight. Laila went to her doctor with these complaints. The doctor thinks it might be Crohn’s disease or ulcerative colitis and refers Laila to a gastroenterologist. Laila has to be admitted and enters the diagnostic phase, for which she has to undergo a number of tests. The gastroenterologist proposes to first take blood to determine her Hb (haemoglobin) level. Her Hb is 4.5 mmol/L. The blood is also tested to see how high her inflammation parameter (CRP) is. This is 250 mg/L. Next, Laila has to undergo a colonoscopy and a gastroscopy so that the gastroenterologist can take a good look at the structure of her gastrointestinal tract. These examinations do not provide enough information for the doctor to make a diagnosis. The doctor suggests that Laila also have a colonoscopy with contrast fluid.

Learning Task 4: Laila Case Watch Video 3.2. The Laila case contains several health problems, including ‘lack of a supportive social network’. Two other main problems are clearly identifiable in the case. 1. Which two problems are clearly identifiable, besides the problem of ‘lack of supportive social network’? 2. Choose one of these and develop it further using the standard questions for diagnostic and etiological reasoning (use blank work sheets Diagnostic reasoning and Etiological reasoning). Use all the information from the written case and the film clip that you consider relevant.

References 3. Gezondheidsraad. Beweegrichtlijnen 2017. Den Haag: Gezondheidsraad; 2017. 4. Higashi K, Medic G, Littlewood KJ, Diez T, Granström O, De Hert M.  Medication adherence in schizophrenia: factors influencing adherence and consequences of nonadherence, a systematic literature review. Therapeut Adv Psychopharmacol. 2013;3:200–18. https://doi. org/10.1177/2045125312474019.

References

75

5. Leucht S, Tardy M, Komossa K, Heres S, Kissling W, Salanti G, et al. Antipsychotic drugs versus placebo for relapse prevention in schizophrenia: a systematic review and meta-analysis. Lancet. 2012;379:2063–71. https://doi.org/10.1016/s0140-­6736(12)60239-­6. 6. Buurman BM, Hoogerduijn JG, De Haan RJ, Abu-Hanna A, Lagaay AM, Verhaar HJ, et al. Geriatric conditions in acutely hospitalized older patients: prevalence and one-year survival, functional decline and cognitive impairment. PLoS One. 2011;6:e26951. https://doi. org/10.1371/journal.pone.0026951.

4

Prognostic Decision: What Can We Achieve?

Today, I took part in a conversation between my work supervisor and Mrs Jacobi, who has a serious lung condition: pulmonary emphysema. To my surprise, Mrs Jacobi told me that she still smokes about 15 cigarettes a day, despite her pulmonary emphysema. She said she really wanted to quit, but that she had only managed to smoke less and not to quit. My supervisor had already discussed this with Mrs Jacobi. She suggested she try the StopStepsPlan, in combination with nicotine patches and supportive conversations to make a new attempt to quit. Mrs Jacobi was willing to try this. A little later, my supervisor asked me to draw up a prognosis and formulate goals for Mrs Jacobi. She added that I should take prognostic factors into account, and then she was gone. But how do I know what the prognosis is? I can’t see into the future, can I? And what are prognostic factors anyway? I later told my supervisor that I would work on it tomorrow.

4.1 What Is a Prognosis? What Is Prognostic Reasoning? A prognosis is a prediction of how a particular health problem is likely to turn out. Will the health problem disappear and the patient recover completely? Or will the problem improve but always remain present? Or will the problem only get worse and the patient therefore deteriorate? It is also possible that the problem will disappear temporarily and then reappear. Or that the problem remains stable: neither improving nor deteriorating. In short, a health problem can end in many different ways.

Supplementary Information The online version contains supplementary material available at https://doi.org/10.1007/978-­3-­031-­27069-­7_4. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 J. Dobber et al., Clinical Reasoning and Evidence-Based Practice, https://doi.org/10.1007/978-3-031-27069-7_4

77

78

4  Prognostic Decision: What Can We Achieve?

Prognostic reasoning is clinical reasoning to find the factors that influence the course of a health problem, weigh them and determine the most likely outcome. Usually, the prognosis is further elaborated in patient goals. Moreover, the prognosis is about the probable outcome. No certainty can be given, because there are almost always individual differences between patients with health problems. It is usually difficult to determine in advance how these differences will develop. For this reason, prognostic statements are always accompanied by a significant degree of uncertainty. This uncertainty is also reflected in the words that are used in healthcare when discussing the prognosis with the patient. These include phrases such as ‘We think that …’, ‘Probably …’, ‘There is a good chance …’ and ‘Within a few weeks …’. If possible, it is preferable to replace these subjective terms with more objective statements. For example, ‘Within a number of weeks …’ could be replaced by: ‘Within 2–6 weeks’ and: ‘There is a high probability that …’ with: ‘There is an 80% chance that …’. In prognostic reasoning, the standard questions focus on the prognosis: how will the health problem progress? Using the standard questions, the nurse considers the prognostic factors that influence the course of the health problem and what maximum result can be expected. Using the standard questions, the nurse comes to an informed decision on the patient’s care goal for a specific health problem and thus answers the question, “What do we think we can achieve?”. 1. Which physical, psychological and personal factors (including causal and related factors) influence the course of the health problem negatively, and which positively? 2. Which environmental factors influence the course of the health problem negatively, and which positively? 3. On what mechanism of action (per factor) is that influence based? 4. All those factors taken together, is their influence positive, neutral or negative? 5. What is the maximum achievable result for this health problem? (Sect. 12.9) The answers to these questions lead to the reasoning and formulation of a care goal for the patient’s health problem.

4.2 Why Is a Prognosis Important? How something really turns out can only be determined in retrospect. Nevertheless, the probable course of the health problem is thought about in advance. This has two important advantages. First, a distinction can be made between the probable outcome of a health problem when no intervention is carried out and when one or more interventions are carried out. If the differences in prognosis are known, then the patient and healthcare professional can together better determine whether an intervention is required or not. After all, if the prognosis without intervention is good anyway, and the patient can be expected to improve by himself, no intervention is probably

4.5  How Do I Know What Factors Influence the Prognosis?

79

necessary. If the prognosis is in any case poor, both without and with intervention, the question is whether it is desirable to carry out a stressful intervention. In situations where an intervention is required, the type of intervention (or combination of interventions) can be discussed with the patient. Secondly, the factors influencing the course of the health problem (or prognostic factors) can be mapped out. If it is known which factors influence the course positively or negatively, an attempt can be made to limit the influence of negative influencing factors and to strengthen the influence of positive factors.

4.3 Should Prognosis Be Reasoned About First and Then Therapy, or Vice Versa? It is not possible to strictly separate prognostic reasoning and therapeutic reasoning. After all, whether or not an effective intervention is available can have a major influence on the prognosis. This also applies to the availability of interventions to influence prognostic factors. The prognosis for both alternatives must also be known if the patient is to be given the choice between carrying out an intervention and doing nothing (waiting to see how the health problem develops). Conversely, it is also good to know the prognosis of the health problem first, without any intervention. This can be used to determine whether it is worthwhile to carry out an intervention. Furthermore, it is necessary to first identify which prognostic factors are likely to play a role. This makes it possible to search for interventions that strengthen positive prognostic factors and limit the influence of negative prognostic factors.

4.4 What Is the Difference Between Risk Factors, Etiological Factors and Prognostic Factors? A prognostic factor influences the health problem if the health problem already exists. A risk factor and an etiological factor influence the occurrence of the health problem. In many cases, etiological factors are also negative prognostic factors, but this is certainly not always the case. For example, the fact that the patient is of the ‘male gender’ is a risk factor for the occurrence of a myocardial infarction, but once a myocardial infarction has occurred, women are at higher risk of recurrence. In that case, ‘female gender’ is a negative prognostic factor for the occurrence of another myocardial infarction.

4.5 How Do I Know What Factors Influence the Prognosis? Before prognostic factors can be sought, it must first be determined in which outcome the nurse is interested. After all, a patient has different prognostic factors for different problems (e.g. the prognostic factors for healing an amputation wound are

80

4  Prognostic Decision: What Can We Achieve?

not the same as those for self-management). And the prognosis for a medical health problem may differ from the prognosis for an important and burdensome symptom of that problem (e.g. pain). The consequences (e.g. limitation in dealing with stress) can also have different prognoses. In many cases, ‘quality of life’ is an important outcome. The prognosis for this outcome can also differ from the prognosis for the other outcomes. The most important source for prognostic factors and the extent of their influence is the foreground literature from scientific research. This foreground literature can be found in professional journals in which scientific research results are published (see Part II). In addition, the most important prognostic factors for some health problems are already known. These factors can be found as background knowledge in the background literature. Sometimes, no scientific knowledge is available yet about a particular prognostic factor. In such cases, it can be reasoned whether this factor probably influences the prognosis. In order to do so, it is necessary to find out the mechanism of action of that factor. For example, smoking is a risk factor for the development of cardiovascular disease [1–4]. It is also clear that smoking after myocardial infarction is a negative prognostic factor for the recurrence of myocardial infarction [5, 6]. But let us pretend that this is not known and that we want to know whether smoking is a prognostic factor for a second myocardial infarction and that there is no scientific knowledge about this yet. In order to reason whether smoking is a plausible prognostic factor, we explore the mechanism of action of smoking. To do this, we examine how tobacco smoke affects the blood supply to the heart. One of the substances in tobacco smoke is nicotine. Nicotine causes a temporary narrowing of the arteries. Furthermore, tobacco smoke damages the artery wall, allowing cholesterol to attach itself to the artery wall. If one or both of these phenomena cause a coronary artery to become occluded, the blood supply, and therefore the oxygen supply, of the heart is blocked, resulting in a heart attack. This is, simply stated, the mechanism of action of smoking [1]. Based on this we can reason that smoking indeed increases the risk of a second heart attack and is thus a plausible negative prognostic factor. There are various types of prognostic factors. An important distinction is the distinction between prognostic factors that can be influenced and those that cannot. For example, for many health problems, age is a prognostic factor that cannot be influenced, whereas lifestyle is a prognostic factor that can be influenced. Another classification is according to (a) physical factors (such as body weight, abdominal size and stamina), (b) psychological factors (such as concentration capacity, stress and stress management) and (c) environmental factors (such as health insurance coverage, quality of the housing and social support). This last classification can also be viewed with the ICF diagram in mind. The physical and psychological factors then refer to body structures, body functions and personal factors, while the environmental factors are called the same in the ICF (Sect. 12.71.3) [6].

4.8  How Can I Formulate Patient Goals from a Prognosis?

81

4.6 What Should I Do Once I Have Identified the Prognostic Factors? The prognostic factors can be divided into positive and negative factors. In addition, some factors have a great deal of influence, and other factors hardly at all. Some factors can be influenced, and others cannot. Using this mix of data, the nurse must arrive at a statement about the prognosis. In doing so, she must first consider what the most important prognostic factors will be for the individual patient. If the nurse has a great deal of experience with the health problems in question, she includes this experience in her reasoning. The next question is whether these factors can be easily influenced. Using the results of this reasoning, the nurse weighs the positive and negative factors and comes to a conclusion about the likely and maximum feasible outcome.

4.7 Which Types of Prognosis Are Possible? The conclusion about the probable outcome is reflected in one of six types of prognosis: 1. prevention 2. resolution 3. improvement or remission 4. delay or minimisation 5. stabilisation 6. deterioration-palliation These types of prognosis can be seen as six types of results [7, 8]. Thus, the estimate of prognosis is the estimation which prognosis type probably suits the outcome of the health problem best. It is based on the inventory of prognostic factors present and on the consideration of the magnitude of their influence on the health problem (Sect. 12.34).

4.8 How Can I Formulate Patient Goals from a Prognosis? Patient goals are goals that articulate what the nurses want to achieve through nursing care. These goals are consistent with the health problem and prognosis. Matching the prognosis means that the patient goal must fit the type of prognosis previously established. For example, if the type of prognosis is stabilisation, the goal targeted by nursing care cannot articulate improvement or recovery from the health problem. If the health problem has previously been formulated as a nursing diagnosis, the nurse uses the PES or PR structure when formulating the goal. This means that in most cases, the goal is focused on what is described in the problem (P). This is

82 Fig. 4.1 Relationship between the elements of nursing diagnosis and the goals

4  Prognostic Decision: What Can We Achieve?

P

goal

intermediate goal E

symptom reducon S Fig. 4.2 Relationship between the elements of the high-risk diagnosis and the goals

P

prevenon

goal

intermediate goal R

because the core of the health problem is described briefly and powerfully in that problem. The etiological and related factors (E) are in most cases the starting point for the interventions (see Chap. 5). Final and interim goals are often formulated. An intermediate goal can be aimed at one or more etiological factors, for example, if the intermediate result is that the etiological factor is no longer present. Ultimately, however, the disappearance of an etiological factor is not a goal in itself, as the focus is on the health problem. In a number of cases, the goal is focused on the defining characteristics (S). After all, if one cannot solve the problem, improve it or allow it to go into remission, reducing the patient’s burden of symptoms might be the highest achievable goal. In this case, the nurse deliberately chooses symptom management. Figure  4.1 shows the relationship between the elements of the nursing diagnosis and the goals in diagrammatic form. In the case of a high-risk diagnosis, the diagram shown in Fig. 4.2 applies.

4.9 How Do I Formulate a Patient Goal? The patient objective must be formulated in such a way that the benefit to the patient is immediately clear. When the patient hears the objective, he should in fact immediately say: ‘Yes, I want that’. Of course, the patient’s health problems and prognosis should be discussed with him beforehand, so that the patient agrees with the goal, and the goal does not come as a surprise to him (Sect. 12.35). Many healthcare organisations have a rule that the goals must be formulated in SMART terms. ‘SMART’ stands for Specific, Measurable, Acceptable, Realistic and Time-related.

4.10  How Can I Use the Goals in Evaluation?

83

• Specific. The goal must be formulated in such a way that it is immediately clear to everyone what is to be achieved, by including concrete behaviours of the patient in the goal. For example, for an immobile patient, the goal: ‘Mr. X can move around’ can be made more specific by adding the method of movement: ‘Mr. X can move around with the aid of a walker’. • Measurable. It must be clear when the goal is achieved. The goal of Mr. X can be made measurable by including a minimum performance: ‘Mr. X can move at least 100 m with the help of his walker’. • Acceptable. The patient must find the goal useful. For example, Mr. X may remark that moving 100 m is of no use to him because the supermarket is 250 m from his home. The goal should then be: ‘Mr. X can move at least 300 m using his walker’. Depending on the setting it can be: ‘Mr. X can use his walker to do his shopping independently’. This converts the ability to move into a goal that goes further than just moving. • Realistic. The goal must be achievable. Perhaps Mr. X would prefer to walk without a walker again, but if that is not feasible, it should not be set as a goal. Unattainable goals give the patient false hope and lead to frustration for all concerned. • Time-related. It matters whether the goal is achieved within a week or within 3 months. Therefore, a time limit is included in the goal. In a hospital setting, a good SMART goal might be: ‘Mr. X will be able to walk from his chair to the bathroom with the aid of a walker within a week’. Within a rehabilitation setting, one can build on that with: ‘Mr. X will be able to move independently within the entire rehabilitation unit with the aid of a walker within a month.’

4.10 How Can I Use the Goals in Evaluation? The assessment of the goals takes place during their evaluation. The goals are set up in such a way (i.e. measurable) that they are easy to test. After 1 week, the patient and the nurse can assess whether Mr. X can walk to the bathroom using his walker. If Mr. X is still unable to do so, they can determine to what extent he can move around with the walker and whether or not this was a good (achievable) goal for him. In the rehabilitation centre, the nurse may choose not to wait a month to evaluate. After all, if a goal is set too high or too low, or if the intervention does not work, it is important to adjust the plans in time. This means that the nurse should continuously monitor the effect of the intervention (see Sect. 2.4). This form of diagnostic reasoning leads to ongoing evaluation and adjustment where necessary. Monitoring and timely adjustment of goals and interventions are important in all settings, even if the goal is to be achieved within 1 week or less. This does not detract from the fact that formal evaluation is desirable to determine the results achieved by nurses in patient care. After all, they are responsible for a significant proportion of the care provided and therefore need to make their performance transparent. In doing so, nurses take responsibility for their part in patient care.

84

4  Prognostic Decision: What Can We Achieve?

Glossary Background knowledge, background literature  Background knowledge is basic knowledge that nurses in training acquire through their textbooks. In contrast to foreground knowledge, background knowledge involves relatively unchanging knowledge, such as anatomy, physiology and developmental psychology. Body functions  Physiological functions of body systems (including psychological functions). Examples: consciousness, sleep, smell function, maintenance appropriate body weight, muscle power [6]. Body structures  Anatomical parts of the body such as organs, limbs and their components. Examples: anatomical structure of the retina, of the heart, of the hip joint [6]. Environmental factors  The physical, social and attitudinal environment in which people live and conduct their lives. Examples: medicines, glasses, educational toys, money, climate, family, social norms, housing, presence or absence of healthcare facilities [6]. Etiological factor  Factor that (partly) causes the health problem. Foreground knowledge, foreground literature Foreground knowledge is relatively rapidly changing knowledge. This knowledge changes as new knowledge and insights are acquired through scientific research. Examples of foreground knowledge are knowledge about the most effective intervention for a particular health condition or about the most important prognostic factors for a particular health problem. Mechanism of action  The way in which the etiological or prognostic factor brings about the health change in the patient. The mechanism of action explains which and how physiological or psycho-social processes deviate from their normal course and how they affect the health status. Monitoring  Repeatedly checking whether the interventions are indeed effective, following the course of the disease, condition and treatment. The purpose of monitoring is to be able to make quick and timely adjustments if complications arise or if the patient’s condition does not improve as expected. Monitoring is a form of diagnostic reasoning. Outcome  Outcome refers to that which is influenced through an intervention. Usually an outcome is stated in one or two words, for example ‘weight’, or ‘social integration’. Please note that there is a difference between ‘outcome’ and ‘goal’ and ‘result of the intervention’. The outcome indicates in general terms what is being worked on, the goal is the intended result in terms of patient behaviour, whereas the result of the intervention indicates precisely what the result of that intervention is for the patient. If the outcome is ‘weight’, the goal may be ‘Mr X weighs between 75 and 78 kg.’, the result may be ‘gained 3 kg’. Personal factors  A person’s individual background. Examples: age, gender, social status, life events [6]. PES  Abbreviation for: Problem—Etiological and related factors—Signs and symptoms. This is the structure to formulate an actual diagnosis.

Cases and Learning Tasks

85

PR  Abbreviation for: Problem—Risk Factors. This is the structure to formulate a high-risk diagnosis. Prognosis  Prediction of how a particular health problem is likely to end. Roughly speaking, the prognoses can be sorted in six types of prognosis: prevention, resolution, improvement or remission, delay or minimisation, stabilisation, deterioration-­palliation [7, 8]. Prognostic factor  Factor that, if the health problem is already present, positively or negatively influences its course and outcome and in this way also influences the health outcome(s). Prognostic reasoning  Clinical reasoning to find the prognostic factors in a health problem, weigh them and determine the prognosis. Prognostic reasoning also includes the goal setting for patient care. Related factors  Factors that do not cause the health problem, but which, once the health problem is present, help to maintain the presence of this problem, or factors that did not cause the health problem, but which helped to cause the problem. Risk factor  Factor that, if the health problem is not yet present, increases the risk of developing that health problem. Therapeutic reasoning  Clinical reasoning about the expected effect of interventions and about the appropriateness of interventions, in order to select the most appropriate intervention(s) for the patient and to choose them together with the patient. Many interventions have both desired and undesired effects. In therapeutic reasoning, both types of effect are included in the reasoning.

Cases and Learning Tasks Mrs Kimanski, a 54-year-old woman of Polish origin, lives in a medium-sized town in the west of the country. She has been a widow for a long time, since her husband died of leukaemia when their son was only 2 years old. This son Dawid has recently moved to a university campus, to study social geography. Mrs Kimanski would have preferred that he had lived closer to home, but her joy at him having been accepted, and her pride in her son, prevailed when it was announced that he could go to university. Mrs Kimanski worked for decades at a large department store, in the drapery department. However, the department store closed its doors due to bankruptcy and most of the employees were made redundant. This was a great loss for her as her work meant a lot to her. She lost not only the security of a steady income and a large part of her circle of acquaintances, but also the feeling that she could make a meaningful contribution to society. Her family lives in Poland, and the contact is actually diluted. In addition, she feels uncertain whether she will be able to support her son financially in the current situation, and she is very worried about this.

86

4  Prognostic Decision: What Can We Achieve?

Mrs Kimanski has been unemployed for half a year now. She is lying in bed more often and for longer periods. She does not take the trouble to use makeup or put on her jewellery anymore, while before she was always neatly groomed. She worries a lot, feels tense and doesn’t sleep well. Because she has little appetite, she does not bother to cook for herself. She is reluctant to go out and do the shopping, and so she isolates herself more and more. Dawid, who regularly comes home at the weekends, sees this happening, and he worries about his mother, because this is not how he recognises her. He decides to contact his family doctor who starts her on antidepressants (medication against depression) and discusses the importance of eating well and going outside every day. However, so far these do not have the desired effect. Mrs Kimanski is becoming increasingly neglectful, and it is suspected that she is not taking her medication. When asked about this by her son, she does not give a clear answer and says, ‘I will never get a new job, my life is useless.’ The GP calls in the community nurse, among other things, to monitor the use of medication twice a day.

The role of the community nurse consists of: • supervising the taking of medication and explaining to Mrs Kimanski the importance of taking it • establishing and maintaining a daily structure • attention to sleep behaviour • food, drink, self-care • activate (also outdoors) • stimulate social contacts • monitoring of mood, with attention for possible suicidal thoughts • attention to life fulfilment, meaning

Learning Task 5: Mrs Kimanski Case Watch Video 4.1. The film clip deals with the problem of ‘non-compliance’. This problem is presented in detail on the process worksheet. The learning task is about one of Mrs Kimanski’s other problems, about which more information can be found in the written case. 1. Examine Mrs Kimanski’s problem of ‘non-compliance’, worked out in keywords on the Prognostic reasoning process worksheet, which also includes the diagnostic and etiological reasoning.

Cases and Learning Tasks

87

2. Now elaborate on one of Mrs Kimanski’s other problems. Do this in running text, so your reasoning (diagnostic, etiological and prognostic) is easy for the reader to follow. Use all the information in the written case study and possibly the film clip that you consider relevant.

Process Worksheet for the Mrs Kimanski Case The worked example on the process sheet deals with the problem of ‘non-compliance’.

Process Worksheet Diagnostic Reasoning 1. What objective and subjective symptoms (complaints, signs, behaviours, abnormal value compared to the reference range) are present in the patient? (a) What data can be distilled from observation? Son Dawid observed the following: • His mother is in bed a lot and has little appetite. When he’s home, she doesn’t bother to cook. He thinks she doesn’t cook when she’s alone either. He has also noticed that his mother does not take care of herself as neatly as he is used to. He also thinks she has hardly any contacts besides him. Finally, he says he is shocked when his mother tells him that her life has become useless because she thinks she will never have a job again. He is glad that the GP prescribed antidepressant medication, but he suspects that his mother no longer takes this medication. (b) What information was obtained from the patient? • Mrs Kimanski avoids questions about taking her medication. (c) Are data from already known information available (file, handover, etc.)? • Son Dawid suspects that his mother is not taking her medication. The family doctor also suspects this. 2. How can these symptoms be explained (which abnormalities in anatomy, physiology and psychosocial functioning)? (a) How does this happen in a healthy person? • If people are prescribed medication, healthcare professionals expect the medication to be used as prescribed. If the patient does not want to use the medication, then the healthcare professionals expect the patient to discuss this with them. (b) What is different in this situation? • She does not want to talk about the medication, so it remains unclear whether she is taking it or not. (c) What explanation can be given for this in this situation? • There is no clear explanation.

88

4  Prognostic Decision: What Can We Achieve?

3. What else would you like to know from or about the patient? Use the answer to this question to supplement the answers to Questions 1 and 2. (a) Is there any missing data that might be less prominent? • It is unclear what the situation is with the medication. What did the doctor discuss with her about this? What does she think about the medication? What does she think about her health situation? Does she feel she needs help and even medication? 4. What is the patient’s hindrance (possibly expressed in the ICF components activities, participation)? (a) How do anatomical characteristics and body functions affect daily functioning? • Mrs Kimanski hardly goes out anymore. • Mrs Kimanski does (almost) no more shopping and does not cook anymore. • Mrs Kimanski is in bed for a long time. • Mrs Kimanski has few social contacts. The contact with her family has been diluted, and she does not speak to former colleagues anymore. • Mrs Kimanski no longer participates in working life because she was laid off after a bankruptcy. 5. How does the patient respond to this (emotional behaviour, [in]effective self-management)? (a) Does the reaction lead to a problem (these are potential Problems in the PES, or potential Etiological factors, or potential Signs and symptoms, or potential risk factors, or potential prognostic factors)? • Mrs Kimanski responds by avoiding the topic of medication. There is ineffective self-management in that she has probably stopped taking her medication, without discussing this with the GP. (b) Does the reaction affect the prognosis (positively or negatively)? • The impact of this response on the prognosis appears to be negative. 6. Which possible health problems does this indicate (from which impairments in body structures, in body functions and limitations in activities and restrictions in participation)? (a) How many and what problems are there? • Several problems are involved, here only the possible ‘therapy non-­ compliance’ is elaborated. (b) Are there problems that are interrelated? • That’s certainly the case. It may be that if Mrs Kimanski takes her antidepressants as prescribed, this will lead to a reduction in the other problems. (c) Which problems are important, which less so and what is the reason for this? • Therapy non-compliance is important. It can have an important contribution in fighting depression. 7. What are the defining characteristics of the major health problems? (a) What is described about this in the literature? • The defining characteristic of medication non-compliance is: not using the prescribed medication as directed.

Cases and Learning Tasks

89

• Furthermore, there is no reduction in depressive symptoms, but rather exacerbation. 8. How can the presence or absence of these defining characteristics be tested? Then test whether these phenomena are actually present. (a) Is what has been described in the literature also the case with the patient? • Engage with Mrs Kimanski about her medication in a supportive, non-­ judgemental manner so that Mrs feels free to tell her what she thinks about it. • Checking the packaging: has as much medication been taken out as is appropriate since the medication was prescribed? • It may also be the case that the medication is taken, but is not effective (‘fails to take effect’).

Conclusion The community nurse met with Mrs Kimanski to discuss the medication. Mrs Kimanski said that she had tried the medication for a few days, but that she did not notice that it helped. She then stopped taking it. When her son inquired with her, she was afraid he would get angry with her. Since then, he has flushed the daily dose down the toilet to make it look like she did take the medication. The nurse confirms that it is unfortunate that she has not noticed any effect after a few days. She explains to Mrs Kimanski that it often happens that medication has to be taken longer before the effect is noticeable. P: S:

Medication non-compliance Tells she is not taking het medication

Process Worksheet Etiological Reasoning 1. Are there possible causes/related factors from diseases/disease processes? If so, which ones? (a) Depression. 2. Are there possible causes/related factors from the (para)medical treatment? If so, which ones? (a) It is unclear to what extent the GP discussed the medication and its effects with Mrs Kimanski. 3. Are there possible causes/related factors from body structures? If so, which ones? (a) Not applicable. 4. Are there possible causes/related factors from body functions? If so, which ones? (a) Not applicable.

90

4  Prognostic Decision: What Can We Achieve?

5. Are there possible causes/related factors from activities? If so, which ones? (a) Not applicable. 6. Are there possible causes/related factors from participation? If so, which ones? (a) Not applicable. 7. Are there any possible causes/related factors from environmental factors? If so, which ones? (a) Mrs Kimanski lives alone, there is no one with whom she can discuss this during the week and no one to support her in taking her medication. 8. Are there possible causes/related factors from personal factors? If so, which ones? (a) Possibly Mrs Kimanski thinks that medication is not a good solution for her problems, not fitting with her views on health and disease. Furthermore, Mrs Kimanski has unrealistic expectations about the speed with which the medication could bring improvement. 9. Can these factors together indeed cause and/or maintain this health problem? (a) What causes and what related factors can perpetuate the health problem? • The depression in combination with the view on the effect of medication, and probably a lack of knowledge about antidepressant medication. Another factor is that she lives alone without support from others with whom she can have a confidential conversation. (b) Can these causes cause the health problem? • Yes. (c) What causes and what related factors can perpetuate the health problem? • The depression and her views on this medication. 10. Are these factors indeed present in the patient? (a) Yes, these factors are present in the case of Mrs Kimanski. 11. Do these factors indeed have this effect in this patient? (a) Yes.

Conclusion E

Depression Views on medication, possibly in combination with a lack of knowledge about this No daily support from a trusted person

Cases and Learning Tasks

91

Process Worksheet Prognostic Reasoning 1. Which physical, psychological and personal factors (including causal and related factors) negatively influence the course of the health problem? Which ones are positive? (a) For a realistic estimation of the optimal achievable result, have the health-­ promoting and -limiting factors been mapped out? • The depression affects her thinking negatively (negative). • Her views on (the effect of) medication are not realistic (negative). 2. Which environmental factors influence the course of the health problem negatively? Which ones are positive? (a) Mrs Kimanski lives alone (negative). (b) Mrs Kimanski has an involved son (positive). (c) Community nursing has now been called in (positive). 3. On what mechanism of action (per factor) is that influence based? (a) How big is the influence of that factor? • Depression is a powerful factor. The content of her thinking about herself is negative, and she finds her life useless. It is unclear whether this prevents Mrs Kimanski from taking her medication. • Cognitions on medication. How a person thinks about the usefulness of medication and about the place of medication in her opinions about health and illness can strongly influence the will to use medication for a longer period of time. • The consequences of living alone for medication therapy are that no one can control the medication intake and no one supports Mrs Kimanski to keep taking her medication even if she doesn’t notice any effect yet. Furthermore, there is no one to help remind her of the reasons why she is taking medication. These are important influences. • The son concerned and the community nurse can together provide daily contact. This organises control, but there are also discussion partners for Mrs. who can talk with her about the ‘how and why’ of the medication. (b) To what extent can the factor itself be influenced? • The depression can be influenced by the medication, but possibly also by having people over on a daily basis who can help Mrs Kimanski become more active. • Often, cognitions on medication are difficult to influence. It strongly depends on the opinions themselves and the strength of these opinions. Sometimes a lot can be achieved through education. • Living alone cannot be influenced, but the control of medication use can be well organised by the community nurse. • Daily contact has already been established. 4. Do those factors add up to positive, neutral or negative? (a) Has it been reasoned how the various factors will interact with each other and with the health problem?

92

4  Prognostic Decision: What Can We Achieve?

• Adequate care and daily support for Mrs Kimanski are now in place (community nurse and son). This weighs heavily as they can also influence the negative prognostic factors ‘cognitions on medication’ and ‘possible lack of knowledge’. 5. What is the maximum achievable result for this health problem? (a) Is the maximum achievable result prevention, resolution, improvement or remission, delay or minimisation, stabilisation, deterioration-palliation? • The maximum achievable is the resolution of the problem. There is a clear reason for the depression, so if this will to be treated, adherence is likely to improve.

Conclusion Type of prognosis: resolution. Mr Ewijk is 81 years old, and after a fall from a ladder with a humerus fracture, he is admitted to the Orthopedic Department. He will have surgery on his arm. Mr Ewijk lives independently (his wife passed away a few years ago), he has two daughters with whom he has a good contact. Every Sunday, one of the daughters comes to pick up her father for a walk and to have dinner together. He cooks for himself every day. He likes to be active. Because he is very handy, he has become the ‘handyman’ of the neighbourhood. He maintains a number of gardens and does repair work. The odd jobs give him a sense of belonging and social contact. However, while pruning an old tree, things went wrong: he fell from the ladder and broke his arm. When he is admitted to hospital, Mr Ewijk’s orientation and memory are good. The anamnesis interview revealed that he had hypertension and angina pectoris complaints. There are no cognitive disorders known. Mr Ewijk does suffer from recurrent urinary tract infections, and shortly before admission, he had a urinary tract infection. This infection was treated with antibiotics. The first night after surgery, a fellow patient calls the nurse, and she finds Mr Ewijk next to his bed. His catheter is on the floor, he is tangled up in the IV line and he no longer recognises the nurse. He is restless, looks around, starts a sentence, does not finish it and starts another sentence, which he also does not finish. Mr Ewijk makes slapping motions at the nurse and then suddenly says clearly, ‘Go away, I have to go, I have to go’. Then he points at the wall and asks why there are little flies everywhere on that wall. It seems as if Mr Ewijk does not hear what the nurse is saying to him, he does not react anymore. The night shift staff warns the doctor on duty, who starts medication. The nurse on duty in the morning will now meet with the ward physician to discuss Mr Ewijk’s situation in order to make further decisions.

Cases and Learning Tasks

93

Learning Task 6: Mr Ewijk Case Watch Video 4.2. In the case of Mr Ewijk, the film clip shows a consultation between a doctor and a nurse. Two main problems are discussed. Elaborate on both of them. Use all relevant information from the written case and the film clip. 1. What problems do you see here? Elaborate on these problems using the basic questions for diagnostic, etiological and prognostic reasoning. Use all information from the written case and the film clip that you consider relevant. Mrs Thurston (69 years old) lives alone in an apartment on the seventh floor of a large apartment building. She moved here together with her husband 3 years ago. A year later, Mr Thurston suddenly died of a heart attack. She feels alone sometimes, but has a lot of support from her two children and enjoys her three grandchildren immensely. Mrs Thurston was recently hit by a car while riding her electric bicycle. As a result, she has a broken hip, wrist and ribs on her right side. After an admission in hospital, she was admitted a week ago for further rehabilitation at a geriatric rehabilitation ward. She has a double room there, together with another lady and the ladies have to share a washbasin. During the history taking at the rehabilitation department, Mrs Thurston said she would like to go home as soon as possible. ‘I am much too young to be admitted at a geriatric rehabilitation ward. It is not necessary either, I can do everything myself. But the doctor and the children want me to come here first.’ She tells that she ‘wants to do everything herself’, to prove that she can go home. She was shocked when she heard that she would have to share a room with another patient. ‘I didn’t count on that. What if we don’t get along? And if she snores at night?’ Mrs Thurston’s right arm will have to remain in plaster for another 5 weeks. She has difficulty coughing and breathing because of the fractures in her ribs, and walking with her new hip is difficult because she does not know how to use the aids. Yet she refuses all help during care. She wants to get up and dress herself. When this is not possible, she accepts help with tears in her eyes. When her daughter comes to visit her the day after her admission, she takes out all her frustration on her and is angry and sad. She feels she does not belong in this place and is afraid she will not be able to return to her life before the accident. She also misses her friends and her neighbour. She had a lot of contact with her neighbour, and she also did small jobs for her. Doing some shopping, ordering something for her through the internet, helping her to fill out a government form. She is therefore worried and wonders whether the neighbour will be able to manage. She also feels that she is not doing her children any good, because one day a week she looks after her grandchildren, and she cannot do this.

94

4  Prognostic Decision: What Can We Achieve?

After a week on the ward, recovery is difficult. Mrs Thurston remains short of breath. She has trouble breathing properly and keeps having pain in her chest. The pain in her wrist and hip also persists. There is a swelling around the fractures. The nursing home physician thinks this is caused by overstraining. Because of her broken wrist, walking with aids is challenging. Mrs Thurston prefers to read in her room and doesn’t like to mix with the other patients. As determined as she was in the beginning to recover quickly and go home, she is now depressed. She does what she has to do, but she takes very little initiative herself.

References 2. Nederlandse Vereniging voor Cardiologie: Hartwijzer. n.d.. https://www.hartwijzer.nl/roken. Accessed 4 Jul 2020. 3. Jellinek. n.d.. https://www.jellinek.nl. Accessed 4 Jul 2020. 4. Minneboo M, Lachman S, Snaterse M, Jorstad HT, Ter Riet G, Boekholt SM, et al. Community-­ based lifestyle intervention in patients with Coronary Artery Disease: the RESPONSE-2 trial. J Am Coll Cardiol. 2017;70(3):318–27. https://doi.org/10.1016/j.jacc.2017.05.041. 5. Thun MJ, Carter BD, Feskanich D, Freedman ND, Prentice R, Lopez AD, et al. 50-Year trends in smoking-related mortality in United States. N Engl J Med. 2013;368:351–64. https://doi. org/10.1056/NEJMsa1211127. 6. Chow CK, Jolly S, Rao-Melacini P, Fox KAA, Anand SS, Yusuf S. Association of diet, exercise and smoking modification with risk of early cardiovascular events after acute coronary syndrome. Circulation. 2010;121:750–8. https://doi.org/10.1161/CIRCULATIONAHA.109.891523. 7. World Health Organization (WHO). n.d.. https://www.who.int/standards/classifications/ international-­classification-­of-­functioning-­disability-­and-­health. Accessed 2 Aug 2022. 8. Carnevali DL, Thomas MD. Diagnostic reasoning and treatment decision making in nursing. Philadelphia, PA: J.B. Lippincott Company; 1993. 9. Albersnagel E, Van der Brug Y.  Diagnosen, interventies & resultaten. 3rd ed. Groningen: Wolters-Noordhoff; 2007.

5

Therapeutic Decision: What Can We Do About the Problem?

I’ve noticed that many patients over the age of 60 are concerned about the deterioration of their memory. Because they forget things more quickly, find it harder to remember names and words and have to think longer about how to get somewhere, some say they are afraid of getting Alzheimer’s disease. Yesterday, there was another patient asking for advice. He wanted to know if there was anything he could do to slow down the decline in his memory. Today, I received an e-mail from the care institution where I am interning, asking me if I wanted to attend a congress on exercise. I read somewhere that exercise has a great influence on healthy and happy ageing. So, when I got the e-mail, I wondered if exercise helps the memory function of the elderly. Maybe that would be a good intervention for yesterday’s patient.

5.1 What Is the Difference Between a Therapy and an Intervention? In this book, we mean by ‘therapy’ or ‘treatment’: the whole of interventions aimed at a specific (para)medical, psychological or psychosocial health problem. By ‘intervention’, we mean one or more patient-related nursing  actions or activities which fall under one denominator or name, with the common objective of influencing a specific health problem. For example, ‘alternating positions’ (or changes in sitting posture), ‘preventing fluid and nutritional deficiencies’ and ‘information and instruction’ are three different interventions to prevent the occurrence of the health

Supplementary Information The online version contains supplementary material available at https://doi.org/10.1007/978-­3-­031-­27069-­7_5. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 J. Dobber et al., Clinical Reasoning and Evidence-Based Practice, https://doi.org/10.1007/978-3-031-27069-7_5

95

96

5  Therapeutic Decision: What Can We Do About the Problem?

problem ‘pressure ulcers’ [1, 2]. If, for instance, in cooperation with the general practitioner, the dietician or the occupational therapist, different interventions are applied simultaneously, together they form the decubitus prevention therapy. ‘Therapy’ is thus broader and may include a number of different interventions. So, why is reasoning about interventions called ‘therapeutic reasoning’? It is international practice to talk about therapy when it comes to the question what can be done about a health problem. In this book, we follow this usage. Hence, the choice of the terms ‘therapeutic reasoning’ and ‘therapeutic decision’.

5.2 Is Therapy Also Performed by Nurses? Nurses are part of a multidisciplinary team. In answering the question, ‘What can we do about the problem?’, we need to look at the nurse’s participation in the treatment provided by other disciplines, on the one hand, and the interventions that influence the nurse’s diagnoses, on the other. The former is therefore collaborative or participatory practice and the latter autonomous practice. The nurse participates in the therapy of practitioners such as doctors, psychologists and physiotherapists, but that practitioner is primarily responsible for the therapy. Within their own professional nursing domain, the nurse is primarily responsible and carries out independent interventions aimed at health problems including prevention and self-management, but excluding treatment. Therapeutic reasoning by the nurse is thus autonomously directed towards health problems including prevention and self-management, but excluding treatment. From a collaborative perspective, the nurse does reason about the therapy of the treating disciplines, particularly where the nurse has been delegated an intervention or aspect of treatment. In doing so, the nurse also reasons about the desired and undesired effects of the therapies carried out by the practitioners, as otherwise, it is not possible to monitor the effects of the treatment properly. This is because without knowledge, and diagnostic and therapeutic reasoning, the nurse would not know what to look for, how to interpret symptoms and assess their significance (see Sects. 1.6 and 12.36).

5.3 What Is Therapeutic Reasoning? Therapeutic reasoning is a form of clinical reasoning. It aims to answer the question of what is the best intervention for the individual patient with a specific health problem. In doing so, the nurse considers what interventions are available and what desired and adverse effects can be expected from these interventions. In therapeutic reasoning, the nurse should consider the scientific evidence for the effect of the intervention, the focus of the intervention and the hypothesised mechanism of change of the intervention. In addition, the nurse considers the risk of

5.4  How Do I Know What Interventions There Are?

97

unwanted effects of the intervention. For example, ‘falling’ is an unwanted effect of encouraging more exercise and going outside in an elderly patient who has not been outside for 3 weeks. Once a first selection of possible interventions has been made, this selection is used to assess the feasibility of the interventions in the patient’s own practice. After this second selection, the remaining interventions can be discussed with the patient. The patient’s preference is given important weight in the final therapeutic decision: the choice of intervention(s) to influence the health problem. The standard questions in therapeutic reasoning focus on the expected effect of the interventions and on the suitability of the interventions. Answering the questions leads to an informed choice of one or more interventions and to an answer to the question, ‘What can we do about it?’. 1. Which interventions are available? 2. What effects (with respect to the goal) can be expected from these interventions? How large is that effect expected to be? 3. What evidence is there that this intervention does indeed have that effect? 4. What is the focus of intervention (is it the causes, related factors, symptomatology, prognostic factors or risk factors)? 5. What is the mechanism of change of the intervention? 6. Is the intervention feasible (resources, practicability)? 7. Is the intervention acceptable to the patient? (Sect. 12.10) Based on the answers to these questions, the nurse reasons about the most appropriate intervention(s) for this patient. Please note: To answer Questions 2 and 3 of the standard questions on therapeutic reasoning, adequate knowledge of critical appraisal of intervention research and of guidelines is required. For this, see Part II of this book (See Sect. 12.71.4 for a process worksheet Therapeutic Reasoning).

5.4 How Do I Know What Interventions There Are? There are various sources of interventions. A large proportion of nursing care involves standard situations, whereby the standard care has already been established. This standard of care is recorded in protocols. Protocols should be up to date: if there is new knowledge and new material, the protocol should be adjusted. When identifying possible interventions, the nurse should consider whether the protocol is indicated for this individual patient in this situation or whether there are circumstances or patient characteristics that would make it inadvisable to use the protocol here. If the protocol is indicated, there is no need to look for alternative interventions. However, it may be necessary to consciously deviate from the protocol. Sometimes a minor adjustment to the care is sufficient, and the protocol can then be followed.

98

5  Therapeutic Decision: What Can We Do About the Problem?

The next source of interventions is the professional literature. Some of the interventions can be found in the background literature. The advantage of this is the ease with which the intervention can be found. The disadvantage is that in many cases, it is unclear whether these interventions are also the best interventions for this patient. Furthermore, the nurse can search for evidence-based interventions in the foreground literature or in the guidelines. Part II of this book is about evidence-based practice. That part covers the necessary knowledge to assess the interventions described in the foreground literature (Sect. 12.37).

5.5 How Do I Know What Effects I Can Expect from an Intervention? And How Do I know If There Is Strong Evidence for These Effects? If the intervention is described in the scientific foreground literature, information on its effects should be included in the article. The authors of such a journal article describe in their research findings not only the benefits of the intervention in terms of effect size per outcome but also the disadvantages, such as side effects. The research design of the intervention study can strongly influence the effects of the study. How robust the evidence is for the effect of an intervention is largely determined by the research design to examine it, how the study was conducted and the amount of evidence. In Chap. 10 of this book, we discuss critical reading and appraisal of scientific articles about interventions. When assessing the effects of interventions, it is important to check whether the outcome that is most important for the patient has been investigated. For example, think of a patient who wants to improve the functioning of his memory. If the article about the intervention only reports on the improvement of the physical memory structure in the brain and does not describe whether the research participants indeed performed better on memory tests after the intervention, the effect of the intervention cannot be properly assessed. The nurse should also identify the adverse effects of the intervention. Some interventions have not only positive effects but also (expected or not) serious or burdensome effects. In assessing the appropriateness of an intervention, the pros and cons should be weighed.

5.6 How Do I Know If the Evidence Applies to My Patient? A first assessment of the usefulness of the intervention for the patient can be made by examining whether the participants in the study are similar to your patient. Are they similar in age, health or disease? Based on these characteristics, a rough assessment of the validity of the evidence for the patient can be made. After that, a more precise assessment is needed. More information on this follows in Part II of this book.

5.8  How Do I Find Out the Mechanism of Change of the Intervention? Fig. 5.1  Contribution of the intervention to the SMART patient goals

99

P goal prognosc factors E intervenon

S

5.7 What Is Meant by the Point of Intervention? The focus of the intervention is the element of the health problem which the intervention tries to influence. This can be any of the following: etiological factors, related factors, risk factors, prognostic factors, symptoms of the health problem. In most cases, interventions focus on etiological factors. After all, if the causes of the problem are reduced or eliminated, the expectation is that the problem itself will also be reduced or eliminated. However, this is not always the case because, in addition to the causes, there may also be related factors which did not cause the problem but nevertheless perpetuate it. The intervention can also target these related factors. Other possible targets of the intervention are risk factors and prognostic factors. Sometimes the intervention is focused on the signs and symptoms of the health problem. In this case, the intervention does not address the causes of the problem but only the symptoms. The intervention should contribute to the SMART patient goal. This is shown schematically in Fig. 5.1.

5.8 How Do I Find Out the Mechanism of Change of the Intervention? In many cases, understanding the mechanism of change of interventions requires in-depth background knowledge. Sometimes this means studying normal and abnormal anatomy and physiology and psychological and psychosocial processes. Which of these processes are affected by the intervention? In what way does the intervention influence these processes? What consequences does this influence have for the course of these processes? How do these processes influence the functioning of the person? In seeking answers to these questions, the mechanism of change becomes increasingly clear. The more the nurse understands about the mechanism of change of an intervention, the more she is able to carry out the intervention with greater insight and evaluate the benefit of the intervention for the individual patient. The mechanism of change of some interventions is not (yet) known. It may be that scientific research has convincingly demonstrated the effect of the intervention,

100

5  Therapeutic Decision: What Can We Do About the Problem?

but it remains unclear how and why the intervention produces this effect. If the mechanism of change is unknown, there may one or more hypothetical mechanisms of change, based on plausibility of the hypothesised mechanism (Sect. 12.38).

5.9 How Can I Assess the Feasibility of the Intervention? Only if the intervention is available, it can be implemented in practice, and it may be feasible. Some interventions are readily available as they have already been implemented in care or treatment in other patients. Other interventions are easy and quick to learn and fit easily into the work program of the nursing team. However, this is certainly not the case for all interventions. If an intervention is new to the team, a number of preconditions must be met to implement it correctly. 1. The team members should have sufficient knowledge and skills to carry out the intervention. Some interventions are complex to carry out and require specific knowledge and skills. If this is the case, it should be checked whether this knowledge and these skills are sufficiently available in the team. If two nurses in the team know the intervention well and can carry it out without further preparation, this may be sufficient. However, it is sometimes necessary for each team member to be able to carry out the intervention. In this case, the other team members need training to be able to carry out the intervention properly. Training may not be available immediately. Moreover, the time and budget may not be available to carry out the training. 2. If equipment, aids, tools or other materials are needed for the intervention, the intervention can only be carried out if these are readily available. It cannot be assumed that equipment, aids and materials are available or can be acquired quickly. 3. If an intervention requires equipment, tools or materials, these may be too expensive to purchase. Interventions that require a lot of implementation time may also be too expensive as this may increase staff costs. 4. Finally, some interventions take too much time to fit into a busy work programme.

5.10 How Do I Choose from Several Alternative Interventions? In the preceding steps of therapeutic reasoning, the interventions were examined. Probably, several interventions have already been discarded as they did not prove to be sufficiently effective or because there was too great a risk that they would have serious disadvantages. It is also possible that some of the interventions were dropped because they were not feasible. All of the remaining interventions should be appropriate for this patient’s health problem. The nurse may have a preference for a particular intervention, but may feel that the costs and benefits of other interventions are sufficiently balanced. These

5.11  How Do I Get from Choosing the Intervention to Implementing It?

101

remaining interventions can be presented to the patient and their relatives to clarify which interventions are acceptable to them and which intervention the patient prefers. Once all of these steps have been completed, the nurse will have identified all of the information that will allow the final choice of intervention to be made.

5.11 How Do I Get from Choosing the Intervention to Implementing It? In standard interventions, the intervention can be carried out according to a protocol. In this case, a concrete implementation plan or action plan is available. If the intervention is not part of the standard repertoire of nurses, further concretisation is required in the form of a plan. The exact form of the plan may vary between interventions. In any case, the plan should make clear: • • • • •

which activities belong to the intervention; what materials, tools and equipment are needed; how to carry out these activities; when to carry out these activities; by whom to carry out these activities.

Glossary Adverse effects  Adverse effects of an intervention are unintended side effects of an intervention that have a negative impact on the patient. For example, side effects of medication or severe fatigue after strenuous physical exercise. Autonomous practice  Autonomous practice means that the nurse makes decisions and carries out actions for which she is primarily responsible. In autonomous practice, the nurse has this autonomy, since these actions fit in the nursing professional domain. Background knowledge, background literature  Background knowledge is basic knowledge that nurses in training acquire through their textbooks. In contrast to foreground knowledge, background knowledge involves relatively unchanging knowledge, such as anatomy, physiology and developmental psychology. Collaborative practice  Participation in treatment or decision-making for which a professional other than the nurse is primarily responsible (also: participatory practice). Etiological factor  Factor that (partly) causes the health problem. Focus of the intervention  The element in the diagnosis and/or prognostic factor that the intervention addresses and that is changed by the intervention. Intervention targets can be (a) causes, (b) related factors and (c) symptoms. In addition, (d) prognostic factors and risk factors can also be a focus of the inter-

102

5  Therapeutic Decision: What Can We Do About the Problem?

vention. If the intervention targets only one or more symptoms, it only addresses the manifestation of the health problem (e.g. burdening symptoms). Foreground knowledge, foreground literature Foreground knowledge is relatively rapidly changing knowledge. This knowledge changes as new knowledge and insights are acquired through scientific research. Examples of foreground knowledge are knowledge about the most effective intervention for a particular health condition or about the most important prognostic factors for a particular health problem. Intervention  One or more patient-related actions that have a common goal aimed at influencing a specific health problem. Mechanism of change  The way in which the intervention (or the etiological or prognostic factor) brings about the desired change in the patient. For example, many drugs affect chemical processes in the body, and many psychological interventions affect psychological processes. The mechanism of change explains through which physiological, psychological or other process the intervention intervenes and makes changes and how those physiological or psychological changes contribute to the desired effect of the intervention. Outcome  Outcome refers to that which is influenced through an intervention. Usually an outcome is stated in one or two words, for example ‘weight’, or ‘social integration’. Please note that there is a difference between ‘outcome’ and ‘goal’ and ‘result of the intervention’. The outcome indicates in general terms what is being worked on, the goal is the intended result in terms of patient behaviour, whereas the result of the intervention indicates precisely what the result of that intervention is for the patient. If the outcome is ‘weight’, the goal may be ‘Mr X weighs between 75 and 78 kg.’, the result may be ‘gained 3 kg’. Participatory practice  Participation in treatment or decision-making for which a professional other than the nurse is primarily responsible (also: collaborative practice). Prognostic factor  Factor that, if the health problem is already present, positively or negatively influences its course and outcome and in this way also influences the health outcome(s). Related factors  Factors that do not cause the health problem, but which, once the health problem is present, help maintain the presence of this problem, or factors that did not cause the health problem, but which contributed to the arising of the problem. Risk factor  Factor that, if a specific health problem is not yet present, increases the risk of developing that health problem. Signs and symptoms  Signs and symptoms that are always or almost always present in a specific health problem. Because these symptoms are (almost) always present, they characterise the health problem. ‘Signs’ are the characteristic phenomena that are measured or recognised by the health professional (such as blood pressure), while ‘symptoms’ are the characteristic phenomena that the patient can describe himself.

Cases and Learning Tasks

103

Therapeutic reasoning  Clinical reasoning about the expected effect of interventions and about the appropriateness of interventions, in order to select the most appropriate intervention(s) for the patient and to choose them together with the patient. Many interventions have both desired and undesired effects. In therapeutic reasoning, both types of effect are included in the reasoning.

Cases and Learning Tasks Sven’s life, before he used coke for the first time, was boring and ordinary, looking back on it now. Ordinary things, like playing sports or drinking a beer with his friends, didn’t really make him feel good. He remembers well the night he received coke for the first time. Sven was at a dance party when one of his friends gave him coke. The effect was overwhelming: he had energy, was bursting with self-­confidence and could continue drinking alcohol without getting nauseous. In the years that followed, his use of cocaine increased. At first, he sniffed it only at festivals, where he went about once every 2 months. In the course of time, he began to use it on Saturdays too. Without coke, going out was no fun for him. His use of drugs became more and more excessive. The weekend started on Thursday evening and ended on Sunday afternoon. Almost always he was awake for more than two nights in a row. In addition to cocaine, he used amphetamine, ketamine and GHB to deal with the side effects of heavy cocaine use. When he wasn’t using, he was craving drugs and collecting them for the weekend. His heavy drug use led to debt and emaciation. He was fired from his job. In 1 year he used 20,000 euros worth of drugs and lost 10 kg. By now he was also able to put his nose almost flat against his face, because the nasal septum was seriously damaged. Sven’s parents could no longer close their eyes to the situation in which Sven found himself. He was rude, did not keep appointments, did not pay the rent of his apartment on time and stole money and things from his parents’ house. He walked around with big pupils. When his parents denied him access to their house, Sven had a wake-up call. He realised that this way, his life was a dead end. Through his general practitioner, he applied for admission to an addiction clinic. After going through a program at the detox ward, he will soon start day treatment. After a period of structure in the clinic, where the focus was on rehab, he has to face the problems he has and think about how to move on with his life. He talks to a nurse at the detox ward about this.

Learning Task 7: Sven Case Watch Video 5.1.

104

5  Therapeutic Decision: What Can We Do About the Problem?

The Sven case contains two main problems, namely ‘lack of a meaningful purpose in life’ and ‘high risk of relapse due to addiction’. The detailed example is about ‘lack of a meaningful life’. The learning task deals with the problem of ‘high risk of relapse from addiction’. 1. Examine the problem of ‘lack of meaningful content in life’ worked out in keywords on the process worksheet Therapeutic reasoning. The diagnostic, ­etiological and prognostic reasoning have also been incorporated into the process worksheet. 2. Now elaborate on the problem of ‘high risk of relapse’, paying attention to the prevention of relapse. Do this in running text, so that your reasoning (diagnostic, etiological, prognostic and therapeutic) is easy for the reader to follow. Use all the information from the written case and the film fragment that you consider relevant.

Process Worksheet for the Sven Case The worked example on the process sheet deals with the problem of ‘lack of meaningful life content’.

Process Worksheet Diagnostic Reasoning 1. What objective and subjective symptoms (complaints, signs, behaviours, abnormal value compared to the reference range) are present in the patient? (a) What data can be distilled from observation? • Not applicable. (b) What information about symptoms was obtained from the patient? • Sven is not sure what to expect in the future. • Looking for a way to use his experiences positively. • No functioning social network. • No day care in the community. • No living quarters. • Relationship with parents disrupted by addiction. • Possible debts. • Doesn’t know how to go about getting his life back on track after discharge from the clinic. (c) Are data from already known information available (file, handover, etc.)? • This information is not available. 2. How can these symptoms be explained (which abnormalities in anatomy, physiology or psychosocial functioning)?

Cases and Learning Tasks

3.

4. 5.

6.

105

(The anomalous process in addiction physiology plays a very limited role for the problem worked out here, and it is therefore not worked out in this process worksheet). (a) How does this happen in a healthy person? • Not applicable. (b) What is different in this situation? • In recent years, his life has been dominated by addiction. • Fear of the future. (c) What explanation can be given for this in this situation? • Breaking free from addiction requires a new purpose in life. What else would you like to know from or about the patient? Use the answer to this question to supplement the answers to Questions 1 and 2. (a) Are there any missing data that might be less prominent? • How is Sven’s financial situation? • How was he educated, and what labour did he perform? • How are Sven’s self-insight, judgement and self-confidence? • Is Sven capable of independently solving more complex problems, making realistic plans and successfully carrying out these plans? What is the patient’s hindrance (possibly expressed in the ICF components activities, participation)? (a) How do anatomical characteristics and body functions affect functioning? • Body function: impairment in experiencing self/identity leads to uncertainty about the future. How does the patient respond to this (emotional behaviour, [in]effective self-management)? (a) Does the response lead to a problem (these are potential Problems in the PES, or potential Etiological factors, or potential Signs and symptoms, or potential risk factors, or potential prognostic factors)? • He responds by seeking help by having a conversation with the nurse; this response is adequate and therefore does not lead to a problem (potential positive prognostic factor). (b) Does the reaction affect the prognosis (positively or negatively)? • Yes, getting help in a timely manner is a positive factor. Which possible health problems does this indicate (from which impairments in body structures, in body functions and limitations in activities and restrictions in participation)? (a) How many and what problems are there? • There are several problems, here we will only deal with meaning problems. (b) Are there problems that are interrelated? • This is definitely the case. (c) Which problems are important, which less so and what is the reason for this? • High risk of relapse into addiction. • Lack of meaningful life fulfilment. • Both are strongly related to building a new, healthy social network.

106

5  Therapeutic Decision: What Can We Do About the Problem?

7. What are the defining characteristics of the major health problems? (a) What is described about this in the literature? • For problems with meaning: difficulty making choices; disappointment about earlier choices; social pressure from the environment; doubts about one’s own abilities and brooding about one’s purpose in life; a feeling of emptiness. 8. How can the presence or absence of these defining characteristics be tested? Then test whether these phenomena are actually present. (a) Is that what has been described in the literature also the case with the patient? • Assessment is possible in a conversation with Sven.

Conclusion P: S:

Lack of meaningful life Brooding about the purpose/content of his life after discharge from the addiction clinic Disappointment about previous choices (addiction) Fear of the future (about how things will go; he does not know how to handle housing, work/day care, the content of life, social contacts)

Process Worksheet Etiological Reasoning 1. Are there possible causes/related factors from diseases/disease processes? If so, which ones? (a) End of the treatment period after addiction. 2. Are there possible causes/related factors from the (para)medical treatment? If so, which? (a) Treatment at the detox ward confronts Sven with reality and the problems that go with it. 3. Are there possible causes/related factors from body structures? If so, which ones? (a) Alterations in the reward system in his brain. 4. Are there possible causes/related factors from body functions? If so, which ones? (a) Uncertainty and doubts about his current identity and self-image. 5. Are there possible causes/related factors from activities? If so, which ones? (a) Lack of daily activities. 6. Are there possible causes/related factors from participation? If so, which ones? (a) Lack of (paid) work or voluntary work which Sven considers meaningful. 7. Are there any possible causes/related factors from environmental factors? If so, which ones? (a) Lack of a functioning social network. (b) Lack of own living space.

Cases and Learning Tasks

107

8. Are there possible causes/related factors from personal factors? If so, which ones? (a) No information on this in the case. 9. Can these factors together indeed cause and/or maintain this health problem? (a) What causes and related factors play an important role? • Discharge from addiction and re-enter society. • Uncertainty and doubts about his current identity and self-image. • No prospect of fulfilling his days/day-to-day tasks. (b) Can these causes cause the health problem? • Yes. (c) What causes and what related factors can perpetuate the health problem? • All, but especially the uncertainty about his identity and self-image. 10. Are these factors indeed present in the patient? (a) Yes. 11. Do these factors indeed have this effect in this patient? (a) Yes.

Conclusion E

The successful detox treatment is almost completed Uncertainty and doubts about his current identity and self-image Lack of a well-functioning social network Absence of daily activities/work/volunteer work/day structure after discharge from addiction centre Lack of own living space

Process Worksheet Prognostic Reasoning 1. Which physical, psychological and personal factors (including causal and related factors) negatively influence the course of the health problem? Which ones are positive? (a) Have the health-promoting and health-limiting factors been mapped out for a realistic assessment of the optimal achievable result? • Degree of self-confidence, understanding of self and judging what he can and cannot do (no data on this at the moment). • Size of his ability to solve problems (no data on this at the moment). • Be able to make plans and implement them (no data on this at this time). 2. Which environmental factors influence the course of the health problem negatively? Which ones are positive? (a) Negative: broken contact with family, very small social network, no daytime activities, possible debts. Possibly also: negative social attitudes and

108

3.





4.

5.

5  Therapeutic Decision: What Can We Do About the Problem?

personal attitudes of family, acquaintances and strangers towards people who have been addicted. (b) Positive: support from clinic. Further ambulatory guidance is also possible. On what mechanism of action (per factor) is that influence based? (a) How does the factor bring about the (positive or negative) change? • Information about the physical, psychological and personal factors is lacking, so that no answer can be given to the question about these factors. • Few positive contacts: positive contacts increase people’s resilience. A lack of them is negative. It is more difficult for Sven to make good choices about the content of his life if he can only talk about it with few people. • No daytime activities: we derive part of our identity from what we do. The lack of meaningful daytime activities makes it difficult to develop the feeling of being useful, of having a place in society. • Professional support and guidance: these can temporarily take the place of an informal social network. Sven can talk to them about what he wants, they can support him where necessary. (b) How big is the influence of the factor? • The negative factors weigh heavily, more heavily than the positive environmental factors. • The psychological and personal factors carry about the same weight as the negative environmental factors. (c) To what extent can the factor itself be influenced? • Self-confidence, self-insight and judgement can be influenced through therapy. • The problem-solving ability and the ability to make plans can be influenced externally to a limited extent through therapy and education. • Sven can be supported in re-establishing contact with his family, and this also applies to building up a healthy social network. • Interventions aimed at daily activities, work and living are possible. • It may be possible to limit any debt through debt assistance. • Negative attitudes are difficult to influence from the outside. • The ambulant support from the clinic can be determined by the clinic itself, but is dependent on financial possibilities. Do those factors add up to positive, neutral or negative? (a) Has it been reasoned how the various factors will interact with each other and with the health problem? • In this case, psychological and personal factors will be the deciding factors. These are still largely unknown. These factors can be supported and strengthened through ambulatory guidance. • All in all, the factors are cautiously positive, but a lot depends on Sven’s psychological and personal factors. What is the maximum achievable result for this health problem? (a) Is the maximum achievable result prevention, resolution, improvement or remission, delay or minimisation, stabilisation, deterioration-palliation? • The maximum achievable result is therefore an improvement.

Cases and Learning Tasks

109

Conclusion Type of prognosis: Improvement.

Process Worksheet Therapeutic Reasoning 1. What interventions are available? (a) Is there an overview of available interventions? • Residential support, assisted living, supported living. • Daily activities in a safe environment, job training, supported employment. • Outpatient supervision of functioning in all areas of life, Rehabilitation, intensive forms of case management such as flexible assertive community treatment (FACT). • Training in problem solving and planning. • Therapy to increase self-confidence and possibly self-insight. (b) Is the situation a standard situation and/or is the intervention part of a protocol at the work unit? • No. (c) Are there interventions described in a guideline, in professional literature or background literature? • Information about the interventions can be found in professional and background literature. 2. What effects (in terms of goal) can be expected from this intervention and how large is this effect expected to be? (a) Because the interventions interact, it is difficult to predict the result. Rehabilitation has positive effects on the achievement of patient goals in various life areas (housing, daily activities and work, education/learning, social contacts, finances). Intensive case management (such as FACT) contributes to maintaining stable housing. The success of a step-by-step development from daily activities to work training, to (supported) paid work, depends strongly on the capabilities and motivation of Sven. At the moment, Sven seems to be very motivated, although he is not sure what kind of daily activities he would like. A care farm might be a first step, but the work there is different from what Sven said he wanted to do, namely to mean something for people who are in a similar position to his own. (b) Sven says that he mainly wants to be physically active and that he ‘doesn’t want to spend too much time in his head’. This pleads in favour of starting with day care and living first, and depending on the course of this, to find out what further guidance, support, training or therapy is needed/wanted. 3. What evidence is there that this intervention does indeed have that effect? (a) Can the intervention be expected to be effective in this particular patient? • There is some evidence of the effectiveness of intensive case management. This evidence has mainly been obtained in studies with other target groups than former addicts.

110

4. 5.

6.





7.

5  Therapeutic Decision: What Can We Do About the Problem?

• There is evidence of the effectiveness of rehabilitation. This evidence has mainly been obtained in research with other target groups than former addicts. • Supported living and day care in a safe environment are accessible interventions to get used to a daily rhythm and independent living. What is the focus of intervention (is it the causes, related factors, symptomatology, prognostic factors or risk factors)? (a) Through which aspect of the health problem does the intervention influence the problem? • The interventions are focused on the etiological and prognostic factors. What is the mechanism of change of the intervention? (a) Does the intervention bring about changes in anatomical and/or physiological structures or in psychosocial functioning? • The interventions are aimed at psychosocial functioning. (b) What changes are these? • In Sven’s case, these interventions are about psychological processes: getting to know other people, trusting them, entering into a relationship with them and maintaining this relationship; developing the feeling that he takes part in society and that he has found his place; that he has a stable positive self-image. (c) How certain is it that this mechanism of change does indeed occur (what is the probability)? • It is not known how likely these psychological processes are to occur. Is the intervention feasible (resources, practicability)? (a) Are any tools, equipment, instruments or other materials needed for this intervention? If so, are these available? • The interventions are beyond the capabilities of the addiction clinic. The clinic must register the client with other care organisations, and in Sven’s case, an intake at the various organisations is necessary. (b) How long does the execution of the intervention take? • The interventions themselves are long-lasting and intensive. • Registration can be done through the addiction clinic’s established collaborative contacts with the other care organisations. (c) Is there any particular knowledge and/or skills required to deliver the intervention? If so, are these knowledge and skills present in sufficient numbers of nurses to ensure implementation of the intervention on a daily basis? • The interventions require specialist skills and are carried out by trained professionals within care organisations other than the addiction clinic. (d) Is the benefit of the intervention balanced with the costs of the intervention? • If the interventions are part normal funded or insured care costs, the costs and benefits are balanced. Is the intervention acceptable to the patient? (a) Is it feasible to discuss possible options with the patient? • Sven wants to be physically active, and this is possible through assisted day care. If this is done with him in the perspective of stepwise guidance

Cases and Learning Tasks

111

to work or voluntary work which suits him, the expectation is that Sven will find this acceptable. • Sven indicates that he does not know how to approach housing, so even residential support or assisted living seems to be acceptable to him. • Sven has also indicated that he needs support, preferably from the current supervisors. Ambulatory forms of support could be an acceptable alternative. (b) Have the options been discussed with the patient? If this is not possible, have the options been discussed with his next of kin? • The options have not yet been discussed with Sven, this should be the next step. Mr Bachoni is an Italian man of 75 years old. Since a few weeks, he lives in the nursing home, because he did not take care of himself anymore and especially because he followed his rules for diabetes less and less. After the death of his wife 2 months ago, he lived alone and his house was rapidly becoming dirty, even though his daughters were willing to help. About 5 weeks after the death of his wife, Mr Bachoni was just sitting in his chair. He only got out of it to go to the toilet. According to his daughters, he didn’t even sleep in bed anymore, but in that chair. On the ward, Mr Bachoni usually sits on a fixed place in front of the window. Last week, he told us he likes this spot very much, he can rest there, so his fatigue doesn’t bother him so much. A few days ago, Mr Bachoni got confused, and he appeared to have a very dry skin and mucous membrane. In addition to dehydration, the doctor noted that Mr Bachoni’s diabetes (diabetes mellitus, type 2) was dysregulated. Mr Bachoni has always been overweight, and he had a BMI of 27.5 upon admission. At the nursing home, Mr Bachoni is placed on a diet, but it is not noticed if Mr Bachoni adheres to the diet or if he eats much or little. In this care home, the nutrition assistant prepares the bread for all residents at the bread meal. The hot meal is delivered by a care catering company. Only the nutrition assistant is present on the ward during meals, while the nurses take a break. Often there are a few family members who help their relative with the meal. When the nutrition assistant was asked if Mr Bachoni sticks to his diet, and how much he eats and drinks during meals, she said she was too busy giving everyone their food to keep an eye on it. Mr Bachoni probably does eat enough, because his daughters often visit with Italian food for their father. They take him from the ward, and he probably eats what they have brought him. Mr Bachoni can go to the toilet independently and can also take water independently. Because of this, it is not clear how much he drinks and how much he urinates. The senior nurse decides it is time to talk to Mr Bachoni about his diabetes. She would like to get an idea of what he knows about diabetes and how he can best be supported in dealing with the disease.

112

5  Therapeutic Decision: What Can We Do About the Problem?

Learning Task 8: Mr Bachoni Case Watch Video 5.2. The problems in the case of Mr Bachoni are related to diabetes. The situation in the case and the film clip show two main problems. Elaborate on both of them. 1. What two main problems do you see here? 2. Elaborate on these problems using the basic questions for diagnostic, etiological, prognostic and therapeutic reasoning.

References 3. Nederlands Huisartsengenootschap. NHG Standaard decubitus. May 2015 version. 2015. https://richtlijnen.nhg.org. Accessed 18 Jul 2020. 4. Halfens RJG, Meesterberends E, Neyens JCL, Rondas AALM, Rijcken S, Wolters S, et  al. Landelijkel Prevalentiemeting Zorgproblemen. Rapportage resultaten 2015. Maastricht: Maastricht University, CAPHRI; 2016.

Part II Evidence-Based Practice (EBP)

6

Clinical Reasoning and Evidence-Based Practice (EBP)

My fellow nurses try to persuade patients that they should eat more healthily and exercise more. They tell patients that this is good for their health and that the patient then runs less risk of becoming ill again. Unfortunately, very few patients follow this advice. I have noticed that this kind of advice does not work for many patients. I started looking for a way in which many more patients would decide to live a healthier life. I found a study by Lundahl et al. on motivational interviewing. This has turned out to be an effective intervention for behavioural change. I read more background literature about it and followed a training course. It has made my work easier, and I am happy with what I am now able to achieve with many patients. I have helped my colleagues to also work in this way. Some of them have also followed a training course. (The full article by Lundahl et al. (2013) can be found online, https://doi. org/10.1016/j.pec.2013.07.01 [1]).

6.1 What Is Evidence-Based Practice? How Does Evidence-­Based Practice Help Improve My Clinical Reasoning? Evidence-based practice (EBP) is a method of making better clinical decisions. In other words, EBP is at the service of clinical reasoning. By using the EBP method, scientific knowledge is involved in making the best decision for the patient. The data collected with and about the patient is combined with scientific knowledge about diagnosis, etiology, prognosis and therapy.

With contribution by M. Offringa.

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 J. Dobber et al., Clinical Reasoning and Evidence-Based Practice, https://doi.org/10.1007/978-3-031-27069-7_6

115

116

6  Clinical Reasoning and Evidence-Based Practice (EBP)

It is not uncommon for the nurse to be in doubt and unsure of the best decision for the patient. For example, Isabella wondered whether there was not a better lifestyle intervention than what had been customary in the department until then, namely convincing the patient to adopt healthier lifestyles. Isabella’s realisation that the method of persuasion does not work for many patients, but that she does not (yet) know of any other method that is better, is referred to as ‘clinical uncertainty’. Recognising and acknowledging this kind of clinical uncertainty are the main reason for combining scientific knowledge with clinical knowledge about the patient via the EBP method. In this book, we use the following definition of evidence-based practice ‘the integration of best research evidence with clinical expertise and patient values to facilitate clinical decision making’ [2].

6.2 What Is the Method of Evidence-Based Practice? Thus, a clinical uncertainty is the starting point of the method of EBP. By formulating a clinical uncertainty in a structured clinical question, a systematic and efficient search can be made in scientific biomedical and psychological databases for the best evidence. If scientific articles are found that seem to provide an answer to the clinical uncertainty, then a critical assessment of these articles is important. To use scientific knowledge appropriately in clinical reasoning, the nurse should ask three questions about this knowledge: 1. Is the new information emerging from the scientific articles used likely to be true? 2. What does this information tell us about patient care? 3. Can this information be used for one’s patient? (Sect. 12.39) If the assessment of this knowledge is positive, then the nurse can use the scientific knowledge in their reasoning when making the clinical decision. Because this is not a simple reasoning and searching process, a critical review by the nurse of how she has carried out this process is necessary. Was the process done in a way that made it likely that, for this individual patient, the best scientific evidence was indeed found? And was this evidence used appropriately? The method of evidence-based practice therefore includes the following five steps [3]: 1. Ask: reformulate the identified clinical problem into a structured clinical question 2. Acquire: search and select the best evidence 3. Appraise: weigh the quality of the evidence, and identify the results 4. Apply: determine whether the results found in the evidence can also be expected for your patient 5. Act: decide whether and how to use the evidence (this often includes shared decision making with the patient)

6.4  How Do I Know What the Best Evidence Is?

117

6.3 How Can I Translate a Clinical Problem into an Answerable Question? An important characteristic of a structured clinical question is that the question cannot be answered with just a simple ‘yes’ or ‘no’. In addition, most clinical questions contain the following three elements: (1) population, (2) determinants and (3) outcome. The following is an explanation of these three elements. Population refers to the group of people who have the same important characteristics as the patient. A 71-year-old man who is admitted to hospital with pneumonia belongs to the population ‘elderly patients with pneumonia in hospital’. Determinants are factors that (a) cause or promote the onset of a disease or (b) influence the course of the disease. For example, smoking is an important determinant for the development of lung cancer and many other diseases. Lifestyle is a factor that influences the course of cardiovascular disease. However, an intervention to support a patient with heart disease to adopt a healthier lifestyle is also a determinant that influences the course of the disease (Sect. 12.40). Finally, outcome is the variable in which the nurse is interested. For example, if asked about a good way to lose weight for obese people, the outcome will be ‘weight’. When asking about the effect of pain management, the outcome would be ‘pain’. A structured clinical question is formulated in a complete sentence, ending with a question mark. For example, the question about weight loss then becomes: ‘What is the effect of 6 months of Weight Watchers (determinant) on the weight (outcome) of men aged 65  years and older who have a BMI of 30 or higher (population)?’ (Sect. 12.41) The databases of scientific articles are searched using a well-chosen combination of keywords. In order to be able to search these databases properly, the question must first be converted into the most important keywords. For this, the PICO system is used. The abbreviation ‘PICO’ means population, intervention, comparison and outcome. For each element, the most important keywords are formulated, and these keywords are used to search the database for relevant articles. The PICO system is most suitable for questions on therapy. The ‘I’ stands for the intervention that the nurse is primarily interested in, and the ‘C’ stands for the intervention with which she would like to compare the first intervention. There are good specialist books on how to search scientific articles adequately using PubMed, or one can use the PubMed User Guide (Sect. 12.42) [4].

6.4 How Do I Know What the Best Evidence Is? The best evidence is evidence where (1) the clinical problem is sufficiently similar to the clinical question, (2) the patient population is sufficiently similar to one’s own and (3) the level of evidence is as high as possible. The level of evidence is represented by the evidence pyramid [5]. The sources of the study type with the highest level of evidence are at the top of the pyramid. For each step down the pyramid, the evidence

118

6  Clinical Reasoning and Evidence-Based Practice (EBP)

level of the source becomes slightly weaker (see Fig. 6.1). The higher the research type is in the pyramid, the less risk there is of bias distorting the evidence from a source with that research type. The research type is therefore important in determining the strength of evidence. In research, and therefore also in every source that reports on it, limitations in the quality of the evidence can occur that weaken the strength of evidence. For example, such a limitation occurred in a study with a much too small sample size (59 participants when there should have been 230) of an intervention to promote breastfeeding in neighbourhoods where only 25% of mothers breastfeed [6]. In the study, 74% of mothers who received the intervention started breastfeeding after delivery, compared to 41% of mothers who did not receive the intervention. However, the quality of the study is moderate, partly due to the sample size being too small. Due to this limitation, the strength of evidence of this study is lowered. On the other hand, factors may also be present that increase the quality of the evidence and thus strengthen its probative value. An example of this is the occurrence of the dose-response effect. Suppose that training in maintaining body balance had the effect that the elderly who followed the training fell less often than other elderly. And suppose that the elderly who received ten training sessions subsequently fall significantly less often than those who received five training sessions. In such a situation, a stronger intensity of the intervention seems to be associated with a stronger effect. This phenomenon is called the dose-response effect. If this effect is found in a study, this strengthens the evidence of the study, and there is stronger evidence for the intervention. Thus, critical appraisal of the evidence is indispensable. Relying on the study design and its place in the evidence pyramid, without critically appraising the study, can lead to errors in clinical reasoning and subsequent decision-making. Chapters 7–11 address this critical appraisal.

evidence-based guidelines

systemac reviews

primary studies (e.g. Randomized Controlled Trials)

Fig. 6.1  Evidence pyramid

6.6  How Can I Weigh the Found Evidence on Methodological Quality and…

119

6.5 Where Do I Find the Evidence? There are various guideline sites. Guidelines can be found via, for example, www. nice.org.uk (British), www.ahrq.gov/gam/index.html (American) and www.nhmrc. gov.au (Australian). See further Sect. 10.3. Systematic reviews are also called ‘syntheses’ because they combine (synthesise) the results of different original studies into the same effects. The strength of evidence is extra strong if the systematic review is accompanied by a meta-analysis. In such an analysis, the results (and sometimes the raw data) of various original studies are combined. The effect is recalculated on the basis of this combination. Because this calculation is based on all these underlying original studies together, the result of the meta-analysis is more reliable than that of the various original studies separately. The most important database for systematic reviews is the Cochrane Database of Systematic Reviews (CDSR). This is a paid database, but there is free access in many countries due to funding the Cochrane Library. In addition, many systematic reviews can be found through the large biomedical search engine PubMed. See further Chap. 10. Original studies can be found through various databases. The most important of these are MedLine (via the search engine PubMed [4]), Embase, Cinahl and PsycInfo. See further Chap. 10.

6.6 How Can I Weigh the Found Evidence on Methodological Quality and Applicability in My Own Practice? The critical appraisal of the quality of the study and the applicability of its results in one’s own practice is carried out by means of checklists. Each of the four types of basic questions involves a different type of research. So, a different checklist is used for each type of research. These checklists can be found in the in-depth chapter. However, in order to understand the checklists, knowledge of the different types of research and the corresponding quality requirements is required. This information can be found in the following chapters.

Glossary Bias (in scientific health research)  In scientific health research, bias means a systematic deviation from the underlying truth because of a feature of the design or conduct of a research study (e.g. overestimation of a treatment effect because of failure to randomise) [3]. Clinical uncertainty  Doubt or uncertainty about what the best decision is in a specific patient situation. Clinical uncertainty may arise in the diagnostic, etiological, prognostic or therapeutic areas.

120

6  Clinical Reasoning and Evidence-Based Practice (EBP)

Critical appraisal  The process of carefully and systematically examining research to judge its trustworthiness, and its value and relevance in a particular context [7]. Determinant  Factors or characteristics that are related in some way to the health outcome being studied. For example, smoking is a determinant for lung cancer, body weight for diabetes and cannabis use for psychosis. Evidence-based practice  ‘The integration of best research evidence with clinical expertise and patient values to facilitate clinical decision making’ [2]. Evidence pyramid  Graphic representation, in the form of a pyramid, of the hierarchy of evidential value of different types of scientific research. At the top of the pyramid is the type of research with the greatest strength of evidence (i.e. evidence-based guidelines); at the bottom of the pyramid is the type of scientific research with the least strength of evidence. In many cases, only the top part of the pyramid is displayed (as in this book). Meta-analysis  Analysis that can form part of a systematic review. In a meta-­ analysis, the results of the individual studies included in the systematic review are combined (pooling). By pooling the results, a new effect estimate can be calculated which is more reliable than the effect estimates of the individual studies. Original studies  Studies in which patients are participants to research a question concerning diagnosis, etiology, prognosis, therapy or patient perspectives and experiences. In contrast to syntheses and guidelines, in which the research design focuses on synthesising the results of other studies. PICO  Abbreviation for: Population, Intervention, Comparison of Control and Outcome. PICO is a way of identifying the most important keywords to perform a systematic literature search for intervention research in scientific databases. Based on the research question, the most important keywords are formulated for each of the four PICO elements. Synthesis  Systematic review in which, based on a research question, information from various scientific studies is combined according to a fixed and strict method. The results have a greater probative value than each of the individual studies used for the systematic review.

References 1. Lundahl B, Moleni T, Burke BL, Butters R, Tollefson D, Butler C, et al. Motivational interviewing in medical care settings: a systematic review and meta-analysis of randomized controlled trials. Patient Educ Couns. 2013;93:157–68. https://doi.org/10.1016/j.pec.2013.07.012. 2. DiCenso A, Guyatt G, Ciliska D. Evidence-based nursing. St. Louis, MO: Mosby; 2005. 3. Guyatt G, Rennie D, Meade MO, Cook DJ. Users’ guides to the medical literature. A manual for evidence-based clinical practice. 3rd ed. New York, NY: McGraw Hill Education; 2015. 4. PubMed. PubMed User Guide. n.d.. https://pubmed.ncbi.nlm.nih.gov/help/. Accessed 9 Aug 2022. 5. Agoritsas T, Vandvik PO, Neumann I, Rochwerg B, Jaeschke R, Hayward R, et al. Finding current best evidence. In: Guyatt G, Rennie D, Meade MO, Cook DJ, editors. Users’ guides to the medical literature. A manual for evidence-based clinical practice. 3rd ed. New York, NY: McGraw Hill Education; 2015.

References

121

6. Wolfberg AJ, Michels KB, Shields W, O’Campo P, Bronner Y, Bienstock J. Dads as breastfeeding advocates: results from a randomized controlled trial of an educational intervention. Am J Obstet Gynecol. 2004;191:708–12. https://doi.org/10.1016/j.ajog.2004.05.019. 7. Burls A.  What is critical appraisal? n.d.. https://www.whatisseries.co.uk/what-­is-­critical-­ appraisal/. Accessed 9 Aug 2022.

7

What May Be the Matter with the Patient: How to Critically Appraise Articles on Diagnostic Tests?

Since 2 days, Mrs Bolscher is in our care. Last week, she fell on the street, and she had a complicated hip fracture. Fortunately, the surgery went without any problems, because Mrs Bolscher is already 89 years old. The day before yesterday, she came home again. We attend to her every day. Her children visit her often, so she is not alone a lot. This morning, Mrs Bolscher told a strange story about ants she saw walking on the wall, and the conversation with her was also strange. When I asked her something, she gave an answer that had nothing to do with the question, and when she told me something, she stopped in the middle of her story. She also suddenly became very sad for no apparent reason. I immediately thought of delirium. Fortunately, last month, our local team searched for screening instruments for delirium. The Delirium Observation Scale (DOS) proved to be a good tool. At lunchtime, I filled in the DOS for Mrs Bolscher at the office of our neighbourhood team. She scored six points, so that means she probably does indeed have delirium.

7.1 What Is a Diagnostic Test? The question ‘what may be the matter with Mrs Bolscher?’ is a diagnostic question. To test the hypothetical diagnosis, preferably a diagnostic measuring instrument is used. Diagnostic questions are common in nursing practice, but can be difficult to answer because often there is no gold standard. Nurses do not make medical diagnoses, but they do often screen for the presence of certain symptoms. Nevertheless, a diagnostic measuring instrument (such as the DOS) can help the healthcare professional to adequately recognise the symptoms of a health condition (such as

With contribution by A.M. Eskes, C.H.M. Latour, and R.J.P.M. Scholten. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 J. Dobber et al., Clinical Reasoning and Evidence-Based Practice, https://doi.org/10.1007/978-3-031-27069-7_7

123

124

7  What May Be the Matter with the Patient: How to Critically Appraise Articles…

delirium). This is important in order to treat patients properly, but also to inform patients and their relatives about the course of a delirium, for example. To make a good choice for a certain measuring instrument, it is important to gain insight in the current literature on the subject and to critically assess the literature found. This literature focuses on the question: How accurately does the diagnostic tool measure the presence or absence of a specific diagnosis? Such a study is also called a ‘diagnostic accuracy study’. In diagnostic accuracy studies, a newly designed diagnostic test is tested. This is done by comparing the performance of the new diagnostic test with an existing diagnostic test, preferably with the test that is currently the gold standard [1]. The new diagnostic test is called the index test, the existing test is called the reference test. Testing is done by comparing the performance of the index test (the number of errors the index test makes in determining the diagnosis) with the performance of the reference test. The reference test has been examined previously, so the number of errors the reference test makes in determining a specific diagnosis is known. The index test is often a simpler, less invasive and cheaper test than the reference test. The assessment of a diagnostic measuring instrument can be divided into three aspects [2]: 1. the validity of the research 2. the importance of the results 3. the applicability of the results

7.2 How Can I Assess the Validity of a Diagnostic Measuring Instrument? The first requirement for any scientific research, including diagnostic research, is validity. The measuring instrument must measure what it is supposed to measure. To determine this, there are five criteria by which the information in the diagnostic study article about the design and conduct of the study is assessed. This information is given in the article under the heading methods (this is called the methods section of the article). Below are five questions with which these criteria which can be assessed (see also Sect. 12.43); this checklist is based on the checklist of Leeflang et al. [2]). 1. Were the study participants admitted to the study sequentially or on the basis of a random sample? 2. Has the index test been compared to a valid reference test? If a gold standard exists, has the index test been compared to this gold standard? 3. Are the assessors of the index test blinded to the result of the reference test? Are the assessors of the reference test blinded to the results of the index test? 4. Was the decision as to whether the reference test should also be performed on the study participants, taken independently of the result on the index test? 5. Was there no selective dropout of study participants? (Sect. 12.62.1)

7.4  What Is a Valid Reference Test?

125

The following text provides background information needed to adequately interpret the validity of a diagnostic study.

7.3 What Is Meant by ‘Sequential’ or ‘Random Sample’? Why Is It So Important? In the ideal case, all patients from the source population (e.g. all patients who present themselves to their general practitioner with a particular type of complaint) are admitted to the study. That is to say, all consecutive patients with the same complaint, without exception. In that case, the study population is a good representation of the source population. Another possibility is to draw a random sample. In a random sample, each person from the source population has an equal chance of being in the sample and, therefore, of participating in the research. A random sample is also a good representation of the source population. This good representation is important because the validity of the measuring instrument partly depends on the selection of the patients. This means that researchers are not allowed to determine on the basis of patient characteristics whether patients should be allowed to participate in the study. Therefore, the study participants should be admitted to the study in a random manner. If selection is made in another way (e.g. on the basis of clinical patient characteristics other than the complaints and symptoms with which the patients present), this may make it more difficult or easier to make the correct diagnosis with the index test or with the reference test [2]. It is more difficult if the clinical features selected are not distinctive of the condition being tested and easier if these features are distinctive.

7.4 What Is a Valid Reference Test? To determine whether a measuring instrument (the index test) works properly, a valid reference test must be used. The ideal reference test indicates with certainty whether a disease is present or not. In that case, it is the gold standard. An important bottleneck in nursing practice is that there is often no real gold standard. New measuring instruments are then compared with tests that, according to the professional group, approach reality best.

7.5 Why Do the Evaluators of the Test Have to Be Blinded? The index test and the reference test must be carried out completely independently by two different persons. In other words, both persons should administer a different test to the same patient. It is important that the assessor of one test is not aware of the result of the other test. The assessors of one test must therefore be blinded to the result of the other test. For example, if it is known that the index test has a positive result (i.e. the diagnosis in question is indeed present according to the index test),

126

7  What May Be the Matter with the Patient: How to Critically Appraise Articles…

this probably will influence the assessor of the reference test. The assessor is then more likely to judge the reference test positive as well. This leads to bias, which overestimates the true agreement between the two tests.

7.6 In What Situation Is the Requirement Dropped That Both Tests Be Performed? In That Case, How Can the Comparison Be Done Properly? Both an index test and a reference test are performed on each study participant. In some cases, the reference test may be too expensive, too invasive or in some other way too burdensome to carry out for everybody. In these cases, the index test may be carried out for everybody, but the reference test may only be carried out for some of the study participants. It is important that researchers do not let this decision depend on the result of the index test (Sect. 12.43) [2].

7.7 What Is ‘Selective Dropout’? What Should I Look for in Assessing Dropout? Another aspect that helps to determine the validity of the diagnostic study is the number of dropouts that do not make it through the entire study protocol. The authors should report the number of dropouts from the study and describe the reasons for the dropout. Selective dropout is the phenomenon that mainly study participants drop out who have certain characteristics (e.g. few complaints). After all, patients who have just a few complaints have less interest in taking two different diagnostic tests than patients with more complaints. Consequently, not every participant has the same chance to drop out. If significantly more patients with the same specific characteristic (e.g. having few complaints) drop out, this dropout is selective with respect to that specific characteristic. This can lead to relevant differences in patient characteristics between healthy and non-healthy participants. Selective dropout can therefore lead to serious differences between the group of healthy participants and the group of non-healthy participants. This makes it easier for the test to distinguish between the participants who do have the diagnosis and those who have not, while in daily practice, this is usually not this simple.

7.8 How Can I Assess the Value of the Measuring Instrument Under Study? In a valid research, it is important to determine how well the measuring instrument can predict the diagnosis. This is called the diagnostic accuracy. This accuracy is important, for example, to prevent as much as possible that women who are screened for breast cancer get a positive test result while they do not have cancer. Conversely, it is also important to prevent as many patients as possible from receiving a negative

7.9  How Can I Assess Whether I Can Apply the Researched Diagnostic Instrument…

127

test result even though they do have the disease. To determine how well a measuring instrument works, the sensitivity and specificity of the instrument must be assessed. These values are determined by comparing the new test with a reference test. For example, the sensitivity of the DOS indicates the percentage of individuals with delirium in whom the test should therefore show that delirium is present. These individuals are correctly identified as being ill by the DOS test and are referred to as true positives. A sensitivity of 92% of the DOS means that of 100 patients with delirium, the DOS correctly indicates that 92 of them have delirium. The DOS therefore incorrectly indicates that eight out of 100 patients do not have delirium [3]. These are the false negatives. The specificity of the DOS indicates the percentage of persons without delirium (for whom the test should therefore indicate that there is no delirium present) who are correctly designated by the test as not being ill: the true negatives. A specificity of 82% of the DOS means that, of 100 patients without delirium, the DOS correctly indicates that 82 people do not have delirium. The DOS therefore incorrectly indicates that 18 patients have delirium: the false positives [3]. In addition to sensitivity and specificity, it is also important, as with intervention research (see Chap. 10), how precise the values found are. This is expressed in the 95% confidence interval (Sect. 12.44). For the healthcare professional, sensitivity and specificity are difficult to use in practice. They predict the chance of a certain test result in the presence or absence of a certain disease. That is why healthcare professionals prefer to use the predictive value. These values reflect the probability that a patient, with a certain test result, actually has the disease. Suppose that a nurse wants to know how high the probability is that a patient with a positive test result will actually have delirium. We call this the positive predictive value. The negative predictive value, on the other hand, indicates the likelihood that a patient with a negative DOS result will not actually have delirium (Sect. 12.45).

7.9 How Can I Assess Whether I Can Apply the Researched Diagnostic Instrument to My Patients? If the study appears to have been conducted validly and the study has demonstrated sufficient diagnostic accuracy, the nurse determines whether the measurement instrument is applicable in practice. Her consideration is based on several aspects: • Appropriateness of this test for the patient. The test is suitable if the patient characteristics of the patient are sufficiently similar to the characteristics of the patients in the study in which the accuracy of the test was investigated. • Purpose of the test. Does the test fit the purpose for which it will be used in the patient? Suppose the result of the test does not determine the use of an intervention. Then the question is whether it is useful to test. After all, the test result does not influence the decision on a consecutive intervention.

128

7  What May Be the Matter with the Patient: How to Critically Appraise Articles…

• Burden of the patient by the test. To what extent is the psychological or physical burden placed on the patient by the performance of the test, balanced with the purpose and importance of the information provided by the test?

Glossary Bias (in scientific health research)  In scientific health research, bias means a systematic deviation from the underlying truth because of a feature of the design or conduct of a research study (e.g. overestimation of a treatment effect because of failure to randomise) [4]. Blinding  Measure to ensure that one or more groups involved in a study (patients, health professionals, treatment providers, effect assessors) do not know whether patients are assigned to the experimental group or the control group. Confidence interval  The confidence interval (CI) indicates the precision of the calculated point estimate of the effect measure. It is common practice to calculate the 95% confidence interval around the point estimate. The 95% confidence interval indicates that if the study is validly conducted, there is a 95% chance that the actual effect lies within this confidence interval. A confidence interval is needed because research results are based on a sample and not the entire population. The confidence interval indicates that there is a 95% chance that the effect in the entire population lies within the interval. Diagnostic accuracy study  Scientific research that focuses on the question: how accurately does the diagnostic measuring instrument measure the presence or absence of a specific diagnosis? Diagnostic question  Question that aims to make a better assessment of the patient’s state of health. The question can be directed towards the nursing, medical and/ or psychological condition of the patient. Diagnostic questions concern the presence or absence of (characteristic) signs, symptoms, complaints and behaviour of the patient. Diagnostic test (diagnostic measuring instrument) Test or measuring instrument that helps to make a diagnosis. The test or measuring instrument increases certainty about the patient’s condition. This is preferably done with the test or instrument that makes the fewest errors. False negative  In diagnostic research, a person is incorrectly labelled as not having the disease on the basis of a diagnostic test. The diagnostic test then indicates that the person is not ill, while he or she does have the disease. False positive  In diagnostic testing, a diagnostic test is used to falsely label a person as having the disease. The diagnostic test then indicates that the person is ill, while he does not have the disease. Gold standard  In diagnostic testing, a diagnostic test to be examined is compared with the test that is best at the time because it makes the fewest errors. An error means that a test incorrectly indicates that a person has or does not have the

7.9  How Can I Assess Whether I Can Apply the Researched Diagnostic Instrument…

129

diagnosis (see ‘false positive’ and ‘false negative’). The test that currently makes the fewest errors is called the ‘gold standard’. Because in many cases no gold standard is available, the best performing test is sought, which is then called a ‘reference test’. Index test  In scientific diagnostic research, a newly designed diagnostic test is tested. In the study, this new diagnostic test is called the index test. Testing takes place by comparing the performance of the index test (the number of errors it makes in making the diagnosis) with the performance of another test. That other test is called the reference test or ‘gold standard’. Predictive value  Positive predictive value: the value that represents the chance, for a diagnostic test, that the patient will actually have a certain condition, after the test has indicated that the patient indeed has this condition. Thus, if the positive predictive value of a test is 0.91, a person for whom the test indicates that he has the condition has a 91% chance of actually having that condition. Negative predictive value: the value with which a diagnostic test reflects the chance that the patient really does not have a certain condition, after the test has indicated that the patient would not have this condition. So, if the negative predictive value of a test is 0.98, then a person for whom the test indicates that he does not have the condition has a 98% chance of not having the condition. Random sample  Sampling in which each person from the source population has an equal chance of being in the sample and hence of participating in the study. The sample is selected from the source population entirely on the basis of chance. Reference test  In diagnostic testing, a diagnostic test to be examined is compared to the test that is best at the time because it makes the fewest errors. An error means that a test incorrectly indicates that a person has or does not have the diagnosis (see false positive and false negative). The test that makes the fewest errors at that time is called the ‘gold standard’. Because in many cases, no gold standard is available, the best performing test is sought, which is then called a ‘reference test’. Selective dropout  Is the phenomenon that the patients who discontinue their participation in the study share a specific characteristic (e.g. they already feel better). This may lead to a specific subgroup of patients leaving the study, the dropout (or attrition) is selected with respect to that characteristic. Selective dropout may lead to bias and thus to overestimation or underestimation of the true effect. Sensitivity  The sensitivity of a diagnostic test is the proportion (percentage) of individuals with a positive test result among those with the condition the test is supposed to diagnose: the true positives. Specificity  The specificity of a diagnostic test is the proportion (percentage) of people with a negative test result among those who don’t have the condition the test is supposed to diagnose: the true negatives. True negatives  Persons who, on the basis of the results of a diagnostic test, are justifiably declared not to be ill (because they do not have the disease). True positives  Persons who, on the basis of the results of a diagnostic test, are justifiably declared to be ill (because they actually have the disease).

130

7  What May Be the Matter with the Patient: How to Critically Appraise Articles…

Validity  Absence of systematic errors. For example, the validity of a measurement instrument is the extent to which a measurement instrument measures what it is intended to measure [4].

References 1. Furukawa TA, Straus SE, Bucher HC, Agoritsas T, Guyatt G. Diagnostic tests. In: Guyatt G, Rennie D, Meade MO, Cook DJ, editors. Users’ guides to the medical literature. A manual for evidence-based clinical practice. 3rd ed. New York, NY: McGraw Hill Education; 2015. 2. Leeflang MMG, Scholten RJPM, Aertgeerts B.  Diagnose. In: Scholten RJPM, Offringa M, Assendelft WJ, editors. Inleiding in evidence-based medicine. Klinisch handelen gebaseerd op bewijsmateriaal. 4th ed. Houten: Bohn Stafleu van Loghum; 2014. 3. Wong CL, Holroyd-Leduc J, Simel DL, Straus SE. Does this patient have delirium? Value of bedside instruments. JAMA. 2010;304:779–86. https://doi.org/10.1001/jama.2010.1182. 4. Guyatt G, Rennie D, Meade MO, Cook DJ. Users’ guides to the medical literature. A manual for evidence-based clinical practice. 3rd ed. New York, NY: McGraw Hill Education; 2015.

8

What May Be the Cause: How to Critically Evaluate Articles on Etiological Factors?

During the coffee break, my colleague John told me that he had noticed that many patients have an unhealthy lifestyle. Then the conversation turned to the question of what our own lifestyle is like. Do we smoke, how much alcohol do we drink, how healthy do we eat, are we overweight and do we exercise enough? We all made a quick list of our scores on these five points. It turned out that most of us did not have a good lifestyle on three of the five subjects. And we are all nurses or nursing students! But we also asked ourselves what the consequences are, do you get sick sooner, do you get heart disease sooner, or cancer? At the end of my shift, I started searching through PubMed. In the process, I found a cohort study on the relationship between a healthy lifestyle and the number of years of life expectancy free of cancer, cardiovascular disease, and type 2 diabetes. The researchers found that for women, a healthy lifestyle (with up to one lifestyle-related risk factor instead of the three that many of the nurses in my team had) resulted in an average of 5.3 years longer life without cancer, cardiovascular disease and DM-2. For men, 5.8 years. (The full article by Li et al. can be found online, https://doi.org/10.1136/bmj.l6669).

8.1 In What Kind of Sources Can I Find the Best Evidence About Etiological Factors? The best evidence on etiological factors is the systematic reviews with meta-­ analyses, which combine previous original studies on the same etiological factors. The research designs for original studies of etiology are the cohort study and the case-control study. Both study designs are referred to as observational studies. This means that the researcher only ‘observes’ the participants in the study. This is

With contribution by M. Offringa. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 J. Dobber et al., Clinical Reasoning and Evidence-Based Practice, https://doi.org/10.1007/978-3-031-27069-7_8

131

132

8  What May Be the Cause: How to Critically Evaluate Articles on Etiological Factors?

observation in a broad sense: the researcher collects all kinds of information, for example, by taking measurements, having people fill in questionnaires and so on. This ‘observing only’ means that the researcher does not play an active role in what happens, that is, the researchers do not carry out an intervention [1].

8.2 What Does a Cohort Study of Etiological Factors Involve? In cohort studies, sometimes participants are followed by the researchers for years. Researchers may observe participants at regular intervals (e.g. once every 2 years) through questionnaires, tests and physical examinations while participants continue to live their normal lives. In this way, one or more predetermined possible etiological factors can be observed for years, and therefore, it can be determined whether the participants who are exposed to these factors are more likely to develop a predetermined health condition than are individuals who are not exposed to this factor or are barely exposed. For example, Li et al. [2] examined two groups of study participants. The first group consisted of 73,196 American female nurses, who completed a comprehensive lifestyle questionnaire biennially beginning in 1976. The researchers used data from the last 28 years. The second group consisted of 38,366 male health professionals, who were followed by researchers for 22 years. During that time, participants were given a comprehensive questionnaire about their current lifestyle and health condition every 2  years. The researchers analysed the influence of five lifestyle-­related risk factors, namely smoking, physical activity, healthy eating, alcohol consumption, BMI. The study participants self-reported if they were diagnosed with cancer, heart attack or stroke or diabetes. This self-reporting was checked by researchers. In this way, it was possible to determine whether lifestyle influences the duration of a life in good health. The researchers divided the participants into five groups. A group with all five lifestyle risk factors, groups with four, three, two risk factors and finally the ‘healthy-living group’, which is the group with zero or one lifestyle risk factor. By dividing the cohort into groups according to the extent to which their lifestyle exposed them to risk of the three diseases (this is called ‘according to the extent to which participants were exposed to the possible etiological factor’), it became clear to what extent a person’s risk of cancer, type 2 diabetes and/or cardiovascular disease increases as the person had more lifestyle-related risk factors (Sect. 12.46). However, there are some important disadvantages to the cohort study as a research form. The first disadvantage is that, in addition to the etiological factors being studied, many other determinants may play a role in the occurrence of the disorder. In the study by Li et al., these other possible determinants include environmental pollution, stress, gender and sleep quality. Researchers know this of course and try to measure these determinants as well. However, this is not always possible.

8.3  What Does a Case-Control Study of Etiologic Factors Involve?

133

In the study by Li et al. [2], the gender of all participants is not difficult to measure, and this was done. The researchers also distinguished the life expectancy of women from the life expectancy of men in their report. However, exposure to environmental pollution is much more difficult to measure. It is likely that in many cases, the participants themselves do not know whether they have been exposed to such environmental pollution. And stress and sleep quality were probably not asked for in the biennial questionnaire, so that these data were not available. As a result, such a determinant can influence the results of the cohort study without the researchers’ knowledge. The second drawback is the length of time required to detect these types of etiological factors. It can take a very long time before the effect of the studied etiological factor manifests itself in a disease. Consequently, the cohort must be followed a very long time until the effects of this etiological factor can occur. Li et al. [2] used two existing cohorts for their study. Neither group had been assembled specifically for their study. They had already been formed in 1976 (first only on diet, later also on other lifestyle factors) and in 1986. At the time, they had been formed from the perspective to be useful in future health research on the influence of diet on health and disease. In many cases, however, this type of data on the etiological factors to be investigated is not yet available, and a new cohort has to be compiled.

8.3 What Does a Case-Control Study of Etiologic Factors Involve? The case-control study does not have the disadvantage of a long time span, because in case-control studies, the process is reversed. Whereas cohort studies start with the etiological factor and then wait, for the occurrence or non-occurrence of the disease (prospective study), case-control studies go back in time: from patients who already have the disease to the extent to which they have been exposed in the past to predetermined etiological factors (retrospective study) (Sect. 12.47). A case-control study thus begins with patients who already have the disease. These patients are checked to see to what extent they have been exposed in the past to the etiological factor in which the investigators are interested. Thus, patients with oesophageal cancer may be asked whether they have consumed alcohol in the past, and if so, how much. At the same time, the researchers will study the other characteristics of the group of persons with oesophageal cancer: how old are they when the disease is diagnosed? Is the disease more common in men than in women? What type of lifestyle have the patients had in recent years? And so on. Based on these characteristics, for comparison, a control group of people with the same characteristics but no oesophageal cancer is sought. The control group is then also checked for their past exposure to the etiological factor under investigation (in this example: alcohol consumption). By comparing the exposure of the patient group and the control group to the possible etiological factor, conclusions can be drawn concerning the role of this factor in the etiology of the disease.

134

8  What May Be the Cause: How to Critically Evaluate Articles on Etiological Factors?

An important disadvantage of case-control studies is that it strongly depends on the memory of the research participants. In the example, it is very doubtful whether everyone can remember how much alcohol on average they used to drink in the past.

8.4 How Can I Determine the Quality of the Cohort Study? The quality of research is determined by the extent to which errors have been avoided. A study is valid when it is free of systematic errors. Systematic errors (or: bias) in the study influence the results, as a result of which the results do not reflect the actual values. The following questions are important in cohort studies (see also Chap. 12); this checklist is based on the checklist of Van Puijenbroek et al. [3]: 1. Have the study participants clearly been described, and are the groups comparable? 2. How high is the risk of selection bias? 3. How high is the risk of information bias? 4. Does the period to follow-up last long enough to measure the effect of the studied factor on the development of the studied condition? 5. Have, during the course of the cohort study, not too many study participants dropped out? 6. Was the analysis adjusted for significant risk factors? (Sect. 12.63.1)

8.5 How Do I Know If the Two Groups Are Comparable? In order to properly determine the influence of the possible etiological factor, the presence of that factor should vary in the studied group. This variation may consist of the presence or absence of the factor, but there may also be differences in the extent to which study participants are exposed to it. For all study participants, the level of exposure should be clearly presented in the article. Furthermore, the criteria used by the investigators to diagnose the condition should be clearly described. Finally, study participants also have other characteristics. These characteristics must also be represented. It is especially important that the presence of factors in the study participants that increase the risk of the disorder is described. After all, these risk factors, as well as the possible etiological factor under investigation, can influence the occurrence of the disorder. For example, in the study by Li et al. [2], heredity is such a factor. Usually, much of this information is given in the first table of the article: the characteristics of the study participants, divided in groups based on the level of exposure to the factor. In the article by Li et al., this is also the case. Table 1 shows the extent to which participants had a family history of diabetes, cancer, myocardial infarction [2]. So, Table 1 provides the readers with the information to check the comparability of the two (or more) groups.

8.7  What Is Information Bias? How Can I Estimate the Risk of Research Results Being…

135

8.6 What Is Selection Bias? How Can I Estimate the Risk of Bias in Research Results? Selection bias means that bias in the study results occurs due to systematic errors in the selection of study participants. In a cohort study, the investigators must first compile their study population (the persons participating in the study) and then measure exposure to the possible etiological factor. This will then naturally produce a group of study participants who have been exposed to that factor and a group of participants who have not been exposed to that factor. However, if exposure is measured first and people are then admitted to the study population or not, the groups have already been formed by the researcher. This causes selection bias. The authors of the article should clearly describe how the groups were composed. If exposure was measured after the study population had been formed, there is a low risk of selection bias. If exposure to the possible etiologic factor was measured before entry into the cohort study, there is a high risk of selection bias. If it is not clear from the article what the procedure was, there is an unclear risk of selection bias. The study by Li et al. [2] used existing cohorts, from which five groups were automatically formed depending on the number of unhealthy lifestyle factors. The process was therefore as described above. However, some selection bias may have slipped into Li et al.’s study in a different way. After all, it is to be expected that people who work as healthcare professionals live a proportionally healthier life than other people. And this is not just limited to the five lifestyle factors measured, but is to be expected in other areas as well.

8.7 What Is Information Bias? How Can I Estimate the Risk of Research Results Being Biased by Information Bias? Information bias means that bias in the study results occurs due to systematic errors in the measurement of exposure to the possible etiological factor, in the measurement of the condition or in the measurement of other factors influencing the condition. Therefore, the measurement of the potential etiological factor and of the condition needs to be performed in exactly the same manner for each study participant. If this is not done in the same way, the information from participants cannot be compared with each other. Working with standardised questionnaires, measuring instruments, diagnostic tests and objective criteria is an important way of preventing information bias. It is preferable that the effect assessor (the person measuring whether the condition has occurred) does not know whether the study participant has or has not been exposed to the possible etiological factor (blinding). By providing a clear description of the measurement and assessment procedures in the article, the authors enable the reader to determine whether there is a low or high risk of information bias. If the description is unclear or missing altogether, there is an unclear risk of information bias. Li et  al. [2] clearly describe the

136

8  What May Be the Cause: How to Critically Evaluate Articles on Etiological Factors?

measurement and assessment procedures. A potential weakness in their procedures is that all participants report on their lifestyle and health via the questionnaires themselves; this self-reporting is not done by everyone in exactly the same way. However, this is unavoidable in large-scale studies involving so many participants over so many years.

8.8 How Do I Know If the Period of Time Until Follow-Up Is Long Enough? The time required by etiological factors to exert sufficient influence and (help) bring about the health condition varies from factor to factor and from condition to condition. The follow-up period should include at least this period of action for the factor. Preferably even longer, as the time to act may also vary from person to person. In addition, it is common that the exposure of study participants to these factors changes over time, for example, increases or decreases. This may also influence the duration of the follow-up required. The authors must make a plausible case that the follow-up period they have chosen is indeed long enough for the occurrence of the health outcome studied for that possible etiological factor.

8.9 Why Does the Intermediate Dropout of Participants Threaten the Validity of the Study? How Can I Determine Whether Too Many Participants Have Dropped Out? The dropout of study participants is usually not equally distributed over the groups. For example, it could be that more participants in the exposed group drop out than in the unexposed group. This makes the groups less comparable. In the study by Li et al., this would be the case if, for example, mainly people with an unhealthy lifestyle in many areas ended their participation in the study prematurely, and there were hardly any dropouts among the participants with a very healthy lifestyle. This phenomenon is called ‘selective dropout’ or ‘selective attrition’. A rule of thumb is that no more than 20% of the study participants may drop out. However, it is important to check that this is not a selective dropout.

8.10 What Is ‘Adjustment for Significant Risk Factors’? How Can I Check Whether These Factors Have Been Controlled for? Besides the possible etiological factor that is being investigated, several other determinants may play a role in the occurrence of the disease. In the occurrence of cancer, type 2 diabetes and cardiovascular diseases, these determinants are, for example,

8.11  Do the Same Requirements Apply to Determining the Validity of…

137

stress, genetic predisposition, sleep quality, environmental pollution and aspirin use. Researchers know this, of course, and try to measure these determinants as well. In determining whether the possible etiological factor is a true etiological factor, the researchers must distinguish the influence of the studied etiological factor as much as possible from the influence exerted by other determinants. If the factor under study indeed turns out to be a real etiological factor, this distinction from other determinants is also important in order to determine how strongly this etiological factor may contribute to the development of the disease. For example, when investigating the influence of the five lifestyle factors on the occurrence of the three diseases, researchers must distinguish the influence of these five lifestyle factors from gender and genetic predisposition. In this study, the influence of gender on the onset of the diseases was distinguished by analysing the men and women in separate groups. This allows the researchers to calculate and present the risks of the five lifestyle factors separately for men and women. In addition, Li et al. describe (in the section Statistical analysis) for which factors they have adjusted their analysis (adjusted for). In the tables in which the results are presented, the authors also indicate which results have been controlled for (or ‘corrected for’, or ‘adjusted for’) [2].

8.11 Do the Same Requirements Apply to Determining the Validity of a Case-Control Study? Questions 1, 2, 3 and 6 of the checklist for cohort studies also apply to case-control studies. However, questions 4 and 5 differ. The complete list of questions for a case-­ control study is as follows: 1. Have the study participants clearly been described, and are the groups comparable? 2. How high is the risk of selection bias? 3. How high is the risk of information bias? 4. Are the cases in which the disease occurred in the case-control study new cases? 5. Was misclassification avoided in the case-control study? 6. Was the analysis adjusted for significant risk factors? (Sect. 12.64)

8.12 What Is Meant by New Cases? Why Is It Important That the Cases Are New? The starting point for patient monitoring is patients in whom the disorder has been diagnosed. New cases are patients who have been diagnosed recently. Such patients are very similar in terms of important characteristics, in contrast to patients who have had the condition for a long time. We say that the former group of patients is more homogeneous and the latter group more heterogeneous. Homogeneity has an important advantage. For example, in new cases, the (measurement of the) possible

138

8  What May Be the Cause: How to Critically Evaluate Articles on Etiological Factors?

etiological factor is less influenced by changes initiated by the disease process. For example, patients who have been diagnosed with cancer tend to eat more healthily, to smoke less and to drink less. New cases are also referred to as incident cases and existing cases as prevalent cases. Incidence is the number of newly detected cases in a population (e.g. all Australian citizens) within a certain period of time (often a calendar year). Prevalence is the number or percentage of people in a population who have the disease at a given time or in a given period.

8.13 What Is Meant by Misclassification? How Can I Find Out If There Is a High Risk of Misclassification? Misclassification occurs when a study participant is exposed to the potential etiologic factor under study, but is recorded as unexposed, or conversely, when a person is not exposed but is recorded as exposed. Misclassification also occurs when a study participant is classified as sick when he is not sick, or as not sick when he is sick. Misclassification is a measurement error and should therefore be regarded as a form of information bias. Measurement errors of this kind can lead to an overestimation or underestimation of the effect of the possible etiological factor and thus distort the results of the study. Such a measurement error may arise, for example, if the study participants without the disease are less well assessed for past exposure to the potential etiological factor than the study participants with the disease. Usually, it is difficult to ascertain whether there has been any misclassification. It is important that the study investigators work with standard procedures for all measurements, using criteria determined in advance. In addition, the researchers could check the classification of all research participants in order to detect and correct misclassification. By including the exact procedures that were followed in the article about the study, the reader can to some extend estimate the size of the risk of misclassification.

8.14 How Can I Assess the Magnitude of the Effect of the Etiological Factor? The effects found are expressed as a measure of effect. This is often expressed in terms of risk, for example, a relative risk (RR), an odds ratio (OR) or the difference between the various groups and a confidence interval. Two aspects are important: the size of the effect, expressed in the effect size, and the precision of the effect, expressed in a confidence interval. The magnitude of the effect is shown with a point estimate (also known as ‘effect size’). The point estimate represents the effect found in the study. Li et al. [2] found in their cohort study that women who live a healthy life (zero or one unhealthy lifestyle factor) can expect to live on average 10.6 years longer in health than women

8.15  How Can I Determine If the Point Estimate Is Really the Actual Effect?

139

with five unhealthy lifestyle factors and 5.3  years longer than women with three unhealthy lifestyle risk factors. Men who live a healthy life live on average 7.6 years longer than men with five unhealthy lifestyle factors, free of cancer, cardiovascular disease and type 2 diabetes. These are the point estimates. These estimates reflect the size of the effect of healthy living. The effect size used is the Mean Difference: men with five unhealthy lifestyle factors have an average expectation of 23.5 healthy years of life at age 50. Men with a healthy lifestyle have an expectation of 31.1 healthy years of life at age 50. The difference between these averages is therefore 31.1–23.5 years, which is 7.6 years (Sect. 12.70).

8.15 How Can I Determine If the Point Estimate Is Really the Actual Effect? For this purpose, the confidence interval (CI) around the point estimate indicates the accuracy of the effect found (or the precision). The point estimate found in the study is the best estimate of the true effect of the possible etiological factor. However, because the study was conducted in a sample of the population, the effect found was influenced by the composition of the sample. In other words, if there had been other individuals in the sample, a slightly different effect would probably have been found. That effect would be close to the present effect size, but would probably deviate slightly from it (e.g. not the mean difference of 10.6 years as found now, but a mean difference of 10.4 years for women with a healthy lifestyle). The phenomenon whereby the point estimate varies depending on the composition of the sample is called sampling variability. This means that it is unlikely that a study will find the exact effect. Therefore, in the article, the authors should include a confidence interval for every point estimate. It is customary to speak of a 95% confidence interval. In the cohort study by Li et al. [2], the researchers found (at the point estimate 10.6) a 95% confidence interval of 10.0–11.3. In articles, this is usually represented as follows: Mean Difference = 10.6 (10.0–11.3). It means that if the study is properly (valid) carried out, there is a 95% chance that the actual effect of the possible etiological factor falls within the confidence interval. Thereby, the true effect is likely to be close to the point estimate and less likely to be close to the extremes of the confidence interval.

8.16 What Does the Confidence Interval Say About Statistical Significance? By looking carefully at the confidence interval, it can be determined whether or not the study has shown a statistically significant effect. As an example, we will again use the mean difference between the women with a healthy lifestyle and the women with five unhealthy lifestyle factors.

140

8  What May Be the Cause: How to Critically Evaluate Articles on Etiological Factors?

The effect of the possible etiological factor healthy lifestyle on the outcome life expectancy free from cancer, type 2 diabetes and cardiovascular disease is expressed as a mean difference. There are three possibilities. 1. A mean difference of 0 means that the factor ‘healthy lifestyle’ makes no difference to the risk of the three diseases. The value 0 is the neutral value of the difference. 2. A mean difference of less than 0 means that the ‘healthy lifestyle’ factor leads to a lower life expectancy free of the three conditions. After all, a difference of (for example) −3.6 would mean that women with a healthy lifestyle (zero or one unhealthy lifestyle factor) can expect to live on average 3.6  years shorter in health than women with five unhealthy lifestyle factors. 3. A mean difference higher than 0 means that the factor ‘healthy lifestyle’ leads to a longer life expectancy free of the three diseases. Only if a difference of 0 is found for ‘healthy lifestyle’, or if 0 falls within the confidence interval, is ‘healthy lifestyle’ not an etiological factor. An example of a confidence interval that includes the 0 would be the confidence interval −1.2 to 3.5. Thus, if this would be the case, the lowest value of the confidence interval is below zero, and the highest value of the confidence interval is above zero. The study by Li et al. [2] found a point estimate of Mean Difference = 10.6. This would mean that women with a healthy lifestyle (zero or one unhealthy lifestyle factor) have a high chance of living free from all three conditions for approximately 10 years longer than women with five unhealthy lifestyle factors. This shows that an unhealthy lifestyle is an etiological factor for the earlier onset of cancer, type 2 diabetes and cardiovascular disease. A Mean Difference = 10.6 and a confidence interval of 10.0–11.3 were found. This means that with 95% certainty, the real effect lies between a longer healthy life expectancy of 10.0 and 11.3 years without any of the three conditions. The real effect could therefore also be, for example, on average 10.1 years longer, but also on average 11.1 years. The result Mean Difference = 10.6 (10.0–11.1) shows increased life expectancy without the three conditions for every value that falls within the confidence interval. The confidence interval shows that a life with the five unhealthy lifestyle factors is indeed an etiological factor for earlier occurrence of (one or more of) the three diseases. Because this unhealthy lifestyle implies an increased risk, it is also referred to as a risk factor for early-onset cancer, type 2 diabetes and cardiovascular disease. The rule for interpreting the significance through using the confidence interval is: If the neutral value of the measure of effect measure (in this case Mean Difference) is enclosed in the confidence interval, no statistically significant effect of this factor has been demonstrated. If the neutral value of the measure of effect measure (in this case Mean Difference) is not enclosed in the confidence interval, then a statistically significant effect of this factor has been demonstrated.

8.18  What Is the Difference Between Association and Causation?

141

8.17 How Do I Know If I Can Use the Results for My Patients? The purpose of reading and appraising research articles is to use evidence in clinical reasoning and therefore in clinical decision-making. Clinical decision-making is always about a specific patient or about the policy to be implemented in a specific patient group. An important component of involving evidence in clinical reasoning is the extent to which the study participants correspond to the nurse’s specific patient or patient group. In other words, are the findings from the cohort study generalisable to one’s own practice? Generalisability relates to the extent to which the study results also apply to patients who did not participate in the study. The characteristics of these patients (for whom the study results may be used) should match the characteristics of the study participants as closely as possible. Usually, the first table of an article provides information about the characteristics of the study participants. By comparing the data from this table with one’s patient, one can check whether the characteristics of one’s patient are sufficiently similar to those of the study participants. In the study by Li et al. [2], the possible selection bias should be taken into account when assessing generalisability: this group of study participants consisted exclusively of healthcare professionals. In addition, the mechanism of action of the etiological factor plays an important role. To what extent can the pathophysiological mechanism triggered by an unhealthy lifestyle in Europeans, Africans, Australians or Asians be different from that in Americans in the study by Li et al. [2]? If it is plausible that the mechanism of action of the factor has the same effect on the patient, this increases the generalisability of the study results. It should be taken into account that pathophysiology and also psychosocial processes can vary greatly between persons. Some persons are much more sensitive to a factor than others.

8.18 What Is the Difference Between Association and Causation? It is important to distinguish between the concepts of association and causation. By association is meant that two variables are related. Take, for example, the variables ‘socio-economic status’ and ‘overweight’. Overweight occurs more often in people with a lower socio-economic status (SES) and less in people with a higher socio-­ economic status. Low SES therefore seems to be a cause of overweight. However, it appears that people with low SES eat less healthily and exercise less than people with high SES.  These factors are probably the cause of the fact that overweight occurs more often in people with low SES. Therefore, low SES is indeed associated with overweight, but overweight is not caused by it. In contrast, there is a causal link between smoking and lung cancer: smoking is a cause of lung cancer.

142

8  What May Be the Cause: How to Critically Evaluate Articles on Etiological Factors?

Glossary Bias (in scientific health research)  In scientific health research, bias means a systematic deviation from the underlying truth because of a feature of the design or conduct of a research study (e.g. overestimation of a treatment effect because of failure to randomise) [4]. Blinding  Measure to ensure that one or more groups involved in a study (patients, health professionals, treatment providers, effect assessors) do not know whether patients are assigned to the experimental group or the control group. Case-control study  Retrospective research designed to determine the association of a specific risk factor and a specific disease. The investigator selects two samples, patients with the disease and a sample with a similar group of persons without the disease. The researchers then look back in time to measure the frequency of exposure to a possible risk factor in both groups [1]. Cohort  A group of people who have specific characteristics in common at the time the group is formed and who are subsequently followed in order to determine whether specific outcomes have developed in them. Belonging to a cohort is of unlimited duration, but can be terminated when someone no longer wants or is no longer able to provide data (e.g. due to death) or when the outcome for which the cohort was composed (e.g. a particular disease) has developed in that person. Cohort study  Research in which the groups (cohorts) to be followed are classified according to the presence of, or exposure to, a specific determinant. The cohorts are followed during the observation period and compared in order to determine whether differences in determinant status lead to differences in specific health outcomes. Confidence interval  The confidence interval (CI) indicates the precision of the calculated point estimate of the effect measure. It is common practice to calculate the 95% confidence interval around the point estimate. The 95% confidence interval indicates that if the study is validly conducted, there is a 95% chance that the actual effect lies within this confidence interval. A confidence interval is needed because research results are based on a sample and not the entire population. The confidence interval indicates that there is a 95% chance that the effect in the entire population lies within the interval. Determinant  Factors or characteristics that are related in some way to the health outcome being studied. For example, smoking is a determinant for lung cancer, body weight for diabetes and cannabis use for psychosis. Effect size (or: point estimate)  The magnitude of the effect as found in a particular study. The point estimate is the best estimate of the actual effect based on that study. But because the study is based on a sample, it is unlikely that the point estimate is exactly the actual effect in the entire population. Etiological factor  Factor that (partly) causes the health problem. Exposure  Means ‘coming into contact with’. Exposure to asbestos therefore means: coming into contact with asbestos.

8.18  What Is the Difference Between Association and Causation?

143

Generalisability (or: external validity)  The extent to which the results found in the sample of a study are also valid for individuals with similar health problems and characteristics who did not participate in the study (e.g. a question might be: are these study results valid for individuals with similar health problems but who live in a different continent or who have a different culture?). Incident cases  All persons who change in health status (newly diagnosed with the disease). Information bias (in an RCT, cohort or case-control study)  Bias in the study results caused by bias in the information because of measurement errors. Information bias can occur, for example, when a number of study participants are assigned to the wrong exposure group or when an incorrect outcome is registered (misclassification). Measure of effect  Measure expressing the effect, for example, mean difference (MD), relative risk (RR), risk difference (RD) and number needed to treat (NNT) Mechanism of action  The way in which the etiological or prognostic factor brings about the health change in the patient. The mechanism of action explains which and how physiological or psycho-social processes deviate from their normal course and how they affect the health status. Observational studies Form of research in which the researcher exclusively observes the study participants for a short or longer period of time. This observation can also take place via written questionnaires. Observational research is therefore non-experimental research (because no intervention is carried out), the researcher tries to limit his own influence as much as possible. The aim is to observe and map out the natural process with regard to the research question. Examples of observational research are the cohort study and the case-­ control study. Point estimate (or: effect size)  The magnitude of the effect as found in a particular study. The point estimate is the best estimate of the actual effect based on that study. But because the study is based on a sample, it is unlikely that the point estimate is exactly the actual effect in the entire population. Prospective research  Research in which research participants are followed and measured ‘over time’. This is in contrast to retrospective research, in which one looks ‘back in time’ to measure the presence of certain exposure factors in the research participants in the past. Relative risk (RR)  Association measure that expresses the ratio of the probability of a particular outcome of the intervention group to that of the control group (also: Relative Risk; Risk Ratio) The RR answers the question: how many times greater or smaller is the probability of the outcome after the experimental intervention than after the control intervention? • An RR of 1.0 means that there is no difference between risk or probability of the outcome. The value of 1.0 is the neutral value of a RR • An RR between 0 and 1.0 means that the intervention group has a lower risk or probability of the outcome than the control group

144

8  What May Be the Cause: How to Critically Evaluate Articles on Etiological Factors?

• An RR higher than 1.0 means that the intervention group has a higher risk or probability of the outcome than the control group. Take a study in which the intervention consists of a specific diet. The outcome is: loss of at least 5 kg within 2 months. Suppose that 10% of the participants in the control group achieved the outcome • RR = 2.0 means that the group that received the experimental intervention was twice as likely to reach this outcome as the control group. So, if the control group had a 10% chance of losing 5  kg within 2  months, the intervention group has 2 × 10% = 20% chance of losing 5 kg within 2 months • RR = 1.0 therefore means that the intervention group has 1 × 10% = 10% chance of losing 5 kg within 2 months. With an RR of 1.0, both groups have an equal chance of the outcome, there is no difference • RR = 0.8 means that the intervention group has a 0.8 × 10% = 8% chance of achieving the outcome. Retrospective research  Research in which one looks ‘back in time’ in order to measure the presence of certain exposure factors in the research participants in the past. This is in contrast to prospective research, in which research participants are followed and measured ‘over time’. Risk factor  Factor that, if the health problem is not already present, increases the risk of developing that health problem. Selection bias (in an RCT or cohort or case-control study)  Bias in the study results caused by systematic errors in the selection of the participants included in the study. Selective attrition (or: selective dropout)  Is the phenomenon that the patients who discontinue their participation in the study share a specific characteristic (e.g. they already feel better). This may lead to a specific subgroup of patients leaving the study, the dropout (or attrition) is selected with respect to that characteristic. Selective dropout may lead to bias and thus to overestimation or underestimation of the true effect. Selective dropout (or: selective attrition)  Is the phenomenon that the patients who discontinue their participation in the study share a specific characteristic (e.g. they already feel better). This may lead to a specific subgroup of patients leaving the study, the dropout (or attrition) is selected with respect to that characteristic. Selective dropout may lead to bias and thus to overestimation or underestimation of the true effect. Validity  Absence of systematic errors. The degree to which the data measure what they are intended to measure, and thus, the degree to which the results of a measurement correspond to the true status of the phenomenon being measured [1]. Variable  Something that can vary in value. Colour can vary in value (blue, green, red), length (1.59 m, 1.82 m, 1.90 m) and gender (female, male).

References

145

References 1. Fletcher RH, Fletcher SW, Fletcher GS.  Clinical epidemiology. The essentials. 5th ed. Philadelphia, PA: Lippincott Williams & Wilkins; 2014. 2. Li Y, Schoufour J, Wang DD, Dhana K, Pan A, Liu X, et al. Healthy lifestyle and life expectancy free of cancer, cardiovascular disease, and type 2 diabetes: prospective cohort study. BMJ. 2020;368:16669. https://doi.org/10.1136/bmj.l6669. 3. Van Puijenbroek EP, Offringa M, Scholten RJPM.  Side effects and etiology. In: Scholten RJPM, Offringa M, Assendelft WJ, editors. Introduction to evidence-based medicine. Clinical practice based on evidence. 4th ed. Houten: Bohn Stafleu van Loghum; 2014. 4. Guyatt G, Rennie D, Meade MO, Cook DJ. Users’ guides to the medical literature. A manual for evidence-based clinical practice. 3rd ed. New York, NY: McGraw Hill Education; 2015.

9

What Can We Achieve: How to Critically Review Articles on Prognostic Research?

If patients have had a myocardial infarction, the cardiologist always provides medical treatment. This significantly reduces the risk of patients dying from cardiovascular causes. In addition, many patients adjust their lifestyle and further improve their prognosis through a healthier lifestyle. As a nurse, I can play an important role in this. For example, I can advise and support patients to eat more healthily. But I wonder which food ingredients contribute to a better prognosis. If I have that knowledge, I can advise patients better about their diet. In a systematic search via PubMed, I found the cohort study by Li et al. into the value of dietary fibre for patients after a first myocardial infarction, as a prognostic factor for later death from cardiovascular and other causes. This study shows that increasing the intake of dietary fibre from cereal products (after experiencing the first myocardial infarction) reduces the risk of death from any cause by about 31%. That sounds a bit strange, because everyone dies of course. But what the researchers mean is that cereal dietary fibre influences the course of life in such a way that patients who eat more cereal fibre are more likely to live longer. Unfortunately, the researchers don’t write in their article how much longer these patients live, because I would love to tell the patient that too. The full article by Li et  al. [1] can be found online (https://tinyurl.com/ dietary-­fiber-­intake).

9.1 What Are Prognostic Factors? Why Are They Important? A prognostic factor is a factor which, once the health problem is present, influences the course or outcome of the health problem in a positive or negative way. This distinguishes it from a protective factor and a risk factor, which influence the

With contribution by W. J. Pim Assendelft. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 J. Dobber et al., Clinical Reasoning and Evidence-Based Practice, https://doi.org/10.1007/978-3-031-27069-7_9

147

148

9  What Can We Achieve: How to Critically Review Articles on Prognostic Research?

occurrence of a health problem if it is not yet present. There are various types of prognostic factors: physical factors (age, nutritional status), psychological factors (stress, intelligence, coping), lifestyle and behaviour (eating habits, smoking, exercise, sleep patterns) and environmental factors (environmental pollution, availability of healthcare facilities). Some of the prognostic factors can be influenced (nutritional status, stress management), while others cannot (age, availability of health care facilities). Knowledge of important prognostic factors influences the choice of meaningful interventions and promotes the selection of realistic goals.

9.2 In What Sources Can I Find the Best Evidence for Prognostic Factors? The research form for prognostic research is the cohort study. However, we can also view treatment as a prognostic factor, and in that case, the effect of treatment as a prognostic factor is studied via a randomised controlled trial (RCT). The latter is regarded as intervention research (see Chap. 10). Two forms of cohort study can be carried out in prognostic research: a prospective cohort study or a historical cohort study. Prospective cohort studies are described in Chap. 8. In a historical cohort study, the researcher starts with the health outcome and then examines the extent to which the prognostic factor under study was present in the past. This is a weaker form of research because it has a number of disadvantages. The first disadvantage is that some of the data the researcher wants to have may not have been recorded in the past or may have been incompletely recorded. The exact complaints and symptoms, and the presence of the phenomena in which the researcher is interested, can no longer be reliably ascertained in retrospect. The second disadvantage is that this form of research excludes part of the patients. Patients who have been cured in the meantime, or who have died, have no further contact with the healthcare system and are not included in the historical cohort. The fact that these patients cannot be included in the sample influences the results of the study.

9.3 How Can I Determine the Quality of a Cohort Study of Prognostic Factors? In order to determine the quality, it is important to check whether the research is valid. For this purpose, a checklist of six questions was drawn up. This checklist is based on the checklist of Assendelft et al. [2]. 1. Is the group of patients clearly described? Is the group composed of patients at a similar point in the course of the disease? 2. Is the measurement at follow-up sufficiently complete?

9.4  What Is Important in the Description of the Patient Group?

149

3. Does the period to follow-up last long enough to measure the effect of the studied factor on the course and outcome of the studied health condition? 4. Is there a clear description of how the results were measured? Was the measurement carried out in a valid, reliable and similar way for all participants? 5. Is there a clear description of how the possible prognostic factors were measured? Was this measurement carried out in a valid, reliable and similar manner for all participants? 6. Does the study report on a first test of a possible prognostic factor or have previous studies preceded it? (Sect. 12.65)

9.4 What Is Important in the Description of the Patient Group? First of all, it is important to check whether the measurements of the prognostic factor are taken at a similar time in the entire patient group. The starting point of the study is usually the moment at which the patient is diagnosed with a health problem. This will often be at the start of the disease. The study by Li et al. [1] examined two groups of individuals. The first group came from 121,700 American female nurses; this group was followed by researchers for 32 years. The second group came from 51,529 male health professionals, who were followed by researchers for 22 years. During that time, the participants filled in a comprehensive questionnaire about their current dietary habits every 4 years. In addition, they were asked to notify the researchers if they were diagnosed with a myocardial infarction. The participants who were diagnosed with a myocardial infarction allowed the researchers to access their medical records to obtain the exact data about the diagnosis. The study participants were those who suffered a myocardial infarction during the study period. These were 2639 women and 2081 men, all of whom were healthy at the start of the cohort study. In the course of the observation period (32 years for the women and 22 years for the men), they suffered a myocardial infarction and survived. Of these individuals, dietary questions from both before and after the myocardial infarction were analysed. From the study group, 381 women and 281 men were excluded because they died before receiving the first dietary questionnaire after their myocardial infarction or because they did not complete the questionnaires [1]. Thus, this study of the prognostic value of dietary fibre used the dietary questionnaires completed by the study participants every 4 years. Subtracting the weight of dietary fibres consumed on the last questionnaire before the myocardial infarction from the weight of dietary fibres consumed on the first questionnaire after the myocardial infarction provided insight into the difference in intake before and after the myocardial infarction [1].

150

9  What Can We Achieve: How to Critically Review Articles on Prognostic Research?

This example shows the importance of clear reporting of the starting point of the study. On the face of it, in the dietary fibre study, this seems the same for every patient. However, suppose patient A has a myocardial infarction 2  weeks before completing his 4-year questionnaire, and patient B has a myocardial infarction 2 weeks after completing his 4-year questionnaire. Then the measurement of the change in dietary habits in patient A already takes place after 2 weeks, but in patient B only after 4 years. That is a big difference! In addition, it should be clear what the source of the patient group is. Is it GP practices, via hospital emergency departments or elsewhere? Information about patient characteristics (demographics) and clinically relevant characteristics is also important. This type of data can usually be found in the first table of a research article. Finally, the inclusion criteria and exclusion criteria should also be clear. This information will allow the nurse to determine if her patient is sufficiently similar to the study participants.

9.5 How Can I Verify That the Follow-Up Measurement Was Conducted with a Sufficient Number of Study Participants? Preferably, follow-up measurements are carried out on all patients who have started the study. This is almost always impossible, for example, because patients move house, decide not to participate, or die. The authors should specify the reasons for dropout (‘attrition’ or ‘lost to follow-up’) in their article. In addition, they should provide sufficient information to allow the reader to assess whether attrition affected the study results. The authors can do this by comparing the patient characteristics of the study participants who have dropped out with those of the patients who completed the complete study pathway. If there is no difference between the two groups, the influence of the dropout is probably limited. Another way is to calculate the results in different ways and to see if this leads to different results. Such a sensitivity analysis can be used to determine whether and how much the observed results change if the missing data from the droppedout patients are filled in with different values [2, 3]. For example, for all patients who dropped out, the researchers fill in the worst possible value (worst-case scenario) or the most favourable value (best-­ case scenario) or a value exactly in between. It is then investigated (and reported in the article) how each of these scenarios affects the study results. In a number of cases, it is also possible to use results of earlier measurements on the patient (last observation carried forward) or to calculate and enter the most probable value on the basis of other patient data (to impute). A rule of thumb is that no more than 20% of patients should drop out. In the dietary fibre study, Li et al. [1] reported that of the 4720 individuals who reported a myocardial infarction, 622 were excluded because they had died before they could complete the first dietary questionnaire after their myocardial infarction or because their dietary questionnaire was incomplete. This amounts to approximately 13% dropout.

9.8 How Can I Assess the Quality of the Measurements of the Potential Prognostic…

151

9.6 How Do I Know If the Period to Follow-Up Is Long Enough to Measure the Effect of the Studied Factor on the Course and Outcome of the Studied Health Condition? The operating time required by mechanism of action of the prognostic factors to influence the course and outcome of the health condition and the chosen health outcome varies from factor to factor and from condition to condition [2–4]. The period to follow-up should include at least that duration of action. Preferably even longer, as the effect time may also vary from person to person. In addition, it often happens that the exposure of the study participants to these factors changes in the interim and becomes more or less. This may also influence the required duration of the follow-up. The authors must make a plausible case that the follow-up period they have chosen is indeed long enough to measure the influence of the prognostic factor on the course of the health problem and the outcome of the health outcome being studied.

9.7 How Can I Assess the Quality of the Measurements of the Outcomes? The authors should clearly describe which outcomes they are measuring. In the methods section of the article, the exact procedure used to measure the outcomes should be clearly presented. If the outcome is survival or death, the criteria for the outcome are clear. But if the outcome is defined as return to work, the criteria must be further specified. Is it only about paid work or does voluntary work also count? Does the research participant have to return to the job he had before his MI or can it be other work? Another question is how the researcher obtains the information about the outcome. From the research participant himself, from the GP or from another source? How certain is it that this source correctly informs the researcher? If the outcome is to be determined or assessed by measurement, this should preferably be done by an effect assessor who is blinded to the extent to which the research participant is exposed to the possible prognostic factor. The information in the methods section should clearly show that the outcomes have been measured in a valid and reliable manner.

9.8 How Can I Assess the Quality of the Measurements of the Potential Prognostic Factor? The same applies to the measurement of prognostic factors as to the measurement of outcomes. There are prognostic factors that are easy to measure (such as age), but they often involve more complicated measurements. For example, is the question of whether a study participant smokes measured by asking him to report the amount of

152

9  What Can We Achieve: How to Critically Review Articles on Prognostic Research?

daily smoked cigarettes himself (self-reporting)? Or is a cotinine test performed to detect the possible presence of nicotine in the blood, urine or saliva? The authors should clearly describe the criteria and the measurement procedure they used to determine the presence, magnitude or intensity of the possible prognostic factor. This should clearly demonstrate that the measurements were valid and reliable. In addition, they must clearly describe the timing of the measurements.

9.9 What Is the Value of Initial Research on a Prognostic Factor? How Does It Differ from Later Research? The better and more frequently a prognostic factor has been studied, the more the nurse can rely on the effect of the prognostic factor, provided that the results of the various studies point in the same direction. If several studies have been carried out into a prognostic factor (or a combination of prognostic factors), a systematic review with a meta-analysis may provide stronger evidence of the effect.

9.10 How Can I Determine the Importance of the Study Results in Relation to the Prognostic Factor for the Patients? To determine the importance of the prognostic factor, the nurse must check the magnitude and precision of the effect of the prognostic factor. The way in which these data should be interpreted is discussed in the chapters on etiology (Chap. 8) and on therapy (Chap. 10). In a prognostic study into the course of a particular disorder, a graphical representation of the prognosis van be used to clarify the prognosis (see Fig. 9.1).

9.11  How Can I Find Out If the Study Results Can be Used for My Own Patients?

153

100 A

percentage of survivors

80 B 60

40

C

20

D

0

0

1

2

3

4

5 6 7 8 time (in years)

9

10 11 12

Fig. 9.1  Examples of four survival patterns, shown in a Kaplan-Meier curve. Horizontal line at 50% indicates median survival time limit. Pattern A: prognosis favourable. Percentage of survivors decreases only slowly. The median survival cannot be given; at the end of the follow-up period (12 years) more than 50% of the patients are still alive. Pattern B: short-term prognosis favourable, long-term unfavourable. The percentage of deceased patients increases rapidly from 6  years of follow-up. The median survival is approximately 11.5 years. Pattern C: prognosis unfavourable throughout the follow-up period. Median survival is approximately 8.5 years. Pattern D: prognosis very unfavourable in the short term. However, the prognosis for patients who survive the first 5 years is much more favourable, mortality then decreases significantly. The median survival is approximately 3 years. (Source: Assendelft et al. [2])

9.11 How Can I Find Out If the Study Results Can be Used for My Own Patients? If the patients from the study are sufficiently similar to the nurse’s patients, the study results can be used. First, the nurse can check the demographics and clinical characteristics of her patient and the study participants for consistency. In addition, the nurse can check if her patient falls within the inclusion criteria and outside the exclusion criteria of the study. The next question is whether the nurse’s patient indeed has the same health problem, with the same severity and at the same stage, or whether the patient is at a more advanced or earlier stage. The patient’s setting (home, hospital, nursing home or psychiatric centre) can also have an influence. Finally, the nurse (or nursing team) must have all the necessary resources (equipment, time, knowledge and skills) to adequately perform the measurements of and/or influence the prognostic factor.

154

9  What Can We Achieve: How to Critically Review Articles on Prognostic Research?

Glossary Cohort study  Research in which the groups (cohorts) to be followed are classified according to the presence of, or exposure to, a specific determinant. The cohorts are followed during the observation period and compared in order to determine whether differences in determinant status lead to differences in specific health outcomes. Exclusion criteria  Criteria that define the characteristics of study participants that must not be present in order to participate in the study. If a potential participant has one or more of these characteristics, he is ‘excluded’, that is, removed from the group of potential participants. Exposure  Means ‘coming into contact with’. Exposure to asbestos therefore means: coming into contact with asbestos. Inclusion criteria  Criteria that define the characteristics of the study participants that must be present in order for them to participate in the study. Mechanism of action  The way in which the etiological or prognostic factor brings about the health change in the patient. The mechanism of action explains which and how physiological or psycho-social processes deviate from their normal course and how they affect the health status. Meta-analysis  Analysis that can form part of a systematic review. In a meta-­ analysis, the results of the individual studies included in the systematic review are combined (pooling). By pooling the results, a new effect estimate can be calculated which is more reliable than the effect estimates of the individual studies. Prognostic factor  Factor that, if the health problem is already present, positively or negatively influences its course and outcome and thus also influences the health outcome(s). Protective factor Factor that, if a particular health problem is not yet present, reduces the risk of that health problem developing. Randomised controlled trial (RCT)  Experimental research into the effectiveness of an intervention. Participants are randomly assigned to an experimental group or a control group. Depending on the group in which the participant is placed, he will receive an experimental intervention or a control intervention. After the intervention, the effects of both interventions are compared to determine the effect of the experimental intervention on predefined outcomes. Reliability  The degree to which repeated measurements of a stable phenomenon by different people and instruments at different times and places get similar results [3]. Risk factor  Factor that, if the health problem is not already present, increases the risk of developing that health problem. Sensitivity analysis  In sensitivity analysis, the researchers calculate the way in which the effect changes if they do assign a result to the dropout participants. It can be determined if and how much the results change if the missing data of the droppedout patients are filled in with different values.

References

155

Validity  Absence of systematic errors. The degree to which the data measure what they are intended to measure, and thus, the degree to which the results of a measurement correspond to the true status of the phenomenon being measured [3].

References 1. Li S, Flint A, Pai JK, Forman JP, Hu FB, Willett WC, et al. Dietary fiber intake and mortality among survivors of myocardial infarction: prospective cohort study. BMJ. 2014;348:g2659. https://doi.org/10.1136/bmj.g2659. 2. Assendelft WJJ, Korevaar JC, Reitsma JB.  Forecasting. In: Scholten RJPM, Offringa M, Assendelft WJ, editors. Introduction to evidence-based medicine. Clinical practice based on evidence. 4th ed. Houten: Bohn Stafleu van Loghum; 2014. 3. Fletcher RH, Fletcher SW, Fletcher GS.  Clinical epidemiology. The essentials. 5th ed. Philadelphia: Lippincott Williams & Wilkins; 2014. 4. Randolph AG, Cook DJ, Guyatt G. Prognosis. In: Guyatt G, Rennie D, Meade MO, Cook DJ, editors. Users’ guides to the medical literature. A manual for evidence-based clinical practice. 3rd ed. New York: McGraw Hill Education; 2015.

What Can We Do About the Problem: How to Critically Appraise Articles on Interventions?

10

In the evidence pyramid (Sect. 6.4), the evidence-based guideline is at the top. This offers the most complete evidence about interventions. If no recent guideline is available, or if the guideline does not answer the search question, a synthesis is the best evidence. In the hierarchy of evidence on interventions, the original studies are next in line: first, the randomised controlled trial (RCT) and then, the controlled clinical trial (CCT). This chapter will discuss these types of evidence-based sources in more detail. First, the original studies will be discussed, and then, it moves up the evidence hierarchy step by step. For other types of research than those mentioned above, the scientific value of research into the effectiveness of interventions is lower. It requires considerable knowledge of research methodology to determine its exact value. For this reason, these research types are not covered in this book.

10.1 Randomised Controlled Trial and Controlled Clinical Trial During my cardiology internship, I saw many patients who maintained an unhealthy lifestyle despite a previous heart attack. They continued to smoke or exercise too little, and there were also quite a few patients who were overweight or had very unhealthy eating habits. Several patients were admitted after a second, or sometimes third or fourth heart attack. Their lifestyle plays an important role in this. How can nurses help these patients who find it difficult to change their lifestyle?

With contribution by Pim Assendelft and Rob Scholten. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 J. Dobber et al., Clinical Reasoning and Evidence-Based Practice, https://doi.org/10.1007/978-3-031-27069-7_10

157

158

10  What Can We Do About the Problem: How to Critically Appraise Articles…

Most people know that smoking, lack of exercise, obesity and unhealthy eating are bad for their heart and blood vessels. But even after a heart attack, 25%–43% of smokers do not even try to quit [1, 2]. Because lifestyle change is difficult for many people, Minneboo et  al. [3] investigated the effectiveness of an intervention in which nurses could refer patients after a heart attack to community-based lifestyle interventions. The full article by Minneboo et al. can be found online (https://doi. org/10.1016/j.jacc.2017.05.041).

10.1.1 What Is a Randomised Controlled Trial (RCT)? A randomised controlled trial (RCT) investigates whether an intervention is effective. This is done by comparing the research or experimental intervention with another intervention (the control intervention), of which the effect is already known. In the RCT by Minneboo et al. [3], the control intervention was the usual care that patients receive according to current national and international guidelines after a myocardial infarction. This usual care consisted of the usual check-up visits to the cardiologist, and up to four outpatient visits to nurses who carried out a secondary prevention programme. This programme focused on lifestyle and medication use. In the experimental intervention, patients received the same usual care, but were also referred to one, two or three lifestyle programmes outside the hospital (Weight Watchers, focused on healthy diet and weight loss; Philips Direct Life, an internet programme focused on exercise; Luchtsignaal, a telephone coaching programme focused on smoking cessation). At the beginning, all participants’ demographic characteristics (such as age and gender) were recorded and their lifestyle profile measured (including BMI, physical activity and fitness, smoking status). This is called the baseline measurement. After 12 months, this lifestyle profile was measured again. An RCT has a set way in which the study should proceed. First, the subjects who can participate in the study are determined: the base population. In the study by Minneboo et al., the base population consisted of adult patients admitted to hospital because of acute coronary syndrome (ACS, this syndrome includes both patients with myocardial infarction and patients with unstable angina pectoris) and/or dotter treatment. This group is then further delineated in a study population. This is done by means of two types of criteria: inclusion criteria and exclusion criteria. The inclusion criteria define the characteristics of the research participants that they should at least have in order to participate in the study. One fixed inclusion criterion is that participants must have the condition or disorder for which the intervention is being studied. In the example of Minneboo et al., this inclusion criterion is: hospitalisation because of ACS and/or dotter treatment, and one or more of the lifestyle-­ related risk factors of weight, exercise and smoking. Furthermore, patients had to be willing to participate in at least one lifestyle programme. Exclusion criteria

159

10.1  Randomised Controlled Trial and Controlled Clinical Trial

determined which characteristics had to be absent from the study participants. If a person has one of the exclusion characteristics, he cannot participate in the study. A common exclusion criterion is comorbidity: because the person has another condition that is not being studied in the study, they cannot participate in the study. For example, one of the exclusion criteria in Minneboo et al. is heart failure class 3 or 4 [3] (Sect. 12.48). The persons remaining after application of the inclusion and exclusion criteria receive information about the study and are asked to participate. If a person agrees to participate (informed consent), they will be part of the sample. Next, all participants are randomised. Randomisation means that participants are randomly assigned to either the experimental group (also called the intervention group) or the control group. For each research participant, chance determines which of the two groups he ends up in. Hereafter, the experimental and the control interventions are carried out. Afterwards, the researchers measure the effect of the intervention in all participants and compare the effects of the control intervention with the experimental intervention. Through this comparison, it is determined whether the experimental intervention has an effect and if so, the size of this effect. The basic scheme of an RCT (using the study by Minneboo et al. as an example) looks like Fig. 10.1.

Populaon 2031 paents admied to hospital, because of ACS and/or doer treatment

Experimental intervenon 411 paents, usual care and lifestyle programme

Outcome

Control intervenon 413 paents, usual care only

Outcome

133 paents met the primary outcome*

Sample 824 paents who fulfil the inclusion and exclusion criteria and who gave informed consent

91 paents met the primary outcome*

*Primary outcome: paents progress on at least 1 lifestyle risk factor, without deterioraon on the other lifestyle factor(s).

Fig. 10.1  Basic outline of an RCT. (Using the study by Minneboo et al. [3] as an example)

160

10  What Can We Do About the Problem: How to Critically Appraise Articles…

10.1.2 How Can I Critically Evaluate the Quality of a Randomised Controlled Trial (RCT)? How Can I Determine the Applicability of the Intervention to Patients in My Practice? To be able to understand the content of an RCT sufficiently and assess it critically, there is a fixed checklist of questions about the article in which the RCT concerned is reported. By reading the article and answering the questions, the quality of the study and the usefulness of the study results for one’s patients are mapped out step by step. The checklist in this book is largely based on the checklist of the Dutch Cochrane Centre [4]. Below is the background information needed to understand the questions of the checklist correctly and to interpret the text of the RCT correctly (Sect. 12.66). Based on an understanding of the RCT, nurses can decide on the application of an intervention within their patient care. This step may seem simple, but it requires background knowledge. Nurses can only apply the evidence from an RCT to their practice if they are able to critically appraise the article on that RCT. For example, a woman with a headache may take a painkiller and the headache will be over in 15 min. But was it really the case that the headache went away because she took the painkiller? Next, consider a patient who is participating in an RCT.  The patient knows he is receiving therapy, but not what therapy. The therapy is called ‘intervention’, but it may not have any effect at all. What effect are the researchers actually looking for? And why do they go to so much trouble to carry out measurements, if the intervention has already ended 6 months ago? This may sound strange, but in an RCT, procedures such as randomisation and blinding are important to measure the true effect of the intervention as accurately as possible and thus minimise bias in the study results. The first six questions of the checklist concern the extent to which bias (systematic errors) occurs in the study. Bias influences the results of the research, and as a consequence, the research results are distorted. Because this reduces the validity of the study, bias should be avoided as much as possible. Questions 7–9 are about the effects of the intervention. If there is no effect or only a small or uncertain effect, it is recommended not to use the intervention for patients in your practice. Finally, the last five questions deal with the applicability of the intervention in the nurse’s practice.

10.1.3 What Is Meant by Randomisation? Randomisation means that the researchers allow chance to determine which participants are in the experimental group and which are in the control group (random sequence generation). This randomisation ensures that the two groups are mutually comparable at the start of the study (the baseline). With good randomisation, all participants have the same chance of being assigned to one of the two study groups.

10.1  Randomised Controlled Trial and Controlled Clinical Trial

161

10.1.4 Why Is Randomisation Important? By allowing chance to determine who ends up in which research group, the characteristics of the participants in the experimental group are easily comparable to those of the participants in the control group, provided the sample is large enough. Randomisation attempts to ensure that patient characteristics that (may) influence the course and thus the outcome are present in both groups in roughly the same way and to the same degree. For example, in the RCT by Minneboo et al. [3], the mean age of the patients in the control group and in the experimental group is very similar. In the control group it is 59.2 years and in the experimental group 58.2 years. There are almost always patient characteristics that influence the course of the disease. Some of these characteristics are known, such as age. However, there are also patient characteristics that influence the course without their influence being known. These are called the unknown factors. The value of a well-executed randomisation is that it strongly reduces the chance of a serious imbalance between both groups in the distribution of known and unknown factors. To find out whether the allocation to the experimental group and to the control group is indeed determined by chance, the first questions of the checklist concern the method of randomisation. There are several ways to perform a proper randomisation. Often, a computer program determines which person goes into which group. In addition, it is very important that the person performing the randomisation cannot influence it at all (allocation concealment). For this reason, randomisation is often carried out by an independent person who is not involved in the study and has no interest in the grouping. However, even in this case, it is necessary that the person randomising cannot know prior to the randomisation what group a participant will be assigned to (Sect. 12.49).

10.1.5 What About a Controlled Clinical Trial? The word controlled refers to the fact that the results of the experimental group are compared with the results of the control group. Sometimes there are reasons not to compose the experimental group and the control group (entirely) by chance. If some degree of coincidence plays a role in the composition of the groups, this is called quasi-random or pseudo-random. There is no real randomisation and therefore no real randomised controlled trial. In fact, it is a non-randomised clinical trial. However, it is customary to call this type of trial a controlled clinical trial (CCT). So, a CCT is a kind of RCT without randomisation. It will be clear that due to the lack of proper randomisation, the quality of these trials is significantly lower than that of RCTs.

162

10  What Can We Do About the Problem: How to Critically Appraise Articles…

10.1.6 Why Is the Comparability of the Participant Groups Important? In an RCT, the experimental and control interventions are compared. The experimental intervention is considered effective if the results of the experimental intervention are clearly better than the results of the control intervention. However, if the experimental intervention is tested on people between 18 and 25 years old, and the control intervention on people between 58 and 65 years old, it is usually easy to demonstrate the superiority of the experimental intervention. After all, usually younger people recover better and faster than older people. It is therefore necessary that the groups are well comparable. This is measured at the start of the RCT. In the article, the author reports the comparability ‘at baseline’ in a table. In that table, three types of characteristics are reported: the demographic characteristics, the prognostic variables and the measurement of the most important outcomes at the start of the study. Demographic characteristics provide information about age, gender, level of education and so on. Prognostic variables are characteristics that influence the course of the disease. Examples include the stage of the disease, medication use and comorbidity. Furthermore, at the start of the study, the values of the most important outcomes should be approximately equal. Table 1 of the article by Minneboo et al. [3] shows that the demographic factor ‘gender’ is very similar in both groups (21% of the patients in both the control group and the experimental group were women). The prognostic variable ‘treatment with beta-blockers’ is more common in the control group (87%) than in the experimental group (83%). The outcome ‘BMI’ was about the same in both groups: the average BMI in the intervention group was 29.8 and in the control group 29.3. If the groups differ on one or more characteristics, it is important to consider how these differences could influence the outcome. In many cases, the authors will pay attention to this themselves in the discussion section of the article. It is also possible that the researchers correct some of the differences by using adjusted analysis techniques. If they have applied this correction, the authors report this in their article.

10.1.7 What Is Meant By Blinding? Blinding, or masking, means that the people involved in the study do not know whether the participants are in the experimental group or the control group. In RCTs testing drugs, the participants themselves are blinded: they do not know whether they are receiving the experimental drug or (for example) a placebo. Although it may seem unethical for a participant not to know which drug they are receiving, blinding is important because of the psychological effect of the intervention. If a person receives an intervention that they expect to work, the likelihood of effect is greater. In contrast, if the person thinks they are getting a placebo and expects it to not work anyway, the chance of effect is actually lower. Also, a participant who thinks they are getting a placebo may not be as committed to carrying out the

10.1  Randomised Controlled Trial and Controlled Clinical Trial

163

intervention as prescribed and may not always take the medicine. Because such prior knowledge influences the behaviour of the participant and the effects of the intervention, blinding is important. Similar mechanisms are at play in other types of interventions than medication, but it is more difficult to blind participants. Not only study participants can be blinded. Another group that can be blinded are the healthcare professionals involved in the care of the participants. For example, this type of blinding prevents healthcare professionals from going the extra mile for participants who are in the control group, because they want to compensate them for not receiving the experimental intervention. When healthcare professionals are blinded, they cannot (consciously or unconsciously) treat the study participants from both groups differently. It depends on the experimental intervention whether it is practically possible to blind the healthcare professionals. A group which is relatively easy to blind is the outcome assessors. These are the people who carry out the measurements with the participants, from which ultimately the effect of the experimental intervention is determined. If the outcome assessor is not blinded, he could unknowingly use a different (e.g. more extensive) measurement method with participants from the experimental group. And such different measurement methods influence the measurement results, making the results of both interventions no longer comparable. The article should clearly report whether blinding was used and which subjects were blinded. Sometimes the term ‘double blind’ is used. This means that two groups of subjects are blinded, in which case it should be described which two groups they are.

10.1.8 Why Is Treatment Comparability Important? In an RCT, differences in the effect of the outcomes in the experimental and control group should be solely attributable to the difference in intervention. Therefore, except for the difference in intervention, both groups should be treated the same.

10.1.9 How Is the Quality of the RCT Affected by Sample Size and Completeness of Follow-Up? The sample size is the number of persons that were randomly assigned to one of the two groups. In the RCT by Minneboo et al. [3], the sample size was 824 persons. A number of general principles apply to sample size. –– In large samples, randomisation leads to more comparable groups than in small samples. In very small samples, the groups are often not very comparable on various factors. –– In large samples (smaller), differences in effect between the interventions can be demonstrated sooner than in small samples, while (very) large differences in effect can also be demonstrated in small samples.

164

10  What Can We Do About the Problem: How to Critically Appraise Articles…

–– The effects can be determined more accurately in larger samples, allowing greater confidence in the effects found. The required sample size depends on several factors. When assessing the sample size in a study, it is best to check what the article says about the number of study participants needed to demonstrate significant effects. This number is calculated by the researchers using a power analysis. In the article paragraph on statistical analysis, the researchers give the results of this power analysis. The sample size should include at least the same number of participants as the researchers report as a result of their power analysis. In the RCT of Minneboo et al. [3], the authors report that at least 790 participants were needed. Since during a research period, almost always a number of participants decide to stop their participation in the research, move, die or drop out for other reasons (attrition or loss to follow-up), it is usually not possible to carry out all measurements for all individuals in the sample. It is not easy to indicate how much dropout is acceptable. It depends, for example, on the power analysis: how much expected dropout is it based on? There can also be selective dropout, for example, when especially people dropout who do not perceive any benefit from the intervention or who even think the intervention harms them. This type of dropout can strongly influence the results. A lot of information on dropout can be found in the flow diagram published with the article. Furthermore, it is important to examine how the researchers have incorporated the dropout in their results. This should be done according to the intention-to-treat principle. This principle means that all participants assigned to the intervention or control group during randomisation are also analysed in the same group (even if they terminated their participation in the study in the interim or if they withdrew immediately after randomisation). There are various ways in which this principle can be implemented, for example, via the method of last observation carried forward. In this method, the value entered at the last measurement before the participant dropped out is also entered as the value for that participant at the measurements after his or her dropout. Sometimes a probable value can be calculated (‘imputation’) based on data about the patient that has already been collected in combination with data from the other patients [5] (Sect. 12.50). Since it is unknown what the actual results of dropouts would have been if they had not dropped out, there is uncertainty about the influence of dropout on measured results [6]. It may be that without the dropout, the effect of the intervention would be smaller and that the results overestimate the effect. But it could also be that the effect would be larger (underestimation of the effect) or that there would be hardly any difference in effect with the results found now. An answer to this uncertainty is to carry out a sensitivity analysis. With the sensitivity analysis, the researchers calculate the way in which the effect changes if they assign a result to the research participants who dropped out. For example, the dropped-out patients in the group with the best results can be assigned the worst possible result, and the

10.1  Randomised Controlled Trial and Controlled Clinical Trial

165

dropped-out patients in the group with the worst results can be assigned the best possible result [7]. In this way, the effectiveness of the best performing intervention cannot be overestimated. The results of the sensitivity analysis are compared with the actually measured results, and then the way the results change and what consequences this would have for the effect of the intervention are investigated. If a sensitivity analysis has been carried out, the researchers report what the difference is with the actual measured results. Sometimes one of these approaches is combined with a ‘per protocol analysis’. In this case, only those patients are analysed who have completed the entire protocol. The dropouts are then left out. This usually overestimates the effect of the intervention. The results of a per protocol analysis cannot be used in clinical practice because the intervention is not carried out exactly according to the protocol for everyone, and in clinical practice, there are also patients whose intervention is terminated prematurely. In the RCT by Minneboo et al. [3], a sensitivity analysis was performed in which the most likely results for all dropouts were calculated and entered via imputation. The results of this sensitivity analysis hardly differed from the analysis without the dropouts. Therefore, in this RCT, there was hardly any influence of dropouts on the measured effect of the intervention.

10.1.10 What Is Meant By Validity? How Can the Validity of an RCT Be Assessed? A study is valid if it contains no systematic errors. Systematic errors (bias) in the study influence the results, causing them not to reflect the actual values. If errors have been made that could have significantly distorted the results, the data from the study cannot be relied on sufficiently. The study cannot then be used to make decisions about patient care. There is then a high risk of bias, but on the other side, hardly any study can be guaranteed to be free of bias. When assessing validity is to determine the magnitude and severity of possible bias. This involves estimating the likely impact of the bias on the study results [8]. Is it likely that the bias will lead to an overestimation (a more favourable result) of the experimental intervention or to an underestimation of the effectiveness of the intervention? Overestimation is more serious than underestimation. If the intervention turns out to work better than expected, that is not so bad. The most important research elements that can be affected by bias are described above: 1. randomisation, 2. blinding, 3. comparability of participants, 4. completeness of follow-up measurements/intention to treat, 5. comparability of treatment.

166

10  What Can We Do About the Problem: How to Critically Appraise Articles…

Of the domains in which biases occur, three can be mentioned which usually lead to the greatest distortions. These three aspects are given extra weight when assessing the validity of a study. These are the following aspects. 1. The lack of allocation concealment. This significantly increases the chance of selection bias. 2. The lack of blinding of the effect assessor. Of the groups that can be blinded, that of the effect assessors is the most important. The absence or breach of blinding the effect assessor leads to an increased risk of ascertainment bias (also known as information bias) and detection bias. 3. Failure to apply the intention to treat principle in combination with too large and/ or selective dropout of research participants. If a large number of participants drop out, or participants drop out selectively (attrition bias), and the dropouts are not included in the analyses, this threatens the comparability of the experimental and control groups. Assessing validity is not about finding as many errors in the study as possible. There are almost no studies without bias, because practical circumstances are never entirely ideal. However, it is important to be able to reason out what influence bias has had on the research results and on the conclusions of the study. So, the question is: ‘To what extent can I trust the results of this study, can I base decisions about my patient(s) on them?’ [7]. If the study is not sufficiently valid, it cannot be trusted that the described effects of the intervention correspond to the actual effects. In this case, the study cannot be used in clinical reasoning about the suitability of the intervention for one’s patient care. If the study is sufficiently valid, it is necessary to further assess its usefulness. To do this, it is first necessary to look at the results of the intervention.

10.1.11 How Can the Results of an RCT Be Interpreted? An RCT often involves more than one effect. Consider, for example, the desired effects of the interventions and any undesired side effects. The desired effects are formulated in advance and are divided into primary and secondary outcomes. The study was set up to measure the primary outcomes, the most important outcomes of the study. The authors of the article describe clearly what the primary outcomes and secondary outcomes are. In the RCT by Minneboo et al. [3], the primary outcome was the percentage of successful patients, that is, patients in whom at least one lifestyle-related risk factor (smoking, physical activity, obesity) had improved after 12  months, without any deterioration in the other lifestyle-related risk factors. Secondary outcomes in this RCT included medication compliance and systolic blood pressure. The measure by which the outcomes are measured is called the outcome measure. For example, the score on the ‘six minutes walking distance’ test (6MWD) was the outcome measure for physical activity [3]. This test involves the patient being instructed to walk as many metres as possible in 6 min, in a responsible manner. The number of metres walked after 6 min is then taken as the measure.

10.1  Randomised Controlled Trial and Controlled Clinical Trial

167

Because this test is performed as a baseline, both prior to the intervention and after the intervention, the difference in meters can easily be measured in each patient (For a checklist of frequently used measures of effect: see Sect. 12.70). The effects found are expressed in a measure of effect, for example, a relative risk (RR), and in a confidence interval. Two aspects are therefore important: the size of the effect, expressed in the measure of effect, and the precision of the effect, expressed in a confidence interval. The magnitude of the effect is shown in a point estimate (or: effect estimate). The point estimate reflects the effect found in the study. The RCT by Minneboo et al. [3] found that in 26% of the patients in the control group and in 37% of the patients in the intervention group, at least one lifestyle-related risk factor had improved, without any deterioration in the other two lifestyle-related risk factors. This means that in this study, patients who received the experimental intervention were 1.43 times more likely than patients in the control group to achieve success on the primary outcome. In research language, this is called the relative risk (RR) = 1.43. The 1.43 is the point estimate of the measure of effect relative risk. There are various measures in which the effect can be expressed. In the Checklists in Chap. 12, there is an overview of a number of frequently used effect sizes. The precision of the effect is represented by a confidence interval (CI). The point estimate found in the RCT is the best estimate of the actual effect of the intervention. However, because the intervention was carried out in a population sample, the effect found is influenced by the composition of the sample. If the sample had consisted of other persons, it is likely that a slightly different effect would have been found. This effect would probably be close to the effect found now (an RR of 1.43), but would deviate from it somewhat (for example an RR of 1.49). The phenomenon that the point estimate varies depending on the composition of the sample is called sampling variability. This means that it is unlikely that an RCT will find the actual effect exactly. This is the reason for calculating a confidence interval. Usually, there is a 95% confidence interval. In the RCT by Minneboo et al. [3], the researchers found (at the point estimate of 1.43) a 95% confidence interval of 1.14 to 1.78. In articles, this is usually represented as follows: RR = 1.43 (95% CI 1.14–1.78). It means that if the study is well (valid) conducted, there is a 95% chance that the actual effect of the intervention falls within the confidence interval. Thereby the true effect is likely to be close to the point estimate and probably not close to the extremes of the confidence interval (Sect. 12.51). The confidence interval calculated by Minneboo et al. is not that wide. However, it does make a difference whether the intervention increases the chance of successfully improving your lifestyle after a heart attack by 1.14 times or 1.78 times. The wider the confidence interval, the more difficult it is to identify the precise effect. It is simpler if the confidence interval is narrower, for example, 1.32–1.59 rather than 1.14–1.78. In this case, we can say that the intervention increases the chance of a successful lifestyle improvement after a heart attack by approximately 1.5 times. Whereas a heart patient who only received usual care had about a 26% chance of achieving successful lifestyle improvement, this chance would be increased to about 26 × 1.5 = 39% by applying the lifestyle intervention at an RR of 1.5 (Sect. 12.52).

168

10  What Can We Do About the Problem: How to Critically Appraise Articles…

The narrower a confidence interval is, the more certain we are of the true effect. The confidence interval represents the accuracy (or precision) of the point estimate. Therefore, a narrow confidence interval is called more precise than a wide confidence interval. Thus, a narrow confidence interval increases the certainty of the magnitude of the true effect. The width of the confidence interval is strongly influenced by the size of the sample. In small samples, the confidence interval is wider than in large samples.

10.1.12 What Is Meant by Significance? It is important to distinguish statistical significance and clinical relevance from each other. The former refers to a statement that is made on the basis of statistical testing in the study: how certain is it that the experimental intervention is more effective than the control intervention? In general, a statistically significant effect is considered to mean that the researcher is at least 95% sure that the experimental intervention is more effective than the control intervention. Sometimes researchers use a different boundary when determining statistical significance. This may be a more stringent limit (at least 99% certainty) or a less stringent limit (at least 90% certainty). Remember that statistical significance is always partly dependent on the control intervention. If an experimental intervention is not statistically significant compared to a placebo intervention or compared to doing nothing, it has no effect. If an intervention is not found to be statistically significant compared to an already known, well-functioning intervention, it has no better effect than that already known, well-­ functioning intervention. Clinical relevance refers to whether the effect found in the experimental intervention is meaningful in daily practice and thus important enough to be implemented in practice. For example, an intervention is designed to help people with trouble falling asleep. The intervention causes people to fall asleep between 5 and 10 min earlier. Although this intervention may be statistically significant in an RCT, those few minutes have little added value for people with trouble falling asleep, and therefore, the intervention is not clinically relevant. There are two ways to determine whether the effects of an experimental intervention are statistically significant: via the p-value and via the confidence interval. The method via the confidence interval is better, because it provides more information about the effectiveness of the intervention. Therefore, it is best to use this method. Below, both methods are briefly explained.

10.1.13 What Does the p-Value Say About Statistical Significance? The “p” in “p-value refers to probability. Probability is expressed as a likelihood, and probabilities are represented by a number from 0 to 1. If the probability of it raining tomorrow is 0, it will certainly not rain tomorrow. If the probability of it raining tomorrow is 1, it will certainly rain tomorrow. A probability of rain of 0.5

10.1  Randomised Controlled Trial and Controlled Clinical Trial

169

means that there is a 50% chance of rain. Although probabilities are expressed as a number between 0 and 1, they are easily converted into a percentage by multiplying the number by 100. A p-value of 0.5 is the same as a 50% chance (because 0.5 × 100 = 50). Statistical significance refers to the question how certain it is that the experimental intervention is more effective than the control intervention. As a starting point, the researcher takes the hypothesis: ‘There is no difference in effectiveness between the two interventions.’ This hypothesis is called the null hypothesis. In addition, the researcher sets up an alternative hypothesis: ‘The experimental intervention is more effective than the control intervention.’ The p-values given in the research article refer to the null hypothesis. Minneboo et al. [3] found a p-value of 0.002 for the primary outcome (successful lifestyle improvement). This means that there is a 0.002 × 100 = 0.2%, or a twotenths (2/10) percent chance that there is no difference in effect on successful lifestyle improvement between the interventions. In short, the lower the p value, the better. For the secondary outcome ‘readmission to hospital’, the researchers found a p-value of 0.63 [3]. This means that there is a 63% chance that there is no difference between the two interventions in their effect on readmission to hospital. It is usual to choose a p-value of 0.05 (= 5%) as the significance threshold. This means that if there is less than a 5% chance that the RCT shows no difference in effect between the two interventions, this is considered a statistically significant effect. If the p-value is 0.05 or higher, this is considered statistically non-significant. Please note that if there is less than 5% chance that there is no difference in effect between the two interventions, this also means that there is more than 95% chance that the experimental intervention is more effective than the control intervention. Sometimes a study uses a significance threshold other than p = 0.05, for example, 0.01 (= 1%). In this case, the effectiveness of the intervention is only demonstrated to be statistically significant if there is less than 1% chance that there is no difference in effect between the two interventions. Other names for significance threshold are alpha (α) and significance level (Sect. 12.53). Thus, the p-value of 0.63 found by Minneboo et  al. [3] is not significant. According to this study, there was a 63% chance that there was no difference in effect between the two interventions in terms of readmission to hospital. Because 63% is higher than the significance threshold value of 5%, the conclusion is that this RCT did not demonstrate statistical significance for this outcome for this intervention.

10.1.14 What Does the Confidence Interval Say About Statistical Significance? The confidence interval indicates that, if the study is well (valid) conducted, there is a 95% chance that the actual effect of the intervention falls within the confidence interval. Thereby, it is likely that the actual effect is close to the point estimate and less likely that it is close to the outer limits of the confidence interval. By looking closely at the confidence interval, it can be determined whether a statistically significant effect has been demonstrated in the study. As an example, we

170

10  What Can We Do About the Problem: How to Critically Appraise Articles…

Table 10.1  Outcome data of successful lifestyle improvement Variable Readmission to hospital

Intervention group (number) 133 of the 360

Control group (number) 91 out of 351

relative risk (95% CI) 1.43 (1.14 to 1.78)

p-value 0.002

Minneboo et al. [3]

will once again use the primary outcome from the RCT by Minneboo et  al. (Table 10.1). The effect of the intervention on the outcome is expressed as a relative risk (RR). There are three possibilities. 1. An RR of 1.0 means that there is no difference between the two interventions. The value 1.0 is the neutral value of the Relative Risk. 2. An RR lower than 1 means that there is a lower probability of the outcome with the experimental intervention than with the control intervention. 3. An RR higher than 1 means that there is a higher probability of the outcome with the experimental intervention than with the control intervention. In this case, a point estimate of RR = 1.43 was found. This would mean that the experimental intervention is more effective than the control intervention. However, because the intervention was carried out in a sample of the population, the effect found is influenced by the composition of the sample. If the sample had contained other persons, the effect would probably have been slightly different. This is the reason why researchers calculate a 95% confidence interval. It means that there is a 95% chance that the real effect of the intervention falls within this confidence interval. Minneboo et al. [3] found a 95% confidence interval of 1.14–1.78. The actual effect lies with 95% certainty between these limits. For example, the actual effect could be 1.71 (the intervention would then work a little better than the point estimate of 1.43 indicates), but also 1.28 (in which case the intervention would be a little less effective). But imagine that the confidence interval would not have been 1.14–1.78, but 0.94–1.99. In that case, the actual effect could also be 1.0 (there is no difference between the two interventions) or 0.95 (the control intervention would even be more effective!). Thus, a result RR = 1.43 (0.94–1.99) means that the experimental intervention can be more effective than the control intervention, that both interventions can be equally effective or that the control intervention is more effective than the experimental intervention. Based on this confidence interval, it can also be seen that there is no certainty whatsoever about the effect of the intervention. The intervention can be both effective and non-effective or even disadvantageous with regard to this outcome! Therefore, the confidence interval provides very useful information. Consequently, it is always necessary to include the confidence interval in the assessment of the effectiveness of an intervention.

10.1  Randomised Controlled Trial and Controlled Clinical Trial

171

If the neutral value of the measure of effect (in this case, relative risk) is enclosed in the confidence interval, no statistically significant effect of this intervention or factor has been demonstrated.

In this case, we have a (notional) confidence interval 0.94–1.99. The neutral value of relative risk is 1.0. Since the value 1.0 falls within the confidence interval, no statistically significant effect was demonstrated.

10.1.15 What Role Do Adverse Effects Play in Assessing an Intervention? An intervention often leads to more than one outcome. This can be seen, for example, in the fact that there are primary and secondary outcomes. It is possible that the intervention leads not only to desired effects but also to undesired effects (adverse effects). For example, in many cases, drugs have side effects that can be mild undesirable, unpleasant and sometimes even dangerous. Other interventions can also have undesirable effects. In addition, events may occur that are contrary to the main outcome of the study. If a patient suffers a heart attack during the period that he is participating in the study into the secondary prevention of heart attacks, this is known as an adverse event. The authors of the article should report whether any adverse effects or adverse events have occurred. If this is the case, the authors should indicate the extent to which these effects and events occurred. This can be expressed in the same type of effect sizes as the desired effects. In this way, the advantages and disadvantages of the intervention can be weighed in the clinical reasoning about the usefulness of the intervention.

10.1.16 How Can It Be Assessed Whether the Experimental Intervention Can Be Generalised To, and Applied In, My Patient Care? The purpose of reading and reviewing research articles is to include evidence in clinical reasoning and thereby use it in clinical decision-making. Clinical decision-­ making is always about a specific patient or patient group. An important component of involving evidence in clinical reasoning is the extent to which the study participants correspond to the specific patient or patient group: can the findings from the RCT be generalised to one’s own practice? The term generalisability refers to the extent to which the study results also apply to people who did not participate in the study. These individuals may have the same condition as the study participants, but

172

10  What Can We Do About the Problem: How to Critically Appraise Articles…

may live in a different country, with a different healthcare system and with different cultural habits. A study from Taiwan into the influence of informal care by relatives after discharge from hospital cannot simply be generalised to a European or American situation. The interpretation of family ties and additional duties and the healthcare system of the countries are too different for that. Another question is whether the results from smoking pregnant women in Glasgow can be used without objection for smoking pregnant women in Plymouth. In some articles, the authors themselves address the representativeness of their study sample and its implications for generalisability in the discussion section. However, it is up to the nurse in her clinical reasoning to consider the extent to which the research findings are generalisable to her patients. If the intervention can be generalised to the nurse’s patients, then the question of the applicability of the intervention in the nurse’s practice must be addressed. In order to do this, the nurse should assess whether the intervention can be implemented in her setting: is the team of nurses competent to carry out the intervention? Are all necessary tools and materials available? Does the intervention not take too much time? In addition, the nurse should check the suitability of the intervention for the patient in whom she wants to apply the intervention. Is the intervention acceptable for the patient? Do the advantages of the intervention outweigh the disadvantages? (Sect. 12.67). The first step consists of asking whether the patient’s problem is sufficiently similar to the problem that the experimental intervention was aimed at. –– Which somatic, psychological or psychosocial problem underlies the health problem? To what extent is the pathophysiology or the deviant psychosocial functioning of the patient comparable with that of the patients studied? –– Which outcomes are important for the patient? Have these outcomes also been measured in the study? The second step is to answer the question of whether the patient is sufficiently similar to the study participants and whether the setting in which care is provided is sufficiently similar. –– Does the patient meet the inclusion criteria? Would he not be excluded from the study based on the exclusion criteria? –– Do the characteristics of the patient adequately match the baseline characteristics of the study participants? (These are usually presented in a table in the article). –– Does the setting of the study (e.g. hospital, district nursing or nursing home) match your setting? Is the setting likely to have influenced the outcome? Then, in the third step, it is necessary to consider the extent to which culture-­ specific aspects have played a role in the research.

10.2 Synthesis: What Is the Value of Systematic Reviews?

173

–– What cultural elements played a role? In what ways were these elements likely to influence the results? –– To what extent do these cultural elements correspond to those in the current patient situation? The fourth step is to ask whether the intervention is feasible in the own practice. –– What knowledge, skills and attitudes do care staff need to adequately carry out this intervention? –– Are these knowledge, skills and attitude already present to a sufficient degree? Are they easy to learn or develop? –– How much time does it take to carry out the intervention properly? Is this time available? –– Are specific equipment or other materials needed to carry out the intervention? Is this available or easily acquired? –– What is the approximate cost of the intervention? Is the budget sufficient to pay for the intervention? Does the budget manager want to spend money on it? Finally, in the fifth step, an assessment has to be made about the usefulness and appropriateness of the intervention for this patient. –– Do the benefits (effect size and precision of the relevant outcomes) clearly outweigh the disadvantages of the intervention? –– Is this intervention worth the direct costs and the indirect costs (time investment)? –– What are alternative interventions and their expected effects? –– Is this intervention acceptable to the patient? –– Which intervention does the patient prefer?

10.2 Synthesis: What Is the Value of Systematic Reviews? A patient complained of lower back pain. I gave him a paracetamol, but the patient said that paracetamol never helps him. So, I decided to find out whether paracetamol is a good painkiller for lower back pain. In searching, I found the systematic review by Machado et al. (2015) on this topic. This article is available online with the title Efficacy and safety of paracetamol for spinal pain and osteoarthritis: systematic review and meta-analysis of randomised placebo controlled trials (https://tinyurl.com/efficacy-­safety-­ paracetamol [9]).

10.2.1 What Is a Systematic Review? A systematic review (SR) combines the results of several original studies. There are different types of SRs. In this section, we will discuss an SR about the effectiveness

174

10  What Can We Do About the Problem: How to Critically Appraise Articles…

of interventions. By collecting all available RCTs (and possibly CCTs) on a clinical question, assessing them and relating the results to each other, a qualitative overview of the existing evidence is created (Sect. 12.54). Sometimes it is possible to pool the results of the individual RCTs (pooling). The individual samples are then merged, as it were, into one large sample. Based on the pooled data, a new calculation of the effect size and associated confidence interval is then carried out. Because this calculation is based on a large sample, there is more certainty about the accuracy of this overall estimate of the effect of the intervention studied than for the effects found in the individual RCTs. If such a new calculation is carried out, this is called a ‘meta-analysis’. In the case of an SR, therefore, no research is carried out with new research participants. Because existing RCTs addressing the research question are the research material, an SR is called ‘secondary research’. Because results of earlier RCTs are combined (synthesised) and calculated at a higher aggregation level, the evidence from an SR is also called aggregated evidence. The value of a well-executed SR is high, especially when a meta-analysis has been performed. Therefore, it is wise to first search for a good SR and not to start with searching for and reading original studies.

10.2.2 How Can the Value of a Systematic Review Be Determined? As the name ‘systematic review’ (SR) suggests, the systematic conduct of a review is crucial to its quality. An SR must be conducted using a fixed, strict method. This method involves the following steps: (a) a clear research question, (b) a good search strategy for original research, (c) a clear and adequate selection of the research found, (d) an assessment of the quality of the original research using fixed criteria, and (e) a clear presentation of the research results. For all these steps, rules of procedure apply. The rules are designed to prevent bias as far as possible. To determine the exact value of the SR, three questions are important: 1 . Is the SR of recent date? 2. Was the SR performed properly (was bias avoided)? 3. Are the results of the SR applicable to one’s own practice? For the first question, new original studies may have appeared after the SR was published. These are obviously not involved in the SR, but could have changed its outcome. Therefore, for a slightly older SR, the nurse should conduct an additional search for recent RCTs. As the search strategy should be well described in the SR, the same search strategy can be used to search for additional literature. To answer questions 2 and 3 and to understand the content of an SR sufficiently to be able to critically assess it, there is a fixed checklist of questions about the article in which the SR is reported. By reading the article and answering the questions, the quality of the research and the usefulness of the research results for one’s

10.2  Synthesis: What Is the Value of Systematic Reviews?

175

patients are identified step by step. The checklist in this book is largely based on the checklist of the Dutch Cochrane Centre [4] (Sect. 12.68). The following is the background information needed to properly understand the questions and adequately interpret the text of the SR.

10.2.3 How Can the Research Question of the SR Be Appraised? An SR should have a clear research question. A clear question statement allows readers to immediately assess the suitability of the SR for their own clinical question. Good clinical questions about interventions have at least four fixed components. 1. Patient. The patient population the SR is about: health problem, age, setting. 2. Intervention. The intervention that is the focus of this SR: exact intervention, dosage, by whom performed. 3. Comparison (or Control). The control intervention with which the effect of the intervention is compared. 4. Outcome(s). The outcome(s) with regard to which the effect of the intervention is examined in this SR. These four parts are summarised in the acronym PICO. The ‘C’ does not always have to be filled in, but the other three parts have to be clearly recognisable in the research question of the SR. In most cases, the question can be derived from the title of the abstract. Sometimes the question follows from the title, in combination with the objective of the study, which is described in the abstract at the beginning of the article. From the SR of Machado et  al. [9], the following information can be extracted from the title: –– –– –– ––

P: spinal pain; osteoarthritis. I: paracetamol. C: placebo. O: efficacy (in this case: pain reduction); safety.

The summary also provides the following information: P: spinal pain refers to neck pain or low back pain; osteoarthritis involves the hip or knee; patients need not have both conditions, but one or the other.

10.2.4 What Requirements Must the Systematic Search for Relevant Studies Meet in a Systematic Review? In order to obtain all relevant studies, the researchers should use a search strategy in which it is almost certain that all relevant articles will be found. This strategy should be clearly stated in the article, enabling the reader to replicate this search. A systematic search includes requirements about the search terms, databases and other

176

10  What Can We Do About the Problem: How to Critically Appraise Articles…

sources and about the limitations (filters or limits) that have been applied in the search. The first step is to check whether the search terms used by the researchers are appropriate to their question. Furthermore, researchers may have omitted important search terms. Therefore, this should also be checked. There are several databases containing journals of importance to healthcare. A systematic search must be conducted in at least the most important of these. The most important databases that should be searched are MedLine (or PubMed, a search engine that searches the MedLine database), EMBASE and the Cochrane Central Register of Controlled Trials (CENTRAL). Other databases that are more specialised in the field of the question can also be searched. If the question is about nursing, CINAHL should also be searched. If the query is related to mental health, a search in PsycINFO is also necessary. In a good search, the researchers also check the reference lists of the articles included in their study. After all, there may be relevant sources among them that the researchers did not find during their search in the databases. Finally, the researchers can also undertake additional searches. For example, they can contact experts in the field of the question, to find out whether the expert knows of another study that was not found in the search. When searching in databases, it is possible to set some limitations in advance, as a result of which articles that are not usable at all will be filtered out of the search yield. These limitations are called limits or filters. An example of such a filter is a language filter. With this filter, the researcher can set up in which languages the articles in the search result may be written, for example, only in English. In the article, all limits should be reported, so that the reader can assess the impact of these limits on the search yield. For example, if the researcher has set the language filter to ‘English’, he will not find any studies published in any other language. However, it is quite possible that several and important RCTs on the question of the SR have been published in another language. These are then missed in the search and do not count when determining the overall estimate of the effect. This may introduce a form of bias: language bias. As part of the search strategy, the researchers also determine the time period to be searched. For example, one can only search for articles that have appeared in the last 10 years. The relevant time period depends on the subject and the research question. The reader must judge for him/herself whether the researchers have made the right choices in delimiting the time period. A notorious form of bias in an SR is publication bias. Because journals prefer to publish research in which a significant difference between interventions is reported (‘positive’ trials), part of the research in which no effect is found (‘negative’ trials) is not published. Also, some negative trials may be published in journals that are more difficult to find. Because these journals are not indexed in the known large databases, these negative trials are more difficult to find in the databases. When searching in PubMed and EMBASE only, articles from these journals cannot be found. If this is the case, positive trials may prevail because a part of the negative trials are not found. Publication bias may therefore lead to an overestimation of the effect found in the SR.

10.2  Synthesis: What Is the Value of Systematic Reviews?

177

10.2.5 What Are the Requirements for the Selection Procedure of the Articles for the SR? Usually, a large number of articles are found in the systematic search. Machado et al. ([9], p. 4) found 4037 articles. After selection, only 12 articles remained. Large numbers of articles, such as these 4037, are not unusual. After all, the researchers do not want to miss any relevant articles. So, the researchers need an adequate method to select the relevant articles. In order to determine which articles are relevant, the researchers have drawn up inclusion criteria and exclusion criteria beforehand, which they use in the selection (Sect. 12.55). When selecting relevant articles from the large numbers of articles found in the search yield, there is a risk that a relevant article will escape the reviewer’s attention. To prevent relevant articles from being missed and therefore excluded, or conversely articles from being wrongly included, the selection must be carried out by two reviewers working independently of each other. After they have both assessed all articles found on the basis of the inclusion and exclusion criteria, they compare their selection. If an article is found that was selected by one of the reviewers but not by the other, the reviewers discuss whether the article should be selected or not. If the reviewers cannot reach consensus on this, a third reviewer is usually asked to give a final opinion on the selection of that article. The selection takes place step by step. Initially, the reviewers select by reading the title and abstract of the articles from the search yield. These provide sufficient information to exclude the majority of the search results. The remaining articles are then assessed for their full text. The full text assessment leads to the final selection. The SR must contain a flow chart of the selection process. The flow chart in the SR of Machado et al. ([9], p. 3) shows that the reviewers had a search yield of 4037 articles, from which 3921 articles were selected by screening the titles and abstracts. The reviewers went through 116 full text articles and ultimately included 12 articles in the review.

10.2.6 How to Assess Whether a Systematic Review Is Based on Good or Weak RCTs? Systematic reviews of interventions are mainly based on RCTs, because these provide the most reliable evidence about the effects of interventions. However, it is almost impossible to conduct 100% bias-free RCTs. Nevertheless, evidence is only reliable if the RCT in which the evidence was found was not (too much) affected by bias. Therefore, the reviewers carry out a quality assessment of the RCTs included in the SR. For each RCT, the reviewers estimate the risk that one or more forms of bias may have influenced the results of the RCT. This quality assessment is intended to prevent potential bias from the RCTs influencing the results of the SR as much as possible. There are several tools to assess RCTs for their quality. The most important of these is the Cochrane Collaboration’s tool [10, 11]. Using this tool, the reviewers

178

10  What Can We Do About the Problem: How to Critically Appraise Articles…

identify the risk of bias for each RCT. The tool distinguishes six domains within which different types of bias can occur (see Table  10.2). For each domain, the reviewers assess whether the risk of bias in each RCT is high, low or unclear. Just as the selection of articles should be carried out by two reviewers independently, so should the quality assessment of the selected RCTs be carried out by two reviewers and independently. Afterwards, both reviewers compare their results, discuss any differences and try to reach consensus. This approach limits the risk of information bias in the SR: not using the correct information, or using the information incorrectly, or adopting it incorrectly. Finally, the results of the quality assessment should be reported clearly in text and via a figure in the results section of the SR. Table 10.2  Sources of bias in RCTs cited in the Cochrane Collaboration’s tool [10, 11] Domain Bias arising from randomisation

Bias type Selection bias

Bias due to deviations from the experimental intervention Bias when measuring the outcome

Performance biasa,b

Bias arising from lack of measurement results of study participants

Attrition biasa

Bias because not all study results were reported Further bias

Reporting bias

Detection biasa,b

Other bias

Note – Arises if the RCT was not adequately randomised (no random sequence generation) – Or arises if in the RCT the allocation concealment is performed inadequately – Arises if in the RCT the participants or the staff were not blinded – Or arises if in the RCT during the study the blinding of the participants or staff is broken – Arises if in the RCT the effect assessors were not blinded – Or arises if in the RCT the blinding of the effect assessors is broken during the study – Arises if in the follow-up to the RCT the measurements were not carried out for all the included participants without this being clearly reported and/or without this being adequately processed in the analyses (incomplete outcome data) – Arises if more outcomes are measured in the RCT than reported in the article (selective reporting) – This occurs if there are biases in the RCT that do not fall under any of the above categories (an example of this is response bias, which often occurs if outcomes are measured solely via participants’ self-reporting. Many participants do not give the correct information, but socially desirable information)

It may be that the risk of these biases varies by outcome within the RCTs Note that performance bias and detection bias arise from unconscious psychological mechanisms. If an effect assessor deliberately performs the measurements differently or adjusts them, this is not detection bias, but fraud (for an example of response bias, see Sect. 12.56) a

b

10.2  Synthesis: What Is the Value of Systematic Reviews?

179

10.2.7 What Is Data Extraction? What Are the Requirements for Reporting on the Data Extraction Process in a Systematic Review? Data extraction is the accurate incorporation of all key data from individual RCTs into the reviewers’ dataset. In the methods section of the SR, reviewers should indicate what information they are extracting from each article. The most important information that the reviewers should include in the RCTs is information about the effects found in the RCTs, also including the confidence intervals. The extracted data also include the information about the study population, the intervention and the outcomes. The review should also state how many reviewers carried out the data extraction and how this was done. For this component too, it is important that two reviewers independently carry out the data extraction. After all, information bias can also arise with this task.

10.2.8 In What Way Should the Original Studies Be Described in the SR? The SR synthesises information from RCTs. If the SR also includes a meta-­analysis, the SR pools the results from those RCTs. The main information from the original RCTs should therefore be presented in the SR in an orderly fashion. Key information includes at least: –– –– –– –– –– –– ––

research design (RCT, CCT, cohort study), sample size, patient population, experimental intervention, control intervention, follow-up duration, primary outcomes,

Because presentation in a table improves the overview for the reader, all these details are best reported in table form. In addition, the text of the SR can elaborate on the main points of these characteristics.

10.2.9 What Is Heterogeneity? How Can It Be Verified That Heterogeneity Has Been Correctly Estimated and Correctly Incorporated into an SR? Heterogeneity is the phenomenon that individual RCTs differ too much from each other to be able to be pooled without further ado. (In the case of homogeneity, there is sufficient similarity of RCTs.) If there is too much heterogeneity, the results

180

10  What Can We Do About the Problem: How to Critically Appraise Articles…

cannot be pooled just without further consideration. There are two forms of heterogeneity: clinical heterogeneity and statistical heterogeneity. In clinical heterogeneity, there are significant differences between individual RCTs with respect to one or more of the following components: –– patient population (e.g. differences in age or health condition), –– setting (e.g. if some RCTs were conducted in a hospital and others in a community setting), –– interventions (e.g. when some RCTs refer to ‘fall prevention’ as removing mats and obstacles, and others to physical exercises to improve body balance), –– outcome measures (e.g. when some RCTs use one cut-off point for a pain scale and other RCTs use a different cut-off point), If one or more of these components between the RCTs on which the SR is based deviate too much from one another, a meta-analysis would be comparing apples with oranges, as it were. Pooling is then impossible. If there is too much clinical heterogeneity, an SR should be limited to an overview and a qualitative summary of the results of the individual studies. Sometimes it happens that a clear subgroup of RCTs can be identified. For example, the RCTs may be heterogeneous in terms of age, but they do contain clear age groups. The reviewers probably saw this coming in advance and beforehand defined a number of subgroups that are comparable. Within each of these subgroups, the RCTs can then still be pooled, because there is clinical homogeneity within the subgroups. There is statistical heterogeneity if there is sufficient clinical homogeneity, but the results of the different RCTs differ too much. For example, consider a study of fall prevention where the same intervention was studied in three RCTs, and there was no clinical heterogeneity. Suppose the following three relative risks with 95% confidence intervals were found: RCT 1—RR = 4.1 (2.0 to 6.2). RCT 2—RR = 0.7 (0.2 to 1.2). RCT 3—RR = 3.4 (0.8 to 6.1). The results of the second RCT are clearly different from the other two RCTs. Indeed, the confidence interval of RCT 1 does not overlap with the confidence interval of RCT 2. This indicates a high statistical heterogeneity. When critically assessing an SR, the nurse should always consider whether there is clinical or statistical heterogeneity. In the case of clinical heterogeneity, this can be done using the table summarising the individual RCTs. If the reviewers have carried out a pooling, they should show a forest plot of the pooled outcome in a figure of the pooled outcome (see Fig. 10.2). Using the link https://tinyurl.com/efficacy-­safety-­paracetamol, go to the article by Machado et al. [9], and then look up Fig. 3 on page 6.

10.2  Synthesis: What Is the Value of Systematic Reviews?

181

Fig. 10.2  Forest plot of comparison: I All patients, outcome: I.I Mortality. Example of a forest plot in a meta-analysis from the Cochrane Database of Systematic Reviews: corticosteroids versus placebo in acute bacterial meningitis. The results of 22 RCTs are clearly presented in this forest plot. The results of each study on the outcome ‘mortality’ are presented in figures (risk ratio, which is another name for relative risk) and in the forest plot. In the forest plot, each horizontal line represents the 95% confidence interval. The square on the line represents the point estimate of that study. This makes it easy to compare the results of the individual original studies. The bottom row (Total) represents the pooled outcome. When calculating the pooled outcome, not every original study carries the same weight. How much weight each study carries is shown in the Weight column. Note that 25 RCTs are included in this figure, but that no risk ratio could be calculated for three studies. Therefore, the pooled outcome is based on 22 RCTs. The vertical line indicates a risk ratio of 1 (which is the neutral value). Point estimates to the left of the line show a risk ratio below 1, indicating that the intervention reduces the risk of death from bacterial meningitis compared with the control intervention (in this case, placebo). Point estimates to the right of the line indicate a more favourable result for the control intervention. The p-value for heterogeneity is 0.18, and the I2 is 21%. Thus, there seems to be no indication of statistical heterogeneity. The pooled effect of the intervention is a risk ratio of 0.90 (95% CI 0.80–1.01). Because the neutral value of 1 is included in the confidence interval, this is referred to as a statistically non-significant result, and according to this meta-analysis, it remains unclear whether corticosteroids are more effective than a placebo in preventing death from acute bacterial meningitis. (Source: Brouwer et al. 2015 [12])

In the SR by Machado et al. [9], pooling did take place. Figure 3 on page 6 shows forest plots of eight subgroups. The square blocks indicate the point estimate of each RCT, while the dashes through the square blocks represent the confidence interval. These data are also shown in the figure as numbers at the end of the line. The bottom line of each forest plot shows the pooled effect: first there is an I2 with a percentage, followed by a diamond in the figure itself, and finally the last column shows the newly calculated effect as point estimate with confidence interval. This shows the effect and confidence interval calculated after pooling the RCTs.

182

10  What Can We Do About the Problem: How to Critically Appraise Articles…

The centre of the diamond indicates the point estimate, while the ends represent the limits of the confidence interval. The first forest plot (Spinal pain [pain/immediate term]) shows an I2 of 0% and a mean difference of 1.4 (95% CI −1.3 to 4.1). This paragraph is concerned with checking for high statistical heterogeneity. There are various possible methods for this check. The first method is to check whether the confidence intervals of the individual RCTs overlap sufficiently in the forest plot. If there is insufficient overlap, the statistical heterogeneity is high. Figure 3 of the SR by Machado et al. [9] shows a forest plot in which there is indeed little overlap between the different confidence intervals (this concerns the forest plot about Osteoarthritis, disability/immediate term). The second method is by looking at the p-value of the chi-square test, which is sometimes listed at the bottom of the forest plot. This test is used to test the difference between studies. A p-value lower than 0.05 indicates heterogeneity. A p-value higher than 0.05 means that there is no heterogeneity. The chi-square test is often expressed with the symbol χ2 or x2. However, the best method is via the I2 statistic. The lower the percentage specified for I2, the lower the statistical heterogeneity. A limit is set at I2 of 60%. Above that level, statistical heterogeneity is generally high. In the past, it was customary not to pool when statistical heterogeneity was high. Nowadays, it is customary to always pool when there is no clinical heterogeneity. The argument for this is that, despite high statistical heterogeneity, the new point estimate with confidence interval calculated by pooling is still the best estimate of the true effect. Because the researchers provide an I2, the reader can see how high the statistical heterogeneity is, and with high statistical heterogeneity, the reader should remember that the point estimate cannot be relied on as strongly as with low I2 [13]. Machado et al. [9] also provide the I2 statistic. The forest plot showing that there is little overlap between the confidence intervals indeed has a high I2 at 84%.

10.2.10 How Should the Results of an SR Be Reported? The way in which the results should be reported depends on whether or not pooling has taken place. If no pooling took place, the results section of the article contains a qualitative description of the studies included and an overview (in a table) in which at least the following components are described per study: –– the primary outcome of the study, –– the measure of effect for the primary outcome, –– the point estimate with confidence interval for the primary outcome, If pooling has taken place, there must be an overview of the results of the included studies with the new effect estimate in the form of a forest plot. It should also be clear how the quality of the included studies (the risk of bias) was processed in the meta-analysis. This processing in the meta-analysis of studies with a high risk of bias can be done in different ways. The reviewers can (1) limit the meta-analysis to

10.2  Synthesis: What Is the Value of Systematic Reviews?

183

the RCTs with a low risk of bias, (2) carry out several meta-analyses, with a different risk of bias for each meta-analysis, or (3) include all RCTs in the meta-analysis and leave the assessment of the risk of bias to the reader. After all, the reader was given an overview of the risk of bias per study earlier in the SR and can therefore make his or her own assessment of how certain the newly calculated effect is. This also enables the reader to determine the severity of the various domains of bias. In general, it is assumed that selection bias, detection bias, attrition bias and reporting bias can distort research results the most [14] (Sect. 12.57).

10.2.11 How to Assess the Validity of an SR? A study is valid if it contains no systematic errors. Systematic errors in the research (bias) influence the results, as a result of which the results do not reflect the actual values. The systematic review assessment checklist allows nurses to identify the key sources of bias in the conduct of an SR (questions 1–8 of the checklist). Based on the answers to these questions, the nurse can assess the validity of the SR. If the validity of the SR is indeed satisfactory, then the nurse should consider whether the research findings of the key outcomes are convincing enough to justify application in their practice.

10.2.12 How Can the Results of an SR Be Interpreted? How to Assess the Applicability of the Intervention in Patient Care? For these questions, both the assessment of the results and the assessment of the applicability of the intervention in patient care are done in the same way as for an RCT (see Sect. 10.1.16).

10.3 Evidence-Based Guidelines 10.3.1 What Is a Guideline? A guideline is a way of making the results of well-conducted scientific research available and usable for healthcare professionals in a simple manner. A working group of experts systematically searches for all the evidence on the basis of various clinical questions relating to a disorder and selects, assesses and interprets the evidence. Together with patients, the workgroup then writes the evidence-based guideline. This contains recommendations for decisions about prevention, diagnosis, treatment and care. International guidelines can be found via, for example, www. nice.org.uk (British) [15] and https://www.nccih.nih.gov/health/providers/clinicalpractice (American) [16] or https://www.guidelinecentral.com/guidelines/# (American) [17].

184

10  What Can We Do About the Problem: How to Critically Appraise Articles…

An evidence-based guideline reflects the scientific state of the art at the time the guideline was written. The recommendations that the guideline gives are not non-­ binding, but are information that determines the decision. If the healthcare professional decides to deviate from these recommendations, he or she must substantiate this deviation with reasons.

10.3.2 What Value Do Guidelines Have for Clinical Reasoning? A good evidence-based guideline is based on a systematic search for evidence, assessment of the evidence by experts, the clinical expertise of those experts and patient preferences. This makes a good guideline a valuable tool in clinical reasoning. If the nurse has access to a recent evidence-based guideline that addresses the nurse’s clinical question, they can base their decision on the guideline. There are three situations in which it is important for the nurse to search for (additional recent) evidence. 1. If the guideline is out of date. If the guideline was published 2 years ago or more, an additional search is recommended. Guidelines older than 5 years are considered outdated. 2. If the guideline does not answer the nurse’s clinical question. Not all topics and questions are covered in the guideline. 3. If the quality of the guideline is insufficient. International guidelines are generally of good quality. However, guidelines of poor quality can also be found on the internet. The recommendations are supported by evidence in the guidelines. For each recommendation, the level of evidence on which the recommendation is based is also indicated. This makes it immediately clear whether the recommendation is based on strong evidence (one or more systematic reviews) or on weak evidence (e.g. expert opinion). Recommendations based on recent evidence with a strong evidence base are more urgent than recommendations based on weaker evidence.

10.3.3 How Can the Quality of a Guideline Be Critically Assessed? Good guidelines are written according to a specific, systematic method. Because a guideline is expected to provide answers to the main difficulties in care decisions on prevention, diagnosis, treatment and care, guideline development begins with an analysis of problematic issues and the formulation of questions for a systematic literature search. Then a systematic literature search is conducted, literature is selected, assessed and interpreted. The guideline is therefore the product of an extensive and systematic literature search conducted by experts. If properly conducted, the guideline has a strong evidence base.

10.3  Evidence-Based Guidelines

185

The checklist used to assess guidelines is AGREE II. This checklist is available online [18]. The AGREE II contains 23 key items divided into six domains: 1 . scope and purpose, 2. stakeholder involvement, 3. rigour of development, 4. clarity of presentation, 5. applicability, 6. editorial independence. Each key item is rated on a 7-point scale, where 1 means “strongly disagree” and 7 means “strongly agree”. It is intended that several appraisers will assess the guideline independently with the AGREE II and then discuss their respective scores. The percentage of the maximum number of points allocated to each domain is calculated. Because the minimum scale score is 1, the assessor must award at least 1 point for each key item. But to arrive at the correct percentage, he must subtract this one point from the calculation: for each key item, he must subtract 1 point from his score and 1 point from the maximum. For example, assume that an assessor has scored 6 + 6 + 5 = 17 points in total for the three key items in Domain 1 (Scope and Purpose). The maximum score in Domain 1 is 3  ×  7  =  21. But when calculating the percentage, the maximum is 21–3 = 18, and the number of points scored is 17–3 = 14. In this example, the calculation of the percentage for Domain 1 would be (14/18) × 100 = 78%. The authors of the AGREE II do not specify limits for the minimum percentage per domain. In the tool, the authors emphasise that the user should determine what they consider acceptable scores per domain based on the way they plan to use the guideline and the context in which they plan to use it. The 23 key items are clearly formulated. As a result, the questions from domains 1, 2, 4, 5 and 6 about the guideline are easy to read and easy for healthcare professionals to assess, without the need for specific additional knowledge. The questions in domain 3 about the rigour of development are crucial to the reliability of evidence-­ based guidelines. However, to assess and answer the key items in this domain of the AGREE II requires specific knowledge. This knowledge can be found in the previous sections of this book.

Glossary Aggregation level  Level at which data are merged. Allocation concealment  Method of distributing patients among the intervention group or the control group, in which this distribution is organised in such a way that no one can influence it. The distribution is completely determined by chance.

186

10  What Can We Do About the Problem: How to Critically Appraise Articles…

Ascertainment bias  Systematic deviation of study results from the true values, caused by knowledge of patient classification in the intervention or control group, for example, because blinding was breached. Attrition bias (selective dropout)  Bias caused by systematic differences in dropout between groups. Dropout of study participants can jeopardise the comparability of the groups, for example, because:   –  in one group many more people drop out;   – more people with certain patient characteristics drop out than in the other group;   – there is a dropout from research participants (from both groups) with a certain characteristic, such as the fact that they only have mild complaints, or there are people who drop out who think that their health problem has already been solved, while the research is still ongoing. As a result of research participants dropping out, no more outcome measurements are possible for them, and the outcome data are therefore incomplete. Bias (in scientific health research)  In scientific health research, bias includes any factor or process that tends to deviate the results or conclusions of a trial systematically away from the truth ([19], p. 28). Blinding  Measure to ensure that one or more groups involved in a study (patients, caregivers, treatment providers, effect assessors) do not know whether patients are assigned to the experimental group or the control group. Clinical heterogeneity  When the original studies in a systematic review differ too much from each other with respect to clinical characteristics of the patients, settings, interventions or outcome measures, this is called clinical heterogeneity. For example, a study of the effect of a particular intervention in children up to 15 years of age will differ considerably from a similar study of the same intervention in people aged 50 years and older. Because of these differences, these clinically heterogeneous studies may not be pooled and cannot be included in a meta-analysis together. Clinical homogeneity If, in a systematic review, the original studies are sufficiently similar with regard to clinical characteristics of the patients, settings, interventions or outcome measures, this is called clinical homogeneity. If there is sufficient clinical homogeneity between studies they can be pooled and included together in a meta-analysis. Clinical relevance  The extent to which the effect found in the experimental intervention is important enough for patients to warrant implementation. Confidence interval  The confidence interval (CI) indicates the accuracy of the calculated point estimate of the measure of effect. It is common practice to calculate the 95% confidence interval around the point estimate. The 95% confidence interval indicates that if the study is validly conducted, there is a 95% chance that the actual effect lies within this confidence interval. A confidence interval is needed because research results are based on a sample and not the entire population. The confidence interval indicates that there is a 95% chance that the effect in the entire population lies within the interval.

10.3  Evidence-Based Guidelines

187

Comorbidity  Having two or more diseases or disorders at the same time. An example is a patient with a heart condition who is also depressed. Control group  Group in which the study participants receive the control intervention. This is an intervention of which the effect is already known and with which the researchers want to test whether the experimental intervention indeed has a better effect than the control intervention. Controlled clinical trial  Intervention study in which the assignment of the study participants to the experimental or control group was not or not entirely carried out at random. Detection bias  Systematic differences between the study groups in the way the outcomes were obtained. The main measure for preventing this bias is blinding the effect assessors. This measure is particularly important when measuring subjective outcomes, such as sleep quality and pain. Evidence-based guideline  Recommendations based on scientific evidence to support clinical decision-making and linked to a specific patient problem or health condition. The evidence on which the recommendations are based is clearly visible in the guideline. Guidelines are written by independent experts in the field of the content of the guideline. Guidelines reflect the state of the art, but also address topics that are still being debated in practice and, in many cases, also practice variation. Evidence-based guidelines translate research results into practice. The recommendations are therefore not non-binding, but departures from them may be justified (see also: guideline). Exclusion criteria  Criteria that define the characteristics of study participants that must not be present in order to participate in the study. If a potential participant has one or more of these characteristics, he is ‘excluded’, that is, removed from the group of potential participants. Experimental group (intervention group)  Group in which study participants are exposed to an experimental intervention, in which the researchers believe that this intervention is better than the alternative interventions available at that time. Forest plot  Graphic representation of pooled study results in a meta-analysis. A forest plot provides an overview of the point estimates and confidence intervals of the individual studies at a glance. This makes it easy to see the extent to which the point estimates and confidence intervals of the pooled studies overlap. A ­forest plot also shows the recalculated point estimate of the effect with the associated confidence interval of the outcome studied. Generalisability (external validity)  The extent to which the results found in the sample of a study are also valid for individuals with similar health problems and characteristics, who did not participate in the study (e.g. a question might be: are these study results valid for individuals with similar health problems but who live in a different continent, or who have a different culture?). Guideline  A guideline is a way of making the results of well-executed scientific research available and usable for healthcare professionals. A working group of experts systematically searches for all the evidence for a condition and selects,

188

10  What Can We Do About the Problem: How to Critically Appraise Articles…

assesses and interprets this evidence. Together with patients, the workgroup then writes the evidence-based guideline, in which the workgroup makes recommendations for decisions about prevention, diagnosis, treatment and care (see also: evidence-based guideline). I2  I squared (or the I squared statistic) is a measure in which the statistical heterogeneity between the individual studies in a systematic review with meta-analysis can be expressed. The I2 is expressed as a percentage. The higher the percentage, the greater the statistical heterogeneity. Inclusion criteria  Criteria that define the characteristics of the study participants that must be present in order for them to participate in the study. Information bias (in an RCT, cohort or case-control study) Bias in the study results, caused by errors in the information due to measurement errors. Information bias can occur, for example, when a number of study participants are assigned to the wrong exposure group or when an incorrect outcome is registered (misclassification). Information bias (in an SR)  Failure to use the correct information, or use information incorrectly, or making mistakes when extracting data from original studies. Intention to treat  Method in an RCT in which all participants who are allocated to the intervention or control group via randomisation are also analysed in the same group, even if they have ended their participation in the study in the interim or if they have withdrawn immediately after randomisation. Intervention group (experimental group)  Group in which study participants are exposed to an experimental intervention, in which the researchers believe that this intervention is superior to alternative interventions available at the time. Language bias Possible deviation from the truth of the results of a systematic review because only studies published in certain languages was searched for. For example, if researchers only searched for publications in English, they may have missed an important French-language research publication. That study would then not be included in the calculation of the new overall effect via a meta-analysis. Measure of effect  Measure expressing the effect, for example, relative risk (RR), risk difference (RD) and number needed to treat (NNT). Meta-analysis  Analysis that may be included in a systematic review. In a meta-­ analysis, the results of the individual studies included in the systematic review are combined (pooling). By pooling the results, a new effect estimate can be calculated which is more reliable than the effect estimates of the individual studies. Original study  Direct patient-reported studies on diagnosis, etiology, therapy or prognosis. Outcome  Outcome refers to that which is influenced through an intervention. Usually an outcome is stated in one or two words, for example, ‘weight’, or ‘social integration’. Please note: there is a difference between ‘outcome’ and ‘result of the intervention’. The outcome indicates in general terms what is being worked on, whereas the result of the intervention indicates precisely what the

10.3  Evidence-Based Guidelines

189

result of that intervention is for the patient. If the outcome is ‘weight’, the result may be ‘gained 3 kg’. PICO  Acronym for: Population, Intervention, Comparison (or Control) and Outcome. PICO is a way of identifying the most important keywords in a systematic literature search for intervention research in scientific databases. Based on the research question, the most important keywords are formulated for each of the four PICO elements. Point estimate (effect estimate)  The magnitude of the effect as found in a particular study. The point estimate is the best estimate of the actual effect based on that study. But because the study is based on a sample, it is unlikely that the point estimate is exactly the actual effect in the entire population. Pooling  In a systematic review: pooling the results of the individual studies included in the systematic review. The aim is to create a large sample on the basis of which an overall effect size can be calculated in a meta-analysis. This estimated overall size is more reliable than the effect estimates of the individual studies. Publication bias  Bias in systematic reviews and meta-analyses which is caused by the preference of scientific journals for publishing articles in which significant effects have been found (‘positive’ results). Such articles have a greater chance of being published than articles without significant effects (‘negative’ results). Because some of the studies with a negative result are not published, they cannot be found in the literature search and thus will not be included in the meta-­ analysis. Publication bias leads to an overestimation of the overall estimate of an effect. P-value  The p-value indicates the size of a probability. This value is expressed as a number between 0 and 1. A probability of 0 for a certain phenomenon means that there is no chance (or 0% probability) of that phenomenon occurring. A probability of 1 means that it is 100% certain that this phenomenon will occur. P-values are easily converted into percentages by multiplying them by 100: a p-value of 0.02 is equivalent to 2% and a p-value of 0.05 is equivalent to 5%. Randomisation  Randomly (i.e. entirely by chance) assigning study participants to an experimental group and a control group. Randomised controlled trial (RCT)  Experimental research into the effectiveness of an intervention. Participants are randomly assigned to an experimental group or a control group. Depending on the group in which the participant is placed, he will receive an experimental intervention or a control intervention. After the intervention, the effects of both interventions are compared to determine the effect of the experimental intervention on predefined outcomes. Relative risk or Risk Ratio (RR)  Association measure that expresses the ratio of the probability of a particular outcome of the intervention group to that of the control group. The RR answers the question: how many times greater or smaller is the probability of the outcome after the experimental intervention than after the control intervention?   –  An RR of 1.0 means that there is no difference between risk or probability of the outcome. The value of 1.0 is the neutral value of a RR.

190

10  What Can We Do About the Problem: How to Critically Appraise Articles…

  –  An RR between 0 and 1.0 means that the intervention group has a lower risk or probability of the outcome than the control group.   –  An RR higher than 1.0 means that the intervention group has a higher risk or probability of the outcome than the control group. Take a study in which the intervention consists of a specific diet. In this study, a successful result is defined as: weight loss of at least 5 kg within 2 months. Suppose that 10% of the participants in the control group achieved this outcome.  – −RR = 2.0 means that the group that received the experimental intervention was twice as likely to reach this outcome as the control group. So, if the control group had a 10% chance of losing five kilograms within 2 months, the intervention group has 2 × 10% = 20% chance of losing five kilograms within 2 months.  – −RR = 1.0 therefore means that the intervention group has 1 × 10% = 10% chance of losing five kilos within 2 months. With an RR of 1.0 both groups have an equal chance of the outcome, there is no difference.  – −RR = 0.8 means that the intervention group has a 0.8 × 10% = 8% chance of achieving the outcome. Reporting bias (selective reporting bias)  Bias in a study report or article whereby not all the outcomes investigated are reported, but only some of them. In most cases of reporting bias, only those outcomes are reported for which statistically significant results have been found. Selection bias (in an RCT or cohort or case-control study)  Bias of study results caused by systematic errors in selecting study participants. Selection bias (in an SR)  Bias of the study results of the systematic review caused by systematic errors in the selection of the studies included in the systematic review. Sensitivity analysis  In sensitivity analysis, the researchers calculate the way in which the effect changes if they do assign a result to the dropout participants. It can be determined if and how much the results change if the missing data of the dropped-out patients are filled in with different values. Statistical heterogeneity (in a meta-analysis) The phenomenon whereby the results of the underlying studies included in the meta-analysis vary widely. For example, a meta-analysis may include both studies with a positive result and studies with a negative result, and where the confidence intervals of these studies do not overlap or overlap only slightly. Statistical significance  A statement made in research, based on statistical testing, about how certain it is that the experimental intervention is more effective than the control intervention (or about how certain it is that a possible risk factor does indeed contribute to the occurrence of a particular health problem; or about how certain one is about the diagnostic accuracy of a particular diagnostic test). Study population  The group of potential study participants after application of the inclusion and exclusion criteria.

References

191

Synthesis  Systematic review in which, based on a research question, information from various studies is combined according to a fixed and strict method. The results have a greater probative value than each of the individual studies used for the systematic review. Validity  Absence of systematic errors. For example, in critical appraisal, validity is the extent to which the limitations in study design and in the conduction of the study leave a study vulnerable to systematic error ([20], p. 677).

References 1. Snaterse M, Deckers JW, Jørstad HT, De Bacquer D, Peters R, Jennings C, et al. Smoking cessation in European patients with coronary heart disease. Results from the Euroaspire IV survey: a registry from the European Society of Cardiology. Int J Cardiol. 2018;258:1–6. https:// doi.org/10.1016/j.ijcard.2018.01.064. 2. Kotseva K, De Backer G, De Bacquer D, Rydén L, Hoes A, Grobbee D, et al. Lifestyle and impact on cardiovascular risk factor control in coronary patients across 27 countries: results from the European Society of Cardiology ESC-EORP Euroaspire V registry. Eur J Prev Cardiol. 2019;26:824–35. https://doi.org/10.1177/2047487318825350. 3. Minneboo M, Lachman S, Snaterse M, Jorstad HT, Ter Riet G, Boekholt SM, et al. Community-­ based lifestyle intervention in patients with coronary artery disease: the Response-2 trial. J Am Coll Cardiol. 2017;70:318–27. https://doi.org/10.1016/j.jacc.2017.05.041. 4. Dutch Cochrane Centre. https://netherlands.cochrane.org/beoordelingsformulieren-­en-­andere-­ downloads. Accessed 27 Sept 2020. 5. Assendelft WJJ, Korevaar JC, Reisma JB.  Prognose. In: Scholten RJPM, Offringa M, Assendelft WJ, editors. Intleiding in evidence-based medicine. Klinisch handelen gebaseerd op bewijsmateriaal. 4th ed. Houten: Bohn Stafleu van Loghum; 2014. 6. Neumann I, Akl EA, Vandvik PO, Agoritsas T, Alonso-Coelho P, Rind DM, et al. How to use a patient management recommendation: clinical practice guidelines and decision analysis. In: Guyatt G, Rennie D, Meade MO, Cook DJ, editors. Users’ guides to the medical literature. A manual for evidence-based clinical practice. 3rd ed. New York: McGraw Hill Education; 2015. 7. Fletcher RH, Fletcher SW, Fletcher GS.  Clinical epidemiology. The essentials. 5th ed. Philadelphia: Lippincott Williams & Wilkins; 2014. 8. Guyatt G, Jaeschke R, Meade MO.  Why study results mislead: bias and random error. In: Guyatt G, Rennie D, Meade MO, Cook DJ, editors. Users’ guides to the medical literature. A manual for evidence-based clinical practice. 3rd ed. New York: McGraw Hill Education; 2015. 9. Machado GC, Maher CG, Ferreira PH, Pinheiro MB, Lin CWL, Day RO, et al. Efficacy and safety of paracetamol for spinal pain and osteoarthritis: systematic review and meta-analysis of randomised placebo controlled trials. BMJ. 2015;350:h1225. www.bmj.com/content/bmj/350/ bmj.h1225.full.pdf. https://doi.org/10.1136/bmj.h1225. 10. Higgins JPT, Altman DG, Gøtzsche PC, Jüni P, Hoher D, Oxman AD, et al. The Cochrane Collaboration’s tool for assessing risk of bias in randomised trials. BMJ. 2011;343:d5928. https://doi.org/10.1136/bmj.d5928. 11. Higgins J, Thomas J, Chandler J, Cumpstom M, Li T, Page M, et  al. Cochrane handbook for systematic reviews of interventions. Version 6.1. 2020. https://training.cochrane.org/handbook/current. Accessed 10 Oct 2020. 12. Bouwer MC, McIntyre P, Prasad K, Van de Beek D. Corticosteroids for acute bacterial meningitis. Cochrane Database Syst Rev. 2015;9:CD004405.

192

10  What Can We Do About the Problem: How to Critically Appraise Articles…

13. Murad MH, Montori VM, Ioannidis JPA, Neumann I, Hatala R, Meade MO, et al. Understanding and applying the results of a systematic review and meta-analysis. In: Guyatt G, Rennie D, Meade MO, Cook DJ, editors. Users’ guides to the medical literature. A manual for evidence-­ based clinical practice. 3rd ed. New York: McGraw Hill Education; 2015. 14. Assendelft WJJ, Scholten RJPM, De Ruijter W, Bouter LM.  Systematische reviews van interventieonderzoek. In: Scholten RJPM, Offringa M, Assendelft WJ, editors. Inleiding in evidence-­based medicine. Klinisch handelen gebaseerd op bewijsmateriaal. 4th ed. Houten: Bohn Stafleu van Loghum; 2014. 15. National Institute for Health and Care Excellence (NICE). www.nice.org.uk. Accessed 21 Sept 2022. 16. National Center for Complementary and Integrative Health (NIH). www.nccih.nih.gov/health/ providers/clinicalpractice. Accessed 21 Sept 21 2022. 17. Guideline Central. www.guidelinecentral.com/guidelines/#. Accessed 21 Sept 2022. 18. AGREE Next Steps Consortium. https://www.agreetrust.org/resource-­centre/. Accessed 21 Sept 2022. 19. Jadad A. Randomise controlled trials. London: BMJ Books; 2007. 20. Guyatt G, Rennie D, Meade OM, Cook DJ, editors. Users’ guides to the medical literature. A manual for evidence-based clinical practice. 3rd ed. New York: McGraw Hill Education; 2015.

How to Better Understand Patients’ Experience: Understanding Psychosocial Processes in Healthcare Through Qualitative Research

11

Still, some questions remain. All that research has made a lot clear to me: which interventions are, and which are not effective? I can find out whether alcohol is an etiological factor for certain forms of cancer. I also understand that once people have a disease or a disorder, there are all kinds of influencers of the course of the disease and of its final state. But what I don’t know is how patients, their carers and nurses experience these interventions, and how they experience the disease. I want to be an empathetic nurse. Then I need to understand the patients and my colleagues as well as possible. What research do I need for that?’

11.1 What Is Qualitative Research? What Is the Difference Between Quantitative and Qualitative Research? Being active, exercising, meeting people, maintaining contacts and taking care of yourself are important aspects for the elderly in growing old as healthily as possible. Quantitative research can, for example, ask questions about the average ADL score of older people living independently and about effective interventions to support ADL. For this, measurement instruments are used, such as the Barthel index, which expresses the extent to which people are capable of ADL. Qualitative research takes a different approach. It places the emphasis on understanding what is going on. How does the process of change to less and less ability take place? What do these changes mean for older people? [1].

With contribution by Pim Assendelft. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 J. Dobber et al., Clinical Reasoning and Evidence-Based Practice, https://doi.org/10.1007/978-3-031-27069-7_11

193

194

11  How to Better Understand Patients’ Experience: Understanding Psychosocial…

Whereas quantitative research is about quantities, numbers and probabilities, qualitative research is about what happens, how that process proceeds and what it means for those involved. In other words, it is about understanding psychosocial processes and the attribution of meaning by those involved (Sect. 12.58). An important characteristic of qualitative research is that it uses observations from the everyday reality of the persons or situations studied, to work towards hypotheses or theories that transcend the situation studied. The data come from the research situation, while the analysis of the data aims to obtain knowledge which goes beyond the research situation. Thus, the hypothesis or theory is the end result of the research. Whereas (for example) in an RCT, hypotheses about effects of interventions are tested in the research, in qualitative research, no hypothesis is tested. Instead, a hypothesis is developed (which may be tested in a later study).

11.2 What Is the Most Important Method in Qualitative Research? In qualitative research, methods are used to arrive at verstehen: to see the perspective of the persons in the study, so that the researcher understands them better. To this end, the qualitative researcher seeks contact with the reality being studied from the very beginning of the study. The researcher does not make a detailed research plan first, but chooses ways to directly take part in relevant situations and possibly participate. This leads to qualitative research data: detailed and concrete. The intention is to develop theories and concepts through the observation of, and participating in, reality. Through this open research procedure, it becomes increasingly clear during the study what is important in everyday reality.

11.3 What Are the Ways of Collecting Data in Qualitative Research? In order to understand people’s perspectives, data from and about psychosocial processes are important. In order to obtain these data, the researcher uses qualitative research methods such as purposeful observation, talking to people, participating with them and studying written sources that report on reality. Observation is an important method of data collection. An intensive method of observation is participant observation. In this, the researcher has to participate and walk along with the people he is studying. The researcher has to register closely, without disturbing the course of events. Anne-Mei The, for example, spent a year and a half at a pulmonary department for her study into the role and function of nurses in end-of-life decisions, and how nurses perceive this role [2]. During the shadowing, the researcher observes, makes field notes of observations and

11.4  How Are Data Analysed in Qualitative Research?

195

conversations, talks over coffee, sometimes deliberately steers the conversation in a certain direction, listens, interviews, reads files and reports and compares situations and events [2]. Interviewing is another method of seeking contact with reality. This can be done through individual interviews, but also through group interviews or focus groups. Qualitative interviews are usually semi-structured. This means that the researcher has established fixed interview topics, but has not formulated fixed questions. This leaves room for differences in the interviews, while the same subjects are discussed in each interview. A third way of collecting data is to study written sources such as patient records, minutes, archives, personal diaries, letters and emails (Sect. 12.59). All these methods require small scale. It is impossible to conduct 500 in-depth interviews. Long-term monitoring and participation means that the researcher has to limit himself to one, at most two, research settings. In qualitative research, the samples are therefore small. The sample does not have to be representative. It is more important to speak to those people and to experience that situations which are most informative for the research question. Therefore, qualitative researchers often use a purposive sample, trying to include the best informants and the most informative situations. It is actually impossible to determine in advance how large the sample should be. When the researcher notices that the interviews or participation no longer provides him/her with new information, this is called saturation. When saturation is reached, the researcher can stop collecting data. The use of different measurements and different types of data (triangulation) usually increases the truth value of the research (internal validity), provided that the right sources and data collection methods are used. By approaching the subject from different angles, in different ways or by different researchers, a more complete picture of the subject is developed (Sect. 12.60).

11.4 How Are Data Analysed in Qualitative Research? Analysis is only possible if there is a question. The question determines the analysis. In the beginning of a qualitative research project, the question is still global, because contact with reality is still being sought. The starting global question is: ‘What is going on here?’ or: ‘Which basic processes and central concepts are important here?’ This exploration phase yields a provisional answer, with provisional central concepts, with which the questions can be posed more precisely and further data can be collected. In qualitative research, all data are not first collected and then analysed, but the analysis runs parallel to the collection of data (continuous comparison). The researcher writes out the data in texts (interview reports, situation descriptions, case reports). Based on the questions, the researcher can then analyse the texts by assigning codes to text fragments. The codes are arranged in categories.

196

11  How to Better Understand Patients’ Experience: Understanding Psychosocial…

For example, Van der Kluit et  al. [3] investigated the goals that hospitalised patients, aged 70 and older, set for themselves for their hospital stay. The researchers interviewed 28 hospitalised elderly patients. From the interviews, they distinguished nine categories of goals. The categories in which most goals fell were ‘alleviating complaints’ (e.g. restoring sleep); ‘improving daily functioning’ (e.g. being able to walk independently again with a walker); ‘regaining/maintaining autonomy’ (e.g. regaining/maintaining independence). The categories were then compared with the usual perspective of healthcare professionals, who generally do not look at goals from the patient perspective and do not ask patients. It also happens that healthcare professionals think that all patients have more or less the same goals. The researchers therefore conclude that there is a strong diversity of goals of elderly patients, and that the conversation about this should be encouraged. After all, the goals of the patients are not automatically clear, and if the healthcare professional knows the goals of the individual elderly patient, this can help in making clinical decisions [3], for example, about relevant outcomes and treatments. The full research article can be found online.

11.5 What Are the Pitfalls of Conducting Qualitative Research? There are a number of pitfalls when conducting qualitative research. The most important are listed in Table 11.1. However, it is difficult to judge from the research article whether these pitfalls played a role in the research. Table 11.1  Pitfalls in conducting qualitative research Pitfall Fallacy of objectivism Researcher bias/ biased viewpoint Going native

Control effect/ reactivity

Validity problem

Short description The researcher attributes the own perspective of meaning to the persons under study The researcher cannot get rid of his own way of looking at things, theories and assumptions. As a result, the researcher does not have an open view of the persons under study and their perspectives The researcher identifies himself so strongly with the persons under study that he cannot take enough distance anymore to study them in a critical-­ scientific way The researcher observes situations or phenomena that he himself has caused by his presence. This happens because persons under study, when they know that they are being observed, consciously present themselves differently and adapt their behaviour The researcher did not speak to the best informants or observe the most informative situations. As a result, reality has not been accurately described

11.6  What Are Quality Criteria for Qualitative Research?

197

11.6 What Are Quality Criteria for Qualitative Research? The results of qualitative research are trustworthy if they meet the quality requirements of qualitative research. Just as in quantitative research, these include internal validity, generalisability, reliability and objectivity. Because the nature of qualitative research is so different from that of quantitative research, the quality criteria associated with these concepts cannot be simply adopted. Qualitative research is subject to a translation of the quantitative concepts (see Table 11.2) and an elaboration of these in terms of quality criteria [4] (for a checklist to evaluate qualitative research, see Sect. 12.69).

11.6.1 Truth Value or Credibility How can the truth of the research results be established? From the point of view of qualitative research, this is a difficult question, because it focuses on psychosocial processes and on meaning. These types of processes largely take place in the human mind, to which the researchers do not have access. Furthermore, such processes have a different content for different people. The content of processes in the mind is also called mental constructs. Because there are differences between people, the reality on which qualitative research focuses consists of a multiple set (collection) of mental constructs. The study results must do justice to this set. The question is therefore whether the set of mental constructs has been adequately reconstructed in the study results. This question can be transformed into another question: is the reconstruction credible for the investigated persons? With truth value, the credibility of the research results must be checked. The credibility of the investigation can be enhanced by different practices and procedures. In Table 11.3, these procedures and measures are summarised [4] (see also Sect. 12.61). Table 11.2  Quality criteria for quantitative research translated into qualitative research Criteria quantitative research Internal validity Generalisability Reliability Objectivity

Translation to qualitative research Truth value or credibility Applicability or transferability Repeatability Neutrality or confirmability

198

11  How to Better Understand Patients’ Experience: Understanding Psychosocial…

Table 11.3  Credibility of qualitative research Point of action to increase the credibility of the research 1. Manner in which the research was conducted

2. Have the results checked by the original source 3. External check of the research process 4. External check of the results 5. Refining the working hypotheses

Measures and procedures to enhance credibility  – Prolonged engagement: by being involved with the study participants or research situation for a long time, it is more likely that the researcher will get to know the reality well  – Persistent observation: if saturation seems to be reached, do not stop observing, but continue and deepen the information obtained  – Triangulation: using different sources or different methods. If the same results are obtained in different ways, this increases the credibility of those results  – Member checking: having the original source read and check elaborated research material (for example a detailed interview)  – Peer debriefing: discussing the research process with peers (other researchers) during the course of the study (e.g. about whether choices are being made at the right time)  – Referential adequacy: part of the collected data is kept separate and stored as raw data. After the study has been completed, others can check whether these raw data support the study results  – Negative case analysis: examining exceptional and anomalous situations (cases), and describing in what way they also fall under the theory (or how they indicate the limit of the theory’s validity)

11.6.2 Applicability or Transferability Applicability concerns the question whether the context of the research adequately matches the context in which the results will be applied. This is something the researcher can never judge. Within qualitative research, small numbers of study participants are used, who do not participate in the study by means of a random sample, but by means of a purposive sample. Due to these small numbers, and the composition of the sample, there is no question of a representative sample. It is the responsibility of the nurse who wishes to apply the research results to their own practice to ensure that their own situation is sufficiently comparable to that of the study. The researcher is responsible for making the data available so that a healthcare professional wishing to apply the study results can verify the degree of similarity between the two situations. To this end, the researcher makes a thick description of the research situation available. This description includes a detailed description of the study situation and the study participants [4]. It is obviously not possible to include the thick description in the research article. However, a table with the characteristics of the study participants must be included in the article. A description of the study situation must also be included in the article, on the basis of which the healthcare professional can make an initial assessment

11.6  What Are Quality Criteria for Qualitative Research?

199

as to whether the match with their own practice is sufficient. The detailed description can then be requested from the researcher. If the own practice is sufficiently similar to the research situation, the results are ‘transferable’.

11.6.3 Repeatability If the same object is measured twice in quick succession, and it has not changed in the meantime, the measurements should give the same measurement result. Such a repeated measurement shows whether the measuring instrument is stable. In qualitative research, observations and interviews are not really repeatable. Repeating research is impossible because people change over time [4]. Only if film or audio recordings have been made, is repetition of a measurement possible. In that case, at least part of the recordings should be double-coded, by two people, independently of each other. Furthermore, it is possible to check whether the research process was conducted correctly and whether the study results are indeed based on the observations: is the reasoning correct? Can the reasoning and the results be justified on the basis of the data collected? If so, then there is dependability: the results indeed depend on the data. This is how repeatability is elaborated in qualitative research. An independent researcher can carry out an inquiry audit to check to what extent the research process has been carried out correctly. However, this is an intensive procedure which is not easy to organise and carry out. By working in a research group (with a number of researchers who work together on the same research), the control of the research process is easier to organise. The researchers exchange data, compare their data and discuss the data and the choices to be made and subsequent steps in the research. In this way they check each other.

11.6.4 Neutrality or Confirmability Rather than emphasising the objectivity of the researcher, neutrality is about whether the data is accurate: what is the quality of the data and results? If data or results are distorted by bias or by the values of the researcher, they are not neutral. The question is then whether the data and results can be confirmed and thus whether they are plausible (confirmability). There are two ways in which neutrality can be checked [4]. The first is to carry out an inquiry audit, as with repeatability. Whereas in the previous point, this focused on the research process, the check should now focus on the research product (the results). The second way is via the reflexive diary that the qualitative researcher has to keep. In this diary, the researcher reports his/her choices, the grounds for these choices, his/her own behaviour and feelings, so that it is possible to retrieve the basis on which decisions were taken and to check the accuracy of these decisions.

200

11  How to Better Understand Patients’ Experience: Understanding Psychosocial…

11.7 How Does Qualitative Research Contribute to Clinical Reasoning? For clinical reasoning, qualitative research usually provides more indirect information. As the nurse gains more insight into the patient’s values, needs, goals and experiences and more insight into processes of care delivery, they better understand the patient and his situation. Because the patient perspective is an important perspective in clinical reasoning and in discussing decisions with the patient, qualitative research can make an important contribution to clinical reasoning.

Glossary Bias (in scientific health research)  In scientific health research, bias includes any factor or process that tends to deviate the results or conclusions of a trial systematically away from the truth ([5], p. 28). Credibility  Term from qualitative research. The credibility of qualitative research is the extent to which it can be believed that the descriptions and statements in that research are valid for the persons and situations studied in that research. A credible study correctly represents the reality studied and therefore has ‘truth value’. And with truth value, the extent to which the descriptions and explanations are free of bias is indicated. Generalisability (external validity)  The extent to which the results found in the sample of a study are also valid for individuals with similar health problems and characteristics, who did not participate in the study (e.g. a question might be: are these study results valid for individuals with similar health problems but who live in a different continent or who have a different culture?). Inquiry audit  Method in qualitative research, whereby the extent to which the research process has been carried out correctly and the research product (the study results and conclusions) are indeed based on the research data, is verified. Internal validity  Degree to which the research is free from bias, and the study results are therefore free from systematic deviation away from the truth. Meaning  The value or interpretation attributed to the research subject by the study participants. Member checking Method in qualitative research to increase the credibility/ truth value of the study results. Elaborated research material (e.g. an elaborated ­interview) is read and checked by the original source, or study participants are asked if they recognise the descriptions in elaborated notes. Mental constructs (in qualitative research)  The output of thought processes that people have stored in their memory as a result of conscious or unconscious learning processes. These mental constructs can take the form, for example, of views, opinions, perspectives, beliefs, meanings. Negative case analysis Method in qualitative research to refine the working hypotheses or to establish the limits of the theory to be developed. In this way,

11.7  How Does Qualitative Research Contribute to Clinical Reasoning?

201

the credibility or truth value of the study results is increased. Exceptional and deviant situations or cases which do not fit the pattern, and to which the hypothesis or theory does not seem to apply (well), are searched for purposefully and further studied. Next, the researcher describes in which way these cases also fall under the theory or hypothesis (refining the working hypothesis) or in which way these negative cases indicate the limit of the theory’s validity. Participant observation  Method of observing study participants and the research setting in which the researcher participates as much as possible in the day-to-day activities of the study participants. Peer debriefing  During the course of the research, the research process is discussed with peers. This can include questions such as: are the values and opinions of the researcher likely to have too much influence on his decisions and choices? Persistent observation  Method within qualitative research to prevent the collection of data being stopped too quickly, namely when it seems that saturation has been reached. According to the persistent observation method, the researcher should not stop collecting information then, but should try to deepen the obtained information. Prolonged engagement  Method within qualitative research to increase the credibility or truth value of the study results. By being involved with the study participants and/or in the study situation for a long period of time, it is more likely that the researcher will get to know the reality well. Prolonged engagement prevents conclusions being drawn too quickly which, on closer inspection, may not be correct. Purposive sample  A sample which deliberately selects those research units which represent certain characteristics. Qualitative research  Research in which the understanding of (psychosocial) processes, experiences and meaning of the study participants is central. The course of the research is not predetermined but is partly determined by the findings during the research. Qualitative research takes place in the natural environment of the research subject. From the start of the study, the researcher seeks direct contact with daily reality. The researcher works towards hypotheses or theories that transcend the situation studied. Quantitative research Research that is carried out systematically via a fixed research plan into observable phenomena, which are measured in a valid and reliable manner. The measurement data are then analysed using statistical techniques. The answer to the research question is given in the form of numbers and in measures (such as the relative risk or the mean difference). Referential adequacy  Method in qualitative research to give others the opportunity to check the validity of the results of the study. To this end, part of the collected data is kept aside and stored as ‘raw’ data. At the end of the study, others can still analyse these raw data and check whether they support the study results. Reflexive diary  Method used in qualitative research to justify and check the neutrality of the data (absence of bias) and the research process. In this reflexive diary, which the qualitative researcher must keep, he or she reports his or her choices, the reasons for these choices and his or her own behaviour and feel-

202

11  How to Better Understand Patients’ Experience: Understanding Psychosocial…

ings, so that the grounds on which decisions were made can be traced and their accuracy verified. Reliability  The degree to which repeated measurements of a stable phenomenon by different people and instruments at different times and places get similar results [6]. Saturation  Term from qualitative research, meaning that no new relevant information is obtained anymore through the data collection method used by the researcher, and thus data collection can be stopped. Thick description Method used in qualitative research to enable users of the research to ascertain whether the research results can be transferred to their own practice. The researcher provides a comprehensive and detailed description (thick description) of the study situation. This description includes extensive information on the study setting and on the study participants. Through this description, the healthcare professional can assess whether her own practice sufficiently matches the research situation, in order to determine whether she can apply the research results. Triangulation  Method to increase the quality of qualitative research. The essence of triangulation is that different approaches are used. Information can be collected through different methods (observation, interviews, file research). Verstehen  Concept from qualitative research. Being able to see the perspective of the study participants.

References 1. Van Nes F.  Everyday activities of ageing couples: changes in the face of declining health. Academic dissertation VUmc. Amsterdam: VU and HvA; 2013. 2. The A. M. ‘Vanavond om 8 uur’. Verpleegkundige dilemma’s bij euthanasie en andere beslissingen rond het levenseinde. Houten: Bohn Stafleu van Loghum; 1997. 3. Van der Kluit MJ, Dijkstra GJ, De Rooij SE.  Goals of older hospitalised patients: a qualitative descriptive study. Br Med J Open. 2019;9:e029993. https://doi.org/10.1136/ bmjopen-­2019-­029993. 4. Lincoln YS, Guba EG.  Establishing trustworthiness. In: Lincoln YS, Guba EG, editors. Naturalistic inquiry. Newbury Park: Sage; 1985. 5. Jadad A. Randomise controlled trials. London: BMJ Books; 2007. 6. Fletcher RH, Fletcher SW, Fletcher GS.  Clinical epidemiology. The essentials. 5th ed. Philadelphia: Lippincott Williams & Wilkins; 2014.

Part III Second Layer with Extras

In-Depth Texts

12

12.1 Examples of the Four Types of Questions Diagnostic questions Etiological questions

Prognostic questions

Therapeutic questions

How do I know this baby is in pain? Is the patient lonely, unable to form or maintain social relationships, unable to regulate his behaviour in company, what is going on? What causes the patient to suffer so much anxiety? Is this patient’s obesity related to energy intake, energy consumption, or both? If it has to do with energy intake, is it about hunger pangs or more about psychosocial factors, such as stress and self-image? For a patient at this age, is it realistic to expect them to learn to manage stress differently? This patient will eventually be able to walk a 100 m, but what use is this when he is home alone again? For example, are the shops within a 100 m walking distance? What is the best treatment for this wound? How can patients regain control over their weight, weight loss and healthy eating?

12.2 Making a Deliberate Decision To make a deliberate decision, the nurse should purposefully seek out information. She carefully considers the extent to which she can trust this information and its value and relevance to answering her questions. The nurse should also consider whether she has enough information or if more is required to reach a decision. In doing so, the nurse is constantly shifting back and forth between clinical information (What symptoms does the patient have? What symptoms are present? What does the patient’s behaviour mean?) and on the other hand, professional knowledge.

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 J. Dobber et al., Clinical Reasoning and Evidence-Based Practice, https://doi.org/10.1007/978-3-031-27069-7_12

205

206

12  In-Depth Texts

Understanding is also important in clinical reasoning. Here, the decision is strongly guided by the nurse’s understanding of the patient’s perspective and experience of the situation. Sometimes ‘ethical clinical reasoning’ is also used. In this case, decisions are based primarily on ethical standards.

12.3 Conditions for Critical Thinking by the Nurse In clinical reasoning, the nurse runs through a set of questions. The nurse’s thinking always involves the following three conditions.

12.3.1 Use Good Information The information the nurse uses should be reliable and valid. Reliability is concerned with whether data measured in the same way by different nurses produces the same results. Or if the same nurse performs the same measurement twice, the result is the same (unless what is measured changes in value between the two measurements). A simple example is the weight of a patient. If a nurse weighs a patient at 10.00 a.m. and the same or a different nurse does it again at 10.02 a.m., both measurements should give the same weight. If the second weighing takes place in the evening at 10.00  p.m., there may of course be a small difference between the measurements. In addition, the more accurately a measurement indicates the true value, the more reliable it is. For example, a precisely measured body temperature of 100.22 °F is more accurate and therefore more reliable than rounding to 100 °F. Validity is often described as ‘measuring what you want to measure’. If a nurse wants to know the height of a patient and asks for his shoe size, this is clearly not a valid measurement. If a nurse wants to monitor a patient’s weight over time to detect weight changes in time and weighs the patient sometimes in the morning at 10 a.m. and sometimes in the evening at 10 p.m. (method 1), this is also not a valid measurement. On the other hand, if she weighs the patient every morning at 07.00, before breakfast and before going to the toilet (method 2), it is a valid measurement. Nevertheless, both the measurements of method 1 and method 2 have been performed correctly. After all, they indicate exactly the right weight of the patient at the moment of measurement. The point is that with method 1 the weight differences between the measurements cannot be interpreted properly for the purpose for which

12.4  Examples of Standard Questions That Are Asked in Order to Eventually Answer…

207

the weight was measured. Method 2 can do this, and therefore method 2 leads to valid data and method 1 does not.

12.3.2 Assess and Weigh Information Properly It is not enough to collect information to make a good decision. The value of the information has to be assessed. There are two important questions here: (a) How sure is the nurse that the information is completely accurate? The nurse must assess the quality of the information. (b) How important is this information? The nurse has a variety of information available through observation, professional knowledge, literature, statements made by the patient, and information from the patient file. This information should be combined to provide the most complete picture possible, even though some of it may contradict other information or appear to do so. The nurse should therefore weigh the information: how important is each piece of information?

12.3.3 Thorough Professional Knowledge Only when the nurse has a thorough, up-to-date professional knowledge, can she assess the reliability, validity, quality and importance of information and interpret it correctly. Clinical reasoning requires a good, deep understanding of human anatomy and physiology, of pathophysiological processes (processes which are physiologically abnormal enough to cause health problems) and of pathology (knowledge of diseases). A good and thorough understanding is also required of psychological and psychosocial processes, of psychopathological processes (abnormal psychological and psychosocial processes that lead to health problems) and of psychopathology (knowledge of psychological disorders).

12.4 Examples of Standard Questions That Are Asked in Order to Eventually Answer the Basic Question Diagnostic questions

Etiological questions Prognostic questions Therapeutic questions

 – What objective and subjective symptoms (complaints, signs, behaviour, abnormal normal values) are present in the patient?  – How can these symptoms be explained? Think of deviations in anatomy, physiology or psychosocial functioning  – What are possible causes and related factors from diseases and disease processes?  – Are these factors indeed present in the patient?  – Which external factors influence the course of the health problem negatively? Which ones are positive?  – What is the maximum achievable result for this health problem?  – What interventions are available?  – What effects (with respect to the goal) can be expected from these interventions? How large is the effect expected to be?

208

12  In-Depth Texts

12.5 Example of Body Structures and Their Body Functions (ICF) Esther (19) feels depressed all the time and says that she has lost all joy in her life. She stays in bed for a long time, has no energy and does not want to get up, does not attend classes anymore and hardly does any activities with her friends. Esther lives with her mother, who makes sure she gets up every day and has something to eat. After a week and a half of waiting, Esther’s mother makes an appointment for Esther with the family doctor. The next day Esther sits with the mental health nurse specialist, who co-works with the family doctor. The neurotransmitter serotonin plays an important role in the mood function. When the serotonergic system is disturbed the functions mood, motivation and energy level are affected. The limitation in social activities and the limitation in participation are also a result of the disruption of the serotonergic system and the disruption of mood, energy level and motivation. Because the hypothalamic-pituitary-adrenal axis is too active, stress regulation is also abnormal. Stress hormones, including cortisol, are released through this axis. This also happens in situations where it is not necessary.

12.6 Deepening Knowledge of Environmental Factors When reasoning about environmental factors, the patient’s perspective on the environmental factors is the starting point. If the environmental factors contribute to the patient’s functioning, the environmental factor is a supporting factor. It is also possible that an environmental factor interferes with the patient’s functioning. In that case, the factor is an impeding factor. In assessing the influence of the factor and reasoning about it, the nurse considers the direction of the factor (supportive or impeding), the quality of the factor (good, moderate or poor), its availability (always, sometimes, randomly, or never) and its relevance (the factor is constantly influencing the patient’s functioning, often, usually, sometimes or rarely). Supportive factors can, for example, prevent or reduce participation restrictions for people with a disorder or disability and, in this way, promote the functioning of that person. Conversely, impeding factors limit the functioning of individuals, thereby increasing their health problems. In the ICF, the following types of environmental factors are distinguished [1]: –– Products and technology, such as products and substances for personal consumption (food, but also medicines), technical devices, communication technology, technology for education and employment, technical aspects that determine the accessibility of buildings, roads and signposting, lighting, financial assets (such as money). –– Natural environment and human-made changes to environment, such as physical geography (e.g., hills, rivers), population: demographic change and density, climate, natural events (e.g., storms and floods), environmental pollution, light

12.7  Explanation of the Standard Questions for a Diagnostic Decision

209

intensity and light quality, natural day and night cycles, sound intensity and sound quality, air quality. –– Support and relationships, such as family, friends, acquaintances, colleagues, caregivers, strangers, pets, health professionals. –– Attitudes, such as attitudes of family members, friends (and so on, see above under Support and relationships), social attitudes (e.g.. opinions and beliefs held by people in a culture or society), social norms and religious norms. –– Services, systems and policies, such as facilities, systems and policies to produce consumer goods, facilities, systems and policies for housing, for communications, transportation, social security, healthcare, education, employment.

12.7 Explanation of the Standard Questions for a Diagnostic Decision Standard question 1. What objective and subjective symptoms are present in the patient (complaints, signs, behaviours, abnormal value compared to the reference range)? 2. How can these symptoms be explained (which abnormalities in anatomy, physiology or psychosocial functioning)?

3. What else would you like to know from or about the patient? 4. What is the patient’s hindrance (possibly expressed in the ICF-components activities, participation)? 5. How does the patient respond to this (emotional behaviour, [in] effective self-management)?

6. Which possible health problems does this indicate (from which impairments in body structures, in body functions, and limitations in activities, and restrictions in participation)? Please indicate for each problem how important this problem is

Note For nurses, a patient’s symptoms are usually the first information about their health condition that they see, hear or read. Therefore, understanding what is going on starts with looking and listening carefully to the patient, and reading carefully the information that is already known Symptoms are manifestations of abnormalities in body structures, or in physiological or psychosocial processes. To understand what is going on, the nurse needs to know how these structures or processes normally work and what abnormality in them explains the symptom. For example, what processes cause someone with a fever to have flushed cheeks? This knowledge is also needed to later assess the usefulness of interventions Data may still be missing, for example, on subjective symptoms such as pain This makes it clear how deviations in body structures and body functions affect the person’s functioning. The results of this can be used in the reasoning later on as a Problem or as Signs/symptoms in the nursing diagnosis The patient’s response may recur in the reasoning later as a Problem in the nursing diagnosis It may also turn out to be a positive or negative factor influencing the prognosis (see Standard Question 1 on prognostic reasoning) Through this problem list, thoughts are ordered: what can be problems from impairments of body structures and the physiological or psychosocial processes? What from the body functions? And so on What problems belong together (such as the impairment of drift control, and the establishment and maintenance of friendships)? What problems are most important right now?

12  In-Depth Texts

210 Standard question 7. What are the defining characteristics of the major health problems?

8. How can the presence or absence of these defining characteristics be tested?

Note By looking, listening and reading carefully, potential problems emerge. It is then necessary to check whether these problems actually exist. First, it must be clear which characteristics / symptoms show the presence of that specific problem. For example, there is only a fever if the body temperature is 100.4 °F or higher When the defining characteristics of the health problems have been determined, it should be checked whether these defining characteristics are indeed present in the patient

After performing this assessment, the nurse can confirm or reject the potential health problems from the problem list, through the presence or absence of the defining characteristics

12.8 Explanation of the Standard Questions for an Etiological Decision Standard question 1. Are there possible causes and related factors from diseases and disease processes? 2. Are there possible causes and related factors from the (para) medical treatment? 3. Are there possible causes and related factors from body structures, or physiological or psychosocial processes? 4. Are there possible causes and related factors from body functions? 5. Are there possible causes and related factors from activities? 6. Are there possible causes and related factors from participation?

Note In many cases, diseases and disease processes lead to other health problems. For example, pneumonia leads to fever. For many health problems, the disease or the disease process is the main cause Disease treatments can cause other health problems. An obvious example is the side effects that some medications cause

From each ICF domain, causes and related factors of health problems may arise. By checking the ICF domains one by one for possible causes and related factors, a fairly complete picture emerges of the causes and related factors of a particular health problem. For example, body structures of the ear: impairment of the structure of the middle and inner ear Examples of possible causes or related factors from ICF domains for impaired self-management: –Mental functions: impairment in judgment Communicative activities: limitation in starting and sustaining a conversation Participation in interpersonal relationships: restriction in engaging in temporary contacts with strangers, with formal relationships and with informal social relationships

12.9  Explanation of the Standard Questions for a Prognostic Decision Standard question 7. Are there possible causes and related factors from environmental factors? 8. Are there possible causes and related factors from personal factors? 9. Can these factors together cause and/or maintain this health problem?

10. Are these factors present in the patient?

11. Do these factors have this effect in this patient?

211

Note Environmental factors: Social security provisions: Being uninsured/not having health insurance.

Personal factors: feminist view of life

Once all the causes and related factors have been identified, it must be determined which of them play a truly important role and which exert some influence, but are not of vital importance The question then arises as to whether the combined key causes/ factors can indeed cause and/or maintain this health problem? The nurse should test whether it is plausible that these combined factors cause and/or maintain the problem. Or whether there is, for example, a missing factor that is always present in case of this health problem The factors that are marked as ‘important’ are checked again to see if they are indeed present in the patient If an important causal factor of the health problem is not present, a targeted investigation is made to see if this factor might be present after all, but was not noticed before. Keep in mind the risk of not looking objectively anymore. If the nurse has already established the diagnosis in her mind, she may take vague indications of the presence of a missing link for granted The final test is whether the factors that are present actually work in this patient in such a way that they lead to the health problem. There can be big differences between people. While one person experiences pain at the slightest, the other almost never suffers

After this review has been conducted, the etiological factors of the health problem can be determined

12.9 Explanation of the Standard Questions for a Prognostic Decision Standard question 1. Which physical, psychological and personal factors (including causal and related factors) negatively influence the course of the health problem? Which ones are positive?

Note Various influences play a role in estimating how the patient’s health problem is likely to progress. Depending on the problem, physical factors (e.g., physique), psychological factors (e.g., susceptibility to stress), and personal factors (e.g., age) can all play a major role in how the health problem may progress A realistic assessment of the optimal achievable result requires that both the health-promoting and the impeding factors are identified

212

12  In-Depth Texts

Standard question 2. Which environmental factors influence the course of the health problem negatively? Which ones are positive? 3. On what mechanism of action (per factor) is that influence based? 4. Do the factors add up to positive, neutral, or negative?

5. What is the maximum achievable result for this health problem?

Note Environmental factors can have a major influence. For example, it can make a big difference whether the patient is well supported by family and relatives, but also which health facilities and interventions are available to the patient or not Through this question, the nurse reasons about (1) the magnitude of the factor’s influence and (2) the extent to which the factor itself can be influenced Ultimately, a weighing of the factors is necessary to reach a final conclusion. The aim is not to determine whether there are more positive or more negative factors. When using prognostic reasoning, the nurse attempts to reason about how the various factors will affect each other and the health problem From the answer to question 4, the nurse formulates the maximum achievable result. This will not always be ‘solving the health problem’. Sometimes the best achievable result is an improvement, sometimes preventing further deterioration and sometimes only slowing down the deterioration

Once the nurse has reasoned out the best possible outcome, a concrete goal for this patient health problem can then be formulated

12.10 Explanation of the Standard Questions for a Therapeutic Decision Standard question 1. What interventions are available?

2. What effects (in terms of the goals) are expected from this intervention? How large is the expected effect?

Note To be able to reason about interventions, first of all an overview of the available interventions is needed. In standard situations, often one and the same intervention will be chosen. Sometimes the intervention is listed in the protocol of the work unit. If the decision has to be made in a non-standard situation, a broader search for different possible interventions should be conducted. Interventions can sometimes be found in guidelines, in the evidence-based professional literature (see Part II of this book), in background literature (such as textbooks) and in the Nursing Interventions Classification (NIC) [2]a Interventions are only useful if they work, i.e., if they contribute to achieving the goals set. It is important to know how much the intervention contributes to the goal: how effective is the intervention? To know if and how effective an intervention is, the nurse can look up the intervention in the evidence-based literature (see Part II of this book) The answer to this question provides an overview of the effectiveness of the available interventions

12.10  Explanation of the Standard Questions for a Therapeutic Decision Standard question 3. What evidence is there that this intervention does indeed have that effect? 4. What is the point of intervention (causes, related factors, symptomatology, prognostic factors, or risk factors)?

5. What is the mechanism of change of the intervention?

213

Note The nurse should answer question 3 in conjunction with question 2. This is because the answer to question 3 comes from the evidence-­ based literature. In answering this question, the nurse should consider the strength of evidence for each intervention; in other words, how confident are we that the intervention will be effective for the patient in whom this intervention will be performed? With questions 4 and 5, the nurse maps out how the intervention is actually doing its job With question 4, the nurse identifies the aspect of the health problem through which the intervention influences the health problem  – If the intervention affects the etiology, then one or more causes or related factors of the problem are reduced or eliminated  – If the intervention influences the symptomatology, one or more symptoms are reduced or eliminated. In other words, there is no influence on the cause of the problem. That is not necessarily a bad thing: If someone has a headache and is given a paracetamol, in many cases that will be a useful and effective intervention  – If the intervention influences the prognostic or the risk factors, then one or more of the negative prognostic factors or of the risk factors are eliminated or reduced, or one of the positive prognostic factors is strengthened The nurse will assess whether and how the intervention brings about changes in anatomical and/or physiological structures or processes or in psychosocial functioning. In other words: how does the intervention work? The answer to this question will help the nurse to assess whether the intervention is appropriate for the patient Example: lifestyle advice and lifestyle interventions (smoking, exercise, nutrition, alcohol) in heart patients are aimed at prognostic factors. Stopping smoking eliminates a negative prognostic factor and reduces the risk of cardiovascular complications by about half The nicotine in tobacco smoke temporarily constricts the arteries and damages the artery wall, allowing cholesterol to adhere to the wall. Therefore, both these pathophysiological processes cause a small passage for the blood in the arteries. In addition, smoking increases the LDL cholesterol level in the blood and thickens the blood. As a result, the blood flows less smoothly and blood clots form more quickly. The blood clots can get stuck in the narrowed part of the artery and close it off, after which a heart attack or stroke can occur. Therefore, the mechanism of change of smoking cessation is putting an end on (the progress of) these pathophysiological processes

214 Standard question 6.Is the intervention feasible (resources, practicability)?

7.Is the intervention acceptable to the patient?

12  In-Depth Texts Note Before answering this question, the nurse has made a selection of potentially appropriate interventions based on the responses to the previous questions. For those interventions that remain, question 6 is answered. With this question, the nurse considers whether the intervention can be implemented by the nurses in the professional unit. The nurse goes through the following questions:  – Are there any tools, equipment, instruments or other materials needed for this intervention? If so, are these indeed available within the work unit?  – How long does it take to implement the intervention?  – Is there any particular knowledge and/or skill required to deliver the intervention? If so, is this knowledge and skill present in sufficient numbers of nurses to ensure delivery of the intervention on a daily basis?  – From the answers to these three questions, the nurse should ask the fourth question: Is the outcome of the intervention in balance with the cost of the intervention? If the costs are high (time consuming, expensive materials) the nurse should consider whether the intervention is justified Ultimately, the nurse selects a limited number of effective and efficient interventions to present to the patient as options. The nurse makes sure that the patient understands the different options. The patient can then tell to what extent each of the options is acceptable to him. In addition, he can indicate whether any of the options is his preference If the patient is unable to answer this question for himself, the nurse may wish to consider discussing the acceptability of the interventions with the patient’s next of kin If no next of kin are available to discuss this with, or if due to an acute situation there is no time to discuss this with the patient or their next of kin, the nurse herself should assess the acceptability of the intervention to the patient

With the answers to the above questions, the nurse can reason on the most appropriate intervention for this patient, with this health problem and in this situation a About the NIC: the NIC is an American initiative to bundle nursing interventions. The classification with interventions is published in book form and on the web. A major disadvantage of the NIC is that the interventions described are consensus-­based and not evidence-based

215

12.11  Forms of Collaboration Between Professionals: Multiprofessional...

12.11 Forms of Collaboration Between Professionals: Multiprofessional, Multidisciplinary, Interdisciplinary or Interprofessional Mutual relationship professionals

Allocation of responsibilities and tasks

Communication between professionals

Multi-­ professional

Hierarchy between professionals, no teamwork

Separate roles and tasks, working separately with the same patient

Multidisciplinary

Hierarchy between disciplines, but teamwork

Separate roles and tasks

No fixed consultation, consultation depends on taking initiative by one of the professionals Scheduled consultations to coordinate care

Interdisciplinary

Limited hierarchy, but teamwork

Separate roles and tasks

Interprofessional

No hierarchy, knowledge and recognition of each other’s expertise

Shared responsibility for common goals and approach

Scheduled consultations to jointly formulate goals and tasks; everyone thinks and talks about the entire care provision, regardless of the boundaries of disciplines Scheduled consultations in which everyone’s ideas and proposals are discussed. Joint goals and approach. Regular evaluation and possible adjustment of goals and plans

Overview of complex patient situation Each professional has separate goals and plans

Patient satisfaction and quality of care Depending on how individual professionals work

There may be common goals, professionals are broadly aware of each other’s actions Shared goals, professionals are well aware of each other’s actions

Depending on the way individual disciplines work

Shared goals, knowledge of each other’s activities. Cohesive and integrated care

The patient’s needs and goals are the starting point. Frequent consultation and coordination with patient and informal carers. Focused on patient experience of consistent and good quality care

Through coordination, an attempt is made to provide cohesive (and therefore high quality) care

Sources [3, 4]: Tsakitzidis G, Van Royen P. Learning interprofessional collaboration in health care. Antwerp: Standard Publishing Professional; 2014; https://interprofessional.global Accessed 1 Jun 2020. Other sources may highlight other characteristics

216

12  In-Depth Texts

12.12 Further Elaboration of the Nursing Professional Domain The International Council of Nurses (ICN) describes the nursing professional domain as follows: Nursing, as an integral part of the healthcare system, encompasses the promotion of health, prevention of illness and care of physically ill, mentally ill and disabled people of all ages, in all healthcare and other community settings. Within this broad spectrum of healthcare, the phenomena of particular concern to nurses are individual, family and group responses to actual or potential health problems. These human responses range broadly from health restoring reactions to an individual episode of illness to the development of policy in promoting the long-term health of a population [5]. Nurses make nursing diagnoses in the following areas: –– responses 1 to, and consequences of, illness and disability, –– burden of (para)medical treatment, –– responses to, and consequences of, (para)medical treatment. These responses can be physical (side effects), psychological (motivation) and emotional; they also include the way in which the patient deals with the treatment, for example, the degree of compliance, –– prevention of and response to health risks, –– self-management. Paramedics and physicians focus on: –– –– –– –– ––

illness and disability, complications of illness and disability, (para)medical treatment complications of (para)medical treatment, prevention of health risks.

In these areas, the nurse does gather information, but she relays this information to the discipline who makes decisions about it. There is overlap between the professional domains of prevention and adverse effects. Sometimes these clearly fall under the (para)medical responsibility, sometimes more under the nursing responsibility. The latter is the case, for example, when side effects are considered to be ‘a burden for the patient, but unavoidable’, as in the case of nausea caused by medicinal treatment. It may also be the case that prevention and adverse effects are regarded as multidisciplinary problems, whereby all disciplines have a share in the decision-making and in the responsibility for the problem.

 ‘Responses’ refers to the way in which the patient reacts to illness and disability. These can be physical reactions (fever, nausea), psychological reactions (fear, mourning) and emotional reactions (anger, sadness). 1

12.13  Short Case with Autonomous and Participative Action

217

12.13 Short Case with Autonomous and Participative Action Mrs. Sanders knows she won’t live much longer. She is in the last stage of life (the terminal phase), which means that she is expected to die within 3 months. She hopes to die at home, with her husband and her children. That’s why she was discharged from the hospital today and returned home. When community nurse Frank drops by for an intake interview, he finds the family doctor there. That is very convenient. Now they can discuss the pain management policy together. ‘Let me know immediately if it is not sufficient’, the GP says. ‘I’m sure you will. And I’ll keep an eye on it myself’, she adds. ‘Of course’ Frank replies. When the doctor has left, Frank turns all his attention to Mrs. Sanders and her husband. Mrs. Sanders tells him that she is not experiencing much pain at the moment, but she is very tired. Frank discusses with Mrs. and Mr. Sanders how the community nurse and the family members can help Mrs. Sanders to focus on the activities that are important to her. Together they agree on the tasks each will take on, so that Mrs. Sanders can organise the last phase of her life as she wishes, despite her fatigue and pain. After that, Frank addresses the pain management plan. As pain is not the main concern at the moment, the GP has prescribed medication ‘if necessary’. Frank indicates that the GP is responsible for pain management. If the patient feels the need, she can also discuss pain management with the community nurses. After all, the community nurse visits her every day and is therefore an easy person to talk to. Frank goes on to say that if the client wishes, he can also consult with the GP about pain medication, but that the GP is always the one who decides on pain policy. ‘We’ll see how it goes’ is their shared conclusion.

12.13.1  Autonomous Reasoning and Acting The nurse has primary responsibility for the health problem of ‘fatigue’. He or she takes decisions on this in consultation with the patient and her partner. After all, in the case of Mrs. Sanders, fatigue is a consequence of her illness and treatment.

12.13.2 Participatory Reasoning and Action The general practitioner has primary responsibility for the drug treatment of the health problem ‘pain’. After all, this concerns medical treatment. The nurses see the patient every day and monitor her experience of pain. Because of this, and because of their professional knowledge about pain management, in many cases the nurses are the first to notice possible problems with pain, and with whom the patient

218

12  In-Depth Texts

discusses her perception and experience. These activities, and those that result from them, fall under the scope of participatory nursing reasoning and action.

12.14 Explanation of the Seven Diversity Aspects The diversity aspects below are an adapted version of the aspects formulated by the Caleidoscopia Network [6]. 1. Ethnicity, nationality, region Ethnicity: identity based on a group culture of the population group from which the person originates. Nationality: belonging to a specific country based on origin or birth. Region: area with a common language/dialect and/or culture. 2. Life stage and generation Region: area with a common language/dialect and/or culture. Generation: the group of people born in the same period, who therefore share a common zeitgeist, history and common experiences. 3. Religion and philosophy of life Religion: a spirituality based on faith in God or gods, with its own stories, rules of conduct and rituals. Philosophy of life: believes and perspectives on the meaning of life, what life is about, and how to live. 4. Sex and gender Sex: the biologically defined category (woman, man or intersex). Gender: the psychological and cultural expression how one identifies oneself. For example, gender identity: whether the person feels male, female or non-­ binary; gender expression: how one expresses gender differences, e.g., in hairstyle and clothing; gender role: male–female roles assigned by society and power relations between men and women. 5. Sexual orientation Sexual orientation such as heterosexuality, homosexuality, bisexuality and so on. This also includes the social position based on sexual orientation. 6. Social class, socio-economic status (SES) Social class: classification of groups of people according to their social status, e.g., upper, middle and lower class). Socio-economic status (SES): position of people in society based on their income, type of work, social network and the status of people in that social network. 7. Talents, limitations, personality Talent: well-developed ability in the physical, cognitive or social area. Limitations: difficulties a person has in performing activities, for example, not being able to understand written texts properly due to low literacy.

12.15  Example of Blood Pressure Measurement by Isabella: First Part

219

Personality: the individual’s ability to react in a particular way to situations, including the mental characteristics that makes the individual distinct from ­others, for example, the degree of introversion or extraversion, conscientiousness, psychological stability, optimism, confidence [1].

12.15 Example of Blood Pressure Measurement by Isabella: First Part Isabella has passed the skills assessment and may now go and do her first work placement. She is looking forward to this. As far as she is concerned, it is high time for the real work. Books in the bookcase and let us get to work. Isabella excitedly starts her first internship in a hospital. Until now Isabella has mainly been observing, but today her supervisor asks her if she wants to start doing the routine checks of the vital signs. Confident in what she practiced at school, Isabella enters the first patient room: ‘ Good morning, did you sleep well? Did you enjoy your breakfast? I would like to take your blood pressure. Please come and sit down’. The blood pressure measurement goes well, she observes. Afterwards her work supervisor asks: ‘And how did the checks go? Did you notice any abnormal blood pressure values?’ Isabella says that none of the patients’ blood pressure has changed much compared to the previous evening. ‘Many blood pressure values are a bit high’ she says, ‘higher than 120/80 mmHg’. She remembers well that in class, this value was mentioned as a normal value. Her supervisor tells her, while looking at the checklist, that you can’t just say for every patient that 120/80 mmHg is normal and in line with expectations. Isabella is not sure what to make of this remark. How is 120/80 mmHg not normal just like that? What blood pressure value can then be expected when you measure the blood pressure? What does it depend on? Isabella is still happy with her checks of the blood pressure, but she is now starting to doubt whether she knows enough about measuring blood pressure. What do checks such as blood pressure, temperature and pulse actually tell us about how someone is doing? How valuable is a blood pressure measurement when a lying patient is asked to sit down? Does it matter if he has just eaten? Whether the blood pressure is measured on the left or the right? Whether it is morning or evening? And suppose that when you measure again you see a difference in the blood pressure value of 5 mmHg? Is that difference relevant or negligible? How precise should a blood pressure measurement actually be? Isabella has also noticed that many patients on the ward take medication to lower their blood pressure. What effect can you expect from this? What side effects should you be alert for?

220

12  In-Depth Texts

Isabella is fed up. How is she going to get answers to all these questions? Actually, I don’t even know how a blood pressure is established, she concludes discouraged. Isabella sighs. She thinks she has forgotten a lot of her lessons and decides to get her physiology book out of the bookcase again. In her physiology book, Isabella reads about how blood pressure is normally kept within certain limits by various control mechanisms. She reads how blood pressure is recorded at various locations in the body and is then regulated by complex feedback systems. She reads how blood pressure can be quickly adapted to the circumstances under the influence of the nervous system and how later the kidneys also come into play. She reads about the sympathetic and parasympathetic nervous systems, about renin, angiotensin, aldosterone, natriuretic peptide and about how nervous and hormonal blood pressure regulation influence each other. Isabella now has a better picture of the normal regulation of blood pressure and realises how sensitive it is. Changes in posture, eating and stress are just a few of the factors that influence the blood pressure within a few seconds, but which she did not take into account during her measurements. How bad is that? Isabella wonders. She quickly realises that it depends on the reason why you want to know how high your blood pressure is. Are you measuring the blood pressure because you want to detect internal bleeding via a drop in blood pressure, or because you want to decide whether blood pressure lowering therapy is necessary, or because you want to assess how well this therapy is working? If you don’t know why you are taking a blood pressure, Isabella thinks, then you also don’t know what blood pressure to expect and how accurate your blood pressure measurement should be. Such a mindless blood pressure measurement is pointless, Isabella observes. At the same time, she realises that for many patients whose blood pressure she has measured that day, she cannot think of a reason why their blood pressure should be routinely be measured every morning and every evening. In two patients, she knows, medication to lower the blood pressure has been started. In their case, at any rate, she should have measured their blood pressure more carefully. Even more if you consider that the effects of the blood pressure lowering medication they are receiving, is often only 5–10 mmHg. The differences in the way in which blood pressure are measured between her and her colleagues probably cause the blood pressure value to fluctuate more than the effect sought by blood pressure lowering medicines. But what is a careful measurement? She recalls the terms validity and reliability from evidence-based practice (EBP) classes. A difficult, very theoretical subject, she thinks, but EBP can be so practical. She also takes the clinimetrics book out of the bookcase again. After all, she wants to be able to measure blood pressure better tomorrow than she did today.

12.15  Example of Blood Pressure Measurement by Isabella: First Part

221

In her clinimetrics book, Isabella reads about careful measuring. The physician-­epidemiologist Feinstein called the adequate expression of the state of health in measures and numbers clinimetrics. Clinimetrics is concerned with the development of clinical measuring instruments that can map out how the patient is doing, and with the assessment of their measurement qualities (validity, reliability and responsiveness). Thus, clinimetric indices can help in diagnostics, prognostics and assessment of treatment effect. Validity is ‘measuring what you want to measure’. A blood pressure induced by stress or changes in posture is not what I want to measure, Isabella says, it should be an unaffected blood pressure on which you can base medication policy. Reliability, she reads, is another word for precision. That means, she thinks, that if I measure the blood pressure twice, I should find the same value and not that the blood pressure happens to fluctuate. It also means that my colleagues and I will all measure the same blood pressure value in the same patient. This will be difficult, Isabella thinks, because it means that we have to agree on how we will measure the blood pressure. Sitting or lying down, before or after eating, and on which arm? She remembers the departmental protocol for measuring blood pressure that her supervisor gave her. This protocol refers to a guideline for determining blood pressure that recommends measuring on both arms during the first patient contact, and twice during subsequent measurements on the arm with the highest value, with the average forming the final blood pressure value. Only then can one speak of an accurate, or reliable, blood pressure measurement. What a job, thinks Isabella. I hardly ever see it done that way. But how useful are blood pressure measurements if they are not valid and reliable? With this question, Isabella closes her clinimetrics book, because that book answers very clearly with ‘not useful’. The next day, Isabella is caring for Mr. Langerhans, a vital man for his 80 years, but for the second time in 6 months his sugar has become so irregular that he needs to be hospitalized. When she walks in to take Mr. Langerhans’ blood pressure, pulse and temperature, he tells her he’s not feeling well. Isabella sees a pale and slightly perspiring man. She suspects that his blood glucose is too low, but a check does not confirm this. When measuring his blood pressure, she notices that the systolic blood pressure is considerably lower at 115 mmHg than the night before. Another measurement, she thinks. The blood pressure monitor then shows the same value. She does not trust it. Out of the corner of her eye she sees doctor Hales walking by. ‘Mr. Langerhans is not feeling well’, she tells him, ‘his blood glucose is normal, but his blood pressure is 25 mmHg lower than last night.’ Doctor Hales also measures Mr. Langerhans’ blood pressure again, just to be sure. He measures a blood pressure of 135/85  mmHg, but then says that he considers the blood pressure value that Isabella measured to be more valuable and that he suspects postprandial hypotension based on that. He advises Mr. Langerhans to lie down for one to 2 h and asks Isabella to measure his blood pressure again in 2 h.

222

12  In-Depth Texts

Isabella feels honoured and perplexed at the same time. How is it possible that doctor Hales takes her blood pressure measurement more seriously than his own? ‘Maybe because of the white coat effect,’ says a classmate. Isabella doubts whether stress at seeing a doctor can explain the difference of 20 mmHg between her measurement and that of doctor Hales. She has not read anything about the white coat effect in her physiology book. She decides to consult Medline, a digital literature database. The search term ‘white coat hypertension’ brings up 193 articles. In one of them she finds the following Fig. 12.1: It becomes clear to her that the white coat effect can be as high as 30 mmHg. Moreover, it appears that nurses give a smaller white coat effect on blood pressure than doctors. A good reason, therefore, to leave the measuring of blood pressure to nurses, and if necessary, to the patient themselves, in order to obtain a valid and reliable value. And what did doctor Hales say about postprandial hypotension again? During a search for postprandial hypotension, Isabella stumbles upon an article by Vloet [8]. It teaches her that practically all elderly people show a drop in systolic blood pressure after eating a meal. Besides physical activity, food intake is an important factor causing variation in blood pressure in the elderly. After her internship, Isabella concluded that measuring blood pressure properly is very difficult. The sensitivity to measuring errors and the number of factors that influence the interpretation are much greater than she could have imagined. The more I know, the more I doubt about what is right, Isabella thinks in frustration. Her supervisor has told her that doubts are good. ‘Without wonder and uncertainty you don’t learn and you don’t explore’. Without doubts, useless or even harmful nursing rituals remain. Clinical reasoning, on the other hand, begins with uncertainty. This example is taken from Scholte op Reimer ([9], pp.  4–8). Reprinted with permission. 25

mmHg

20

Systolic diastolic

15 10 5 0

BP measurement BP measurement BP measurement by doctor by doctor after by nurse 10 min waiting

Fig. 12.1  Effect of measuring BP by doctor and nurse on intra-arterial blood pressure [7]

12.16  Blood Pressure Measurement by Isabella: Second Part

223

12.16 Blood Pressure Measurement by Isabella: Second Part 12.16.1 Health Risk: The Whole Is More Than the Sum of the Parts Isabella is doing her second internship in community nursing. On her way to the next client, cycling between fields full of flowers, Isabella thinks back to her visit to Mr. and Mrs. Raas. Mr. Raas is recovering from a stroke. Mrs. Raas has no complaints, but she asks Isabella not only to check her husband’s blood pressure, but also her own. The blood pressure meter indicated 135/90 mmHg. So now what? The fact that high blood pressure is known as a silent killer haunts Isabella’s mind. She knows that high blood pressure, for example, greatly increases the chance of a brain infarct. She also knows that this risk increases as someone has more risk factors for cardiovascular disease, such as a high cholesterol level, a disturbed sugar balance, or smoking. With these risk factors, the whole is more than the sum of the parts. It is therefore important that all risk factors are mapped out, in order to be able to properly estimate the total risk of disease or death. A good risk assessment is necessary in order to determine who needs professional help. Does someone without cardiovascular disease with a blood pressure of 135/90 mmHg need professional attention? This is what Isabella is wondering on her bicycle. For Isabella’s question, epidemiology can help us. Various risk functions are available for estimating the 10-year risk of death from cardiovascular disease. The current guidelines recommend for Europe the Systematic Coronary Risk Evaluation (SCORE) risk function. This is given practical form in a colourful map to estimate the 10-year risk of death from cardiovascular disease for men and women, for smokers and non-smokers, per age category based on systolic blood pressure and total cholesterol. For example, a 66-year-­ old woman like Mrs. Raas with a systolic blood pressure of 135 mmHg without smoking has a 3% to 6% chance, depending on her cholesterol, of dying from cardiovascular disease within 10 years. It strikes Isabella that the guidelines emphasise looking at the mortality risk of patients systematically. That is, using a measurement tool such as the SCORE cards, because it has been shown that a non-systematic, global ‘by eye’ assessment often under- or overestimates the mortality risk. Isabella decides to bring the SCORE card tomorrow to map out Mrs. Raas’ cardiovascular risk profile. But what if it becomes clear that the risk of health damage requires attention, Isabella mulls, what kind of attention would be best for her? Suppose Mrs. Raas’ blood pressure is too high in view of her total cardiovascular risk profile, what advice or therapy would be appropriate? You need to know this before you confront someone with a high risk, according to Isabella. And how do you observe whether the treatment is working? (Fig. 12.2).

224

12  In-Depth Texts

On Wikipedia, she reads that in antiquity, in the Middle Ages and up to the sixteenth century, bloodletting, possibly in combination with various herbal drinks and spices, was often recommended as a therapy. The amount of blood that would flow out after opening a vein could be as much as 2.5 L. Bleeding was considered wise for many conditions. For centuries, there was a strong conviction in the effectiveness of venesection. Nowadays, most people would not accept venesection as a therapy. So does Isabella, and she searches further. Then she reads about a therapy in the form of a medallion with healing powers. According to the makers, this pendant neutralizes electromagnetic stress. It would give relief from complaints such as insomnia, fatigue, pains, etc. Although wearing such a medallion is nowadays a more accepted therapy than bloodletting, Isabella argues, it would probably not be recommended by healthcare professionals who follow the principles of evidence-based practice. But what then? And anyway, is the reasoning of today so much better than that of yesterday? The arguments of the past were also underpinned by a theory. However, in those days theory was not based on systematic clinical observations, but more on philosophical considerations. The theory underlying today’s professional guidelines is based on scientific research. A characteristic of evidence-based practice is that scientific knowledge rather than experiential or philosophical considerations are the primary determinants of treatment choices. Isabella refers to the guidelines for cardiovascular prevention. She reads that for high blood pressure, no venesection or medallion is recommended, but a lifestyle with a healthy diet, not smoking and enough exercise. If a healthy lifestyle does not work sufficiently, medication is also recommended.

This example is taken from Scholte op Reimer ([9], pp.  8–12). Reprinted with permission.

12.17  Considering Hypothetical Diagnoses

225

female SBD

male age

smoker

non-smoker

non-smoker

smoker

>50 >50 >50 >50 >50

>50 >50 >50 >50 >50

45

48 >50 >50 >50

>50 >50 >50 >50 >50

37

40

42

44

46

49 >50 >50 >50 >50

32

30

32

34

36

38

40

37

45 >50

25

30

36

44 >50

45 >50 >50 >50 >50

27

32

39

18

21

26

32

40

33

39

47 >50 >50

16

19

23

28

12

15

18

23

29

23

28

34

42 >50

10

11

14

17

20

9

11

13

16

21

17

20

24

30

23

20

23

28

34

42

22

26

32

40

50

40

48 >50 >50 >50

16

14

17

20

24

30

15

19

23

29

36

29

35

42 >50 >50

9

12

10

12

14

17

21

11

13

16

20

26

20

25

30

38

47

5

7

8

7

8

10

12

15

8

9

12

15

19

14

18

22

27

34

6

8

10

12

10

12

15

18

22

13

16

20

26

32

25

31

38

47 >50

4

5

7

9

7

8

10

13

16

10

12

15

18

23

18

22

27

34

43

3

3

4

5

6

5

6

7

9

11

7

8

10

13

17

13

16

19

24

31

120

2

2

3

3

4

4

4

5

6

8

5

6

7

9

12

9

11

14

17

22

180

2

3

4

5

6

5

6

7

9

11

8

10

12

15

20

15

18

23

28

36

160

2

3

3

3

4

3

4

5

6

8

6

7

9

11

14

11

13

16

20

26

140

1

1

2

2

3

2

3

3

4

6

4

5

6

8

10

7

9

12

15

19

120

1

1

1

2

2

2

2

2

3

4

3

3

4

6

7

5

7

8

10

13

180

1

1

1

1

1

1

1

1

2

2

3

3

4

6

7

5

6

8

10

13

160